A PRACTICE BOOK IN ELEMENTARY METALLURGY BY ERNEST EDGAR THUM, E.M. >\ Assistant Professor of Metallurgy University of Cincinnati FIRST EDITION NEW YORK JOHN WILEY & SONS, INC. LONDON: CHAPMAN & HALL, LIMITED 1917 ' IV ttbt Copyright, 1917 BY ERNEST EDGAR THUM ' PRESS OF BRAUNWORTH & CO. BOOK MANUFACTURERS BROOKLYN, N. Y. PREFACE SOME two years ago the author of this volume was confronted with the problem of presenting intensive lecture and laboratory courses in metallurgy to cooperative students in mechanical, civil, and electrical engineering at the Engineering College, University of Cincinnati. Drawing upon his own professional experience, he concluded that for such students the subject matter should most profitably be arranged to throw light par- ticularly upon the metallic materials of engineering construc- tion; how they are gained from mother nature; how they are further refined and worked; and how their chemical composi- tion and past history influence their various physical properties, and their adaptability for the duty expected of them. Some difficulty was experienced in discovering recent books of moderate price covering the field of non-ferrous metals and alloys, as well as iron and steel, which could be used as texts. Fortunately, Mr. A. P. Mills, of Cornell University, brought out his excellent book on " Materials of Construction " at about that time, which fitted our needs excellently. The problem of securing a laboratory manual, however, presented greater difficulties. The only note-books then known to us were those by H. M. Howe of Columbia University, and by Albert Sauveur and H. M. Boylston of Harvard University. These books, altho excellent, did not seem to be wholly adaptable to our needs, so they were studied carefully, together with some older laboratory exercises inherited from former instructors, and a set of mimeographed instructions hastily prepared to carry the class thru the first course. Many weak points appeared in every experiment performed. iii iv PREFACE In the first place, the limited time scheduled for the lecture course (one semester of alternate biweekly periods 45 lectures in all) allowed little or no time for the proper discussion of the theory underlying the experiments, and their applications to practical metallurgy. In the second place, the sketchy instruc- tions left a good deal to the student's own resources he would be running back and forth to the stock room continually for some needed accessory (which might not be on hand), or wasting hours of valuable time because he had neglected some minor precau- tion. It was thought that engineering students would be better benefited by following closely a set of nearly " fool-proof " instructions which would give the correct results from which they could draw conclusions and correlate allied information, rather than by spending their time devising a proper experimental procedure. With this in mind, and with the added necessity of handling classes of fifty men in a laboratory equipped with a minimum of simple, every-day apparatus, the following book was gradually evolved. Nearly the whole number of experi- ments have been performed by four different classes, and it is hoped that by now the text and instructions are sufficient and free from ambiguity. At the present time, the laboratory at the University of Cin- cinnati has available the following general equipment which is used in this particular course: Twelve small gas oven furnaces. One large hardening furnace, gas fired. One tilting crucible furnace, gas or oil fired. One lead softening pot, gas fired. One core oven. Fifteen millivoltmeters. One Wanner pyrometer. One Morse pyrometer. Ten microscopic photographic sets. Ten sets of grinding equipment. PREFACE v One Brinell hardness machine. Two scleroscopes. One Olsen impact machine. Three Howe drop hammers. One " Hy-temp " resistance furnace. Two anvils, forges, tool kits, etc. It is therefore apparent that approximately fifty students, divided into twelve squads of about four each, will have a gas furnace and pyrometer for each squad. A certain rotation of the work is of course necessary after the first few days, in order that each squad may have a separate day at the refractory experiment, No. 4, which uses the electrical resistance furnace; the hardness experiment, No. 13, which uses the Brinell machine and the scleroscopes; the metallography experiment, No. n, which involves individual microscopes; and so on. While the instructions on first glance seem to be for one particular laboratory and set of equipment, a closer inspection will discover a sincere effort to make the text applicable to considerable variations in material. Entire generality would have defeated its own purpose, as explained at some length, above. Thus, while experiment No. i must illustrate one particular furnace, and that furnace naturally is the one in use at this laboratory, still the methods of operation are uni- versally applicable to all gas furnaces. The thermo-couple experiments Nos. 6, 7, and 9 require merely a millivoltmeter or a potentiometer. And so on in fact, a very large percentage of the pieces of apparatus called for has been taken from stock in the chemical storeroom, or purchased at local hardware stores. While it is hoped that this book and its methods may appeal to officers now in charge of better equipped laboratories, and help emancipate them from the nuisance of mimeographed instruc- tion sheets, it is thought certain that it will appeal more force- fully to instructors who, like the author, are confronted with the problem of building up a respectable laboratory course with a modest amount of equipment, funds, time, and assistance. vi PREFACE It is a pleasure gratefully to acknowledge my indebtedness to Mr. E. P. Stenger, Met.E., Instructor in Metallurgy, Uni- versity of Cincinnati, for his continual care in the immediate supervision of the laboratory work, which has brought forth numberless perfections in the details of the experiments; and to Mr. Clyde William Park, A.M., Associate Professor of English, and Mr. W. Otto Birk, M.A., Instructor in English, University of Cincinnati, who have each critically examined portions of the manuscript. E. E. THUM. CINCINNATI, OHIO, June i, 1917- TABLE OF CONTENTS PAGE GENERAL RULES AND INSTRUCTIONS i EXPERIMENTAL GROUP I 1 . Furnace Operations 7 2. Oxidizing Reactions 12 3. Reducing Atmospheres and Reactions 18 4. Refractories 25 5- Slags 32 EXPERIMENTAL GROUP II 6. Thermo-couple Elements 41 7. Thermo-couple Construction 49 8. The Cooling Curve of a Pure Substance 56 9. Thermo-couple Calibration 63 10. Lead-antimony Alloys 70 EXPERIMENTAL GROUP III 11. Metallography 79 1 2. Photomicrography 86 13. Hardness 94 14. Electric Furnaces 102 15. Radiation and Optical Pyrometers in EXPERIMENTAL GROUP IV 16. Transformation Points 125 1 7. Crystallization of Steel 133 18. Hardening of Steel 141 19. Quenching Media 147 20. Tempering and Toughening 152 21. Tool Making 162 22. Metallography of Steel 167 23. Case Carburizing 180 24. Corrosion 194 vii viii CONTENTS EXPERIMENTAL GROUP V PAGE 25. Molding 209 26. Composition of Cast Iron 213 APPENDICES A. Elementary Metallurgical Calculations 225 B. Foundry Practice 269 C. Instructions for Written Work 292 INDEX 302 METALLURGICAL LABORATORY GENERAL RULES AND INSTRUCTIONS Enrollment. Students cannot be enrolled for the laboratory work until they have exhibited receipts from the registrar show- ing that they have paid the fees, and made the necessary cash deposits. A coupon ticket will be given the student in exchange for his breakage deposit receipt. Extra supplies and new apparatus to replace breakage may be purchased with these cou- pons. Whenever this ticket is reduced to a value of $2.00, the stock clerk may require an additional deposit of $5.00 before issuing further supplies. A certificate of refund will be issued on the completion of the course, covering the amount of the deposit, less any deductions punched out of the coupon ticket. Attendance. The laboratory is regularly open from 1:00 P.M. until 5:00 P.M. The roll will be taken at one o'clock by visiting the various furnaces and ascertaining the names of the men present for work. While in the laboratories, the students will be expected to conduct themselves in an industrious and orderly manner. Squad Organization. The members of the class will be divided into squads at the beginning of the semester. Each squad is intended to work as a unit at the assigned furnace, under the general direction of the captain. A student will act as cap- tain of his squad for a period of two weeks; his name will be posted on the bulletin board at the beginning of the period, and he will be responsible for the orderly prosecution of the work, the care of the equipment, and the condition of the laboratory work tables during that time. 2 GENERAL RUIES AND INSTRUCTIONS General Apparatus. There is a certain amount of equipment more or less common to all experiments; this apparatus will be furnished each squad and it is to be cared for by them. The movable parts are to be kept in the lockers assigned for that purpose and locked with the padlocks furnished by the stock keeper. The metal parts of furnaces kept on top of the desks are to be blackened at the end of each week. At the end of the semester the contents of the lockers will be inspected by the storekeeper, and any abuse of equipment or loss of material will be charged equally among the members of the squad on their breakage tickets. This equipment for each squad is as follows: Oven furnace, Pot furnace and lid, Blast lamp, Three pieces rubber gas- tubing, Bunsen burner, Button mold, 4o-mesh screen, pan and cover, Horseshoe magnet, Tripod, Wire gauze, Box of matches, Taper holder, Ring stand and clamp, Quart pail, Pint glass tumbler, Scorifier tongs, Crucible tongs, Machinist's hammer, Blacksmith's tongs for f in. Screw-driver, rounds, Two hack-saw blades, Wire-cutting pliers, Spatula, Hack-saw frame, Mixing cloth, Triangular file, Piece of emery cloth, Marking pencil, Asbestos mittens, Counter brush, Can of stove polish, Used clay crucible, 10 gm., Polishing brush. Personal Apparatus. Each student in the metallurgical laboratory should own a pair of safety goggles, which must be worn when lighting furnaces, welding at the arc, using the emery wheel, or pouring hot metal. Disregard of these elementary safety precautions will be sufficient cause for expulsion from the laboratory. Any accident must be reported immediately to an instructor, who will give necessary aid, or summon a physician. GENERAL RULES AND INSTRUCTIONS 3 Students should also carry a rule or tape of some design. The use of a slide-rule is recommended. Special Apparatus. The captain of each squad will each day receive a tray from the stock room containing the special appa- ratus needed for each experiment. He will sign a receipt for this material, and be responsible for its safe return in good order at the end of the period. Any loss or breakage will be assessed among the members of the squad. Supplies. A sufficient quantity of supplies for the per- formance of each experiment will be issued to the captain of each squad at the beginning of the afternoon. In case of care- less wastage, a new supply must be purchased with the coupon book. Supplies are not returnable, but should be reserved in a locker, as many times the materials are to be used in a subse- quent experiment. Hot crucibles may be left overnight under the furnace in an orderly array, but will be removed by the janitors at the week end. Laboratory Equipment. A certain amount of large or expensive apparatus for common use at intervals by all squads is listed in each experiment. This equipment will be located at suitable points in the laboratory. A decent consideration of the rights of the other students will insure its proper use and care. Texts. The laboratory instruction book is to be used as an auxiliary to the class-room text, and the lesson assignments which will be posted at the beginning of the year, or from time to time, should be studied at the date specified. The actual laboratory work for each day will have been posted at some preceding period. Each student is expected to have read over the instructions for the day's work before entering the laboratory in order that the work may proceed promptly and intelligently. The instructions have been worked out in detail so that correct results may be expected if the precautions are carefully observed. The references given may be consulted when convenient, to obtain a more elaborate discussion. Inspection. When the experiment assigned has been com- 4 GENERAL RULES AND INSTRUCTIONS pie ted, an instructor should be called to your furnace and the results of the experiment together with all preliminary computa- tions and sheets of neat tabulations of acquired data should be exhibited for his inspection and "O.K." On receiving this approval, students are at liberty to leave the laboratory if the squad is up to schedule. The captain will be expected to see that the laboratory tools are left in an orderly arrangement and in their correct location. All dirt is to be brushed off the table tops; movable apparatus locked up; special apparatus returned to the stock room; and the furnace cleaned on the inside, on top, and underneath. Written Work. The results of each line of procedure should be recorded in a clear, concise, and definite statement. Discus- sion of these results should follow in the order of procedure listed in the printed instructions. The statement of results and the solution of the assigned queries are the only written exercises required of the student. It is not necessary to recount the method of performing the experiment, since this would be merely a paraphrase of the printed instructions. All written work should be well and neatly done on standard paper, following the general directions for written work in Appendix C. Queries. Following the procedure in each experiment is a list of queries, some of which are for more advanced students. Each student should therefore consult the bulletin board for information as to the particular queries required from his class. These are to be worked out and written up by each man indi- vidually. Problems and computations must be solved in an orderly manner, with all steps shown and noted, so that they can be readily checked over. The instructor can observe the results of the experiment, and will grade the student's skill in manipulation during the laboratory period. The queries, on the other hand, are designed to show him whether the individual student comprehends the theory underlying the experiment and can interpret the results. A high grade will be given work evidencing thought and indi- viduality, even tho the correct solution is not attained. For GENERAL RULES AND INSTRUCTIONS 5 this reason, it is also clear that squad work on the written part cannot be accepted. Stand on your own feet! When you are past school age it will not always be so easy to turn to a " wizard " at your elbow for an explanation of the knotty points. Curves. Acceptable curves must be neatly drawn to scale and lettered with India ink on standard cross-section paper. Title the curves and the coordinates, following the directions in the appendix. Sign and date each curve sheet. We want curves made by engineers. Submission of Written Work. All written work, queries, curves, and tabulations of data, are to be bound together in a regular Metallurgical Laboratory binder of cardboard and submitted to the instructor in charge at the end of each bi- weekly school period. This binder would therefore contain reports on all the work since the last submission of data. After examination and grading, this work will be returned to the student for his own reference, and is to be retained subject to call by the English Department. It is important that the covers and con- tents be preserved intact, as they must be resubmitted in case of conditions or unsatisfactory work. Grades. Final grades for laboratory courses will be based upon daily marks on ist, the queries; 2d, the results of the laboratory work, inspected day by day; 3d, written work other than queries; 4th, general conduct and industry during the laboratory period ; 5th, attendance. It is essential for the orderly conduct of affairs in the labor- atory that all notes be kept up to date and be submitted promptly according to schedule. In case a student fails in punctuality, his work will receive but 75 per cent the mark it otherwise would obtain. EXPERIMENTAL GROUP I FOREWORD TO THE STUDENT In Experiments Nos. i to 5 inclusive are presented a number of general experiments which form an introduction to the study of metallurgy. Laboratory work in metallurgy is largely confined to the study of the behavior of solids at moderately high heat. It is essential, therefore, that the first experiment should instruct the student in the furnace provided for obtaining these temperatures. Systematic metallurgy naturally starts with the production of the pure metals from the compounds or salts found in nature. The chemical processes involved in these operations are quite dissimilar to the test-tube reactions with which the student is already familiar. Experiments Nos. 2 and 3, on oxidizing and reducing reactions, are given in order that the student may adjust his state of mind to the metallurgical viewpoint. In the reduction of metallic oxides, high temperature reac- tions proceed continuously in certain containers. These fur- naces or crucibles must be able to withstand the high heat with- out melting or corroding. Experiment No. 4 on refractories is designed to show the relative infusibility of some common brick-making materials. These selfsame refractory materials, however, are normally present in most metallic ores; and Experiment No. 5 on slags is designed to demonstrate how these relatively infusible mate- rials can be removed from a furnace in a molten condition. Each experiment endeavors to present in a logical manner a clear statement of the underlying principles of the process under consideration, together with notes as to its commercial applications. Carefully follow the details of the procedure, and the proper results are assured. Then attempt to visualize the grand scale of metallurgical operations in the light of your laboratory practice. EXPERIMENT NO. 1 FURNACE OPERATIONS Object. This experiment is an introduction to the opera- tion of gas furnaces. General Explanation. The oven furnace in use in this lab- oratory (Fig. i) consists of an iron box lined with fire brick, with three gas burners operating thru each side wall, near the bottom. A short distance above the bottom is placed a fire-clay shelf which constitutes the working floor of the furnace and forms the roof of a com- bustion chamber in which the incoming gas and air combine. The hot products of combus- tion pass thru the opening be- tween the edge of the shelf and the side walls into the furnace " laboratory " above, and thence escape thru two vents in the ceiling. Natural gas is the fuel used; sufficient air under press- ure is also supplied to become intimately mixed with the gas and support perfect combustion. The maximum temperature may be attained by close regulation of the gas and air an excess of gas will as effectively cool the flame as an excess of air. Therefore, a neutral atmosphere will generally give a higher calorific intensity than either an oxidizing or a reducing flame. Methods of measuring the temperatures in a furnace are of 7 FIG. i. Oven Furnace. 8 EXPERIMENTAL GROUP I course very important. One of the earliest systems in use is that of observing the melting-point of Seger cones, so called after their originator. These cones are made of various mix- tures of alumina, silica, alkaline oxides, etc., whose melting or softening temperatures are fairly well determined. Each cone has a number stamped on the side corresponding to a defi- nite composition. When the cone is gradually heated up and approaches its melting-point, the mixture softens and the peak of the cone starts to bend over. When the apex is bent over in an approximately horizontal direction it has attained the tem- perature corresponding to the number which it bears, as given in the following list: No. Temp. No. Temp. No. Temp. No. Temp. 022 600 070 960 9 1280 -29 1650 021 650 o6a 980 IO 1300 30 1670 O2O 670 050 IOOO II 1320 31 1690 OIQ 690 040 IO2O 12 I3SO 32 1710 018 710 030 IO4O 13 1380 33 1730 017 730 02(1 IO6O 14 1410 34 I7SO 016 750 Old I080 15 1435 35 1770 oi 50 790 Id IIOO 16 1460 36 1790 014*1 815 2d 1 1 20 17 1480 37 1825 0130 835 3 1140 18 1500 38 1850 0120 855 4a 1160 19 1520 39 1880 oirfl 880 5 1180 20 I53P 40 1920 oioa 900 6a 1200 26 1580 4i 1960 090 920 7 1230 27 1610 42 2OOO o8a 940 8 1250 28 1630 Note that all temperatures in this list are degrees Centigrade. Special Apparatus. The special apparatus needed is as follows: Seven scorifiers. Supplies. The supplies needed are as follows: Seger cones Nos. 021, 017, 013(1, o8a, 040, ia, 40. Sand. Charcoal. FURNACE OPERATIONS 9 Procedure, a. Exhibit your receipts to a laboratory officer showing proper payment of laboratory fees and deposit. This is an absolute requirement which the student must fulfil before he will be allowed to do any work in the laboratory. b. Properly registered students will be grouped into squads for laboratory work. The personnel of these squads will remain fixed for the semester and can be changed only by permission. c. A laboratory officer will read the " General Rules and Instructions " (pages i to 5), and identify the general appa- ratus comprising the equipment of each squad. He will also demonstrate the method of lighting the furnace. d. After the blower has been started, open the door of the furnace and place a flame on the inside of the furnace by means of the taper holder. Open the gas cock wide; and when the flame flares up, remove the taper holder from the furnace. The valve with the hand wheel regulates the gas supply, while the valve which is operated by means of a lever regulates the air. The flame of the burning gas will now be projecting from the door and the vents of the furnace. e. Turn on the full air supply and then throttle down the gas very slowly, constantly shortening the flame until the mixture of gas and air is such that the combustion takes place entirely within the furnace and under the hearth. Be sure that the furnace door is open and that no one is standing in front during these manipulations, because when the mixture of gas and air is just right, an explosion wave carries the flame from the furnace openings back to the gas burners in the combus- tion chamber. Note the tune when the flame begins to burn satisfactorily, and close the door. /. Fix each of the Seger cones upright in a scorifier of sand. Arrange these scorifiers in the furnace in a definite order so that the location of each cone is known. Note the condition of the cones from time to tune while the furnace is heating up. Record the time when the apex of each cone is bent over hori- zontally; the furnace has then reached the temperature corre- sponding to that number. Note also the color of the light 10 EXPERIMENTAL GROUP I radiated by the interior of the furnace. An experienced man can give a remarkably close estimate of the temperature of his furnace or of a piece of hot metal by its color. g. After the furnace has apparently reached its maximum temperature, request an instructor to inspect its condition, remove the scorifiers, and place a piece of charcoal in the fur- nace. Then throttle down the amount of gas carefully until the charcoal glows brightly. This glow shows that oxidizing conditions prevail in the furnace, since the charcoal is now burning in the excess of air present. By turning on an excess of gas, the reducing atmosphere is produced, which is evidenced by a thin gas flame burning on the exterior of the furnace. Note the condition of the charcoal under these conditions. The neutral condition occurs when perfect combustion is taking place; that is, when neither an excess of gas nor air is present. This condition is obtained when neither the gas burns at the exterior of the furnace, nor the charcoal glows on the hearth. h. Re-read this entire experiment carefully, and be sure that you understand the text, have performed all the manipula- tions, and can answer the required queries. i. When the experiment is finished, exhibit the data to a laboratory officer. If it is satisfactory, return the special apparatus to the stock-room, and clean up your premises. Queries, a. Briefly describe the construction of the furnace, together with the gas and air piping, and illustrate the descrip- tion with a neat pen-and-ink sketch. b. Draw a curve showing the rate at which the furnace was heated, according to the instructions on page 5, and Appendix C. Note the colors corresponding to various tem- peratures on this curve. c. What causes the difference in the appearance of the gas flame under oxidizing and reducing conditions? d. Explain fully the reasons for the statement that a neutral flame will give the highest temperatures. e. Cite various heating operations where it is important FURNACE OPERATIONS 11 to maintain a reducing, an oxidizing, or a neutral atmos- phere. /. Why are Seger cones placed in a scorifier and not on the bottom of the furnace? g. Discuss the limitations in using Seger cones for pyrom- eters. EXPERIMENT NO. 2 OXIDIZING REACTIONS Object. The object of this experiment is to reproduce in the laboratory some of the oxidizing reactions used in metallurgy. General Explanation. By the term " oxidizing reaction " the metallurgist means a chemical interchange which will convert some metal or metallic compound into a corresponding metallic oxide. Oxygen is the most abundant element known, comprising perhaps 50 per cent of the entire earth mass. In the free state it forms 23 per cent of the atmosphere, by weight. Combined, it forms eight-ninths of the water, half the sand, and a large proportion of most of the existing mineral and organic substances. Oxygen has taken part in important reactions during the for- mation of the earth materials, quartz or sand being the result of the oxidation of the metalloid silicon: = Si0 2 + 196,000. Silica is an important substance, inasmuch as large quantities are found pure, and doubly so as it is an essential component of all volcanic rocks. Water is likewise the product of the oxidation of the metal hydrogen: 2H 2 +O 2 = 2H 2 O+i38,ooo (liquid). Owing to the presence of an excess of free or combined oxygen in all processes, unless carefully excluded, and to the strong affinity of most metals for oxygen, the wasting away of pure metals by the formation of their oxides is a comparatively rapid process. This oxidation, when it goes forward slowly as the result of exposure to atmospheric agencies is known as " corro- 12 OXIDIZING REACTIONS 13 sion." Oxidation of heated metal in rolling or forging is much more rapid, and causes large losses as " mill scale "; the reac- tion is as follows: Because of these facts, the common metals are found in nature in an oxidized condition, the principal ores of iron, for instance, being hematite, limonite and magnetite. The oxidation of pure metals is a common operation in mak- ing pigments, or in recovering rare metals. In the first case, is extensively practiced to make the body for the best white paints. In the second instance, the metallic lead produced in lead smelting carries with it the gold and silver contained in the furnace charges, which metals are often separated by a proc- ess called " cupellation," wherein the molten lead is exposed to a current of air which oxidizes it to litharge. This lead oxide is then drawn off or " skimmed " from the remaining metal while the rarer metals will resist oxidation to the last. The reac- tion is as follows: 2Pb+O 2 = 2PbO+ 101,600. Sulfur is an element somewhat similar in chemical proper- ties to oxygen in fact, ore bodies of metallic oxides found near the surface often change by degrees into bodies of the corre- sponding sulfides, the oxidized portions having been formed from the original sulfide enrichment by the action of percolating surface water containing oxygen and carbon dioxide in solu- tion. In fact, most metallic ores are deposited in the first instance as sulfides from hot alkaline solutions rising from the depths thru fissures in the earth's crust. The oxidized ores are distinctly secondary that is, have been formed from the primary sulfide deposits at a later time. The elimination of sulfur from these sulfide ores is the prime reason for " roasting " operations (see pages 262, 548, 568, Mills, " Materials of Con- 14 EXPERIMENTAL GROUP I struction "), and they may be classed under oxidizing reac- tions because the metallic sulfides are changed into oxides, often with an increase in valence, while the sulfur which is eliminated burns to sulfur dioxide (SCb). Non-ferrous smelting plants liberate such enormous quantities of this gas as to become verit- able nuisances in spite of the most modern appliances for puri- fying the smoke. Special Apparatus. The special apparatus needed is as follows : One piece of f-in. iron pipe, about 6 in. long. One roasting dish. Two test tubes. One test-tube holder. One funnel. One blowpipe, and burner-tip. Supplies. The supplies needed are as follows: Zinc button, about 20 gm. Lead button, about 125 gm. Heavy iron wire, about 8-gage, 3 ft. long. Thin iron wire, about 4o-gage, 12 in. long. Two scorifiers. One piece of soft glass tubing, 6 in. long. Laboratory Equipment. The laboratory equipment needed is as follows: Anvil. Two bucking-boards and mullers. Piece of |-in. pine board. Package of filter papers. Coke. Pyrite. Charcoal. Sodium carbonate crystals. Ammonia. Distilled water. OXIDIZING REACTIONS 15 Procedure. NOTE : Start parts b and c simultaneously. a. Make a " rabble " for stirring pyrite in this manner: Heat one end of the 8-gage iron wire, flatten it out thin on an anvil, and bend the thin end over square, forming a small hoe, the blade of which will be about f in. wide by \ in. high. 6. Polish the fine wire until it is bright, twist it into a spiral around a pencil, then insert it into the iron tube and place the whole in a cold furnace. At the end of the afternoon examine the condition of the wire, and reserve it for comparison with the results of Experiment No. 3. c. Break up to rice size sufficient pyrite (FeS2) to cover the bottom of the roasting dish f in. deep; place the dish in a cold oven furnace near the open door. Bring the heat up to a low red with an oxidizing flame, keeping the front door open all the time. Note the condition of the ore and stir carefully but thoroly with the rabble at ten-minute intervals. CAUTION: Do not overheat the roast. d. Near the end of the laboratory period remove the dish, cool, and pulverize the contents until it will all pass thru a 4o-mesh screen. Place this ground material on a piece of paper, and separate magnetite (FeaCU) and pyrrhotite (FeeS?) with a magnet. Test for soluble iron sulfate (FeSCU) by placing the non-magnetic part in a test tube and boiling with distilled water. Filter and add a few drops of ammonia to the filtrate. Ferrous hydrate is white; a darker colored precipitate shows the presence of soluble ferric compounds. Test the magnetic part and also the residue held on the filter paper by igniting a little of each with sodium carbonate on charcoal in a reducing blowpipe flame. Crush the resulting bead, place it on a clean silver coin and moisten. Black silver sulfide forming will evidence the presence of undecomposed sulfides in the roasted material. e. Powder 500 gm. of coke. Place 20 gm. of zinc in a scori- fier and heat in a pot furnace in a strongly oxidizing flame until it catches fire, then smother the flame with powdered coke. /. Draw the glass tube down to form an air nozzle about 16 EXPERIMENTAL GROUP I i mm. in diameter. Place a scorifier containing about 125 gm. of lead upon a used crucible, set inverted in the pot furnace. Cover the furnace and heat the lead to a dull red color. Skim off any dross with a sliver of wood, and adjust the flame of the blast lamp by carefully controlling the gas supply until the surface of the lead button remains clear. Then carefully pro- ject a blast of air thru the glass nozzle upon the molten lead with sufficient force just to dimple the surface, but not forceful enough to splatter the hot metal. The formation of litharge (PbO) by reaction between the molten lead and the oxygen of the air must proceed rapidly in order to counteract losses by volatilization of the oxide, and the formation of lead silicates by reaction with the material of the scorifier. When the sur- face of the lead is nearly covered with the molten oxide, pour the whole rapidly into a button mold. When cool, break off the brittle oxide on a bucking-board by hammering the lead button into a cube. Pulverize the oxide to 40 mesh and screen out any shot lead. Retain the powdered litharge. CAUTION: Avoid breathing poisonous lead fumes. g. Re-read this entire experiment carefully, and be sure that you understand the text, have performed all the manipulations, and can answer the required queries. h. When the experiment is finished, exhibit the data and results to a laboratory officer. If it is satisfactory, return the special apparatus to the stock-room, and clean up your premises. Queries, a. Describe and explain the results of procedure b. Write an equation showing the reaction, and prove your assumption by testing the products. b. Describe and explain the results of procedure c. c. What are the end products of the roasting operation c? Write reactions showing how each was formed. Write all the reactions performed in the blowpipe test for undecomposed sulfide. d. Describe the experiment with zinc. Why should zinc emit copious white fumes before it catches on fire? What is the OXIDIZING REACTIONS 17 composition of these fumes? What is the action of the powdered coke on the burning zinc? e. Describe and explain the appearance of the scorifier after procedure /. /. If you had some pig iron (essentially an alloy of iron with from 2 to 5 per cent carbon) and treated it as in procedure /, what would have happened? Give cogent reasons for your statements. g. Had the iron wire weighed 3 oz. and 25 per cent of it been converted into the oxide, how many B.t.u. of heat would have been evolved? How many grams of oxide would have been formed? h. How many cubic feet of air at normal conditions will be required to roast i Ib. of FeS2 completely into Fe20a and S02? i. Describe the roasting of a zinc ore. j. Describe the oxidized ores of iron, and note the principal localities in which they are produced. EXPERIMENT NO. 3 REDUCING ATMOSPHERES AND REACTIONS Object. The object of this experiment is to reproduce in the laboratory some of the reducing reactions used in metallurgy. General Explanation. An oxidizing atmosphere is one which contains free oxygen (such as air) or contains a gas which can easily furnish oxygen by the decomposition of its molecule, for example, carbon dioxide: C0 2 -CO+0-68,o 4 o. Such an atmosphere will give up oxygen to the surrounding sub- stances; the intensity of the oxidizing reaction varying with the temperature, pressure, and relative amount of the reagents available. A reducing atmosphere, on the contrary, contains an excess of gaseous molecules which possess an " affinity " for oxygen; for example, acetylene (C2H 2 ), carbon monoxide (CO), or a mixture of hydrocarbons like natural gas. Such an atmosphere tends to abstract oxygen or similar elements from the substances it surrounds; the reaction, as before, varying with the three conditions mentioned. The meaning of the terms oxidation and reduction was for- merly restricted to the addition or subtraction, respectively, of oxygen from substances. The scope of the words has now been extended to include the addition or subtraction of other elements (sulfur, particularly), or the raising or lowering of the valence of polyvalent elements. The third type of atmosphere, viz., the neutral atmosphere, is one which has neither oxidizing nor reducing effect, as these terms have been denned. In the laboratory, pure nitrogen furnishes a convenient neutral atmosphere, for it is an extremely 18 REDUCING ATMOSPHERES AND REACTIONS 19 inert gas under all ordinary circumstances. In commercial furnaces a neutral atmosphere is ordinarily a mixture of oxidiz- ing and reducing gases in a state of chemical equilibrium at the existing temperature; the state is one of balanced activity in which the oxidizing effect of one is continually undone by the reducing action of another. The mechanical working of iron and steel ingots into mer- chantable shapes is most easily done at high temperatures. Even those special operations which may be performed cold, like wiredrawing, cold-rolling and pressing, must occasionally be interrupted by a heat treatment to relieve dangerous internal stresses and to restore ductility to the substance. It has been shown in Experiment No. 2 that such operations in the open air are accompanied by the formation of an excessive amount of mill scale on the surface of the piece, the relative quantity of which increases rapidly with the decreasing size of the objects, because the ratio of superficies to mass becomes correspondingly larger. Heating and annealing furnaces for small or thin articles must therefore operate with a neutral or reducing atmosphere. It has been mentioned that most ore bodies are masses of metallic oxides or sulfides mixed with a variable quantity of barren rock materials called gangue. The minerals containing the valuable metals can be obtained from the ore in a relatively pure state by mechanical processes called " concentration. " (See Richards, " Ore Dressing.") Enormous concentrating mills employing hundreds of men and treating thousands of tons of ore every day are in operation in many mining centers. Leaving out of consideration at this time the treatment of any remaining impurities in the ore, which matter will be the sub- ject of Experiment No. 5, the commercial production of the metal from oxides and sulfides is effected by reducing reactions operated on a grand scale. Oxidized ores, either found in nature or produced by roast- ing, are in most cases reduced to metal in the " blast furnace." (See Hofman, " General Metallurgy," pp. 384, 475.) This is a vertical shaft furnace into which solid ore, flux, and fuel are 20 EXPERIMENTAL GROUP I charged at the top. Air for the combustion of the fuel is blown in under pressure thru openings called tuyeres which are located in the walls of the furnace near the bottom. The intense heat generated at this zone reduces and melts the metal in the ore, which is tapped molten from the bottom of the fur- nace. In the case of iron oxide, much of the reduction of hema- tite is accomplished by the furnace gases, rich in carbon monoxide ascending thru the porous column of descending ore, thus * FeO+C-+Fe+CO-36,5 4 o (at 800 C). In copper and lead blast furnaces the charge consists quite generally of mixtures of oxides and sulfides, which react upon each other as follows: - 52,540. Any surplus of oxide may be reduced by the carbon in the fuel, thus: CuO + C = Cu +0-8540, altho it is usual to provide rather a large excess of sulfides, which will melt without great change. This alloy of metallic sulfides will collect together as a substance called " matte," which is withdrawn in a molten condition from the furnace, is separated from slag and metal, and then further smelted in machines called " converters." (See Mills, " Materials of Con- struction," p. 552.) Special Apparatus. The special apparatus needed is as fol- lows: One piece of f-in. iron pipe, about 12 in. long, threaded. One f-in. gas cock. One o.i-gm. trip balance and weights. Spatula. One button brush. One test tube. See Stoughton, " Metallurgy of Iron and Steel," p. 27. REDUCING ATMOSPHERES AND REACTIONS 21 Supplies. The supplies needed are as follows: One piece thin iron wire, about 4o-gage, 16 in. long. One oooo graphite crucible. Twenty-five gm. hematite. Thirty gm. litharge. Twenty gm. copper oxide (CuO). Two lo-gm. clay crucibles. Laboratory Equipment. The laboratory equipment needed is as follows: Coke. Bucking-boards and mullers. Borax glass. Galena. Salt. Covellite. Dilute hydrochloric acid. Procedure, a. Prepare all mixtures before starting the fur- nace, and place them into the furnace immediately after light- ing. Maintain a slightly reducing atmosphere at all times, as determined by Experiment No. i. b. Sandpaper the thin wire until it is bright, twist it into a spiral around a pencil and insert it into the iron tube. Screw the gas cock on the pipe and connect to a gas supply with a rubber hose. Fix the pipe with a condenser clamp in such a manner that the flames from the lid of the pot furnace will strongly heat that part which contains the wire. Pass a slow current of gas thru the tube during the entire operation; heat strongly for an hour, remove the pipe from the flame, and cool without shutting off the gas current. CAUTION: protect the rubber hose from the heat. c. Remove the wire and examine it critically, comparing its condition to that of a short untreated piece reserved for the purpose, and to that' of the oxidized wire from Experiment No. 2. Observe the color, luster, and general appearance. Test the 22 EXPERIMENTAL GROUP I ductility by counting the number of times it must be bent back and forth to break it. Test the hardness with a file. Test the magnetic qualities and solubility in dilute HC1 of any scale which can be loosened. Prepare a quenching bath by filling a pail with tap water, and place it as close to the furnace as possible. Grasp a short piece of the original wire and of the heat- treated wire side by side in the jaws of a small tongs, and heat them to a bright red in the reducing flame issuing from the pot furnace. Quench the wires by quickly plunging them at the high heat into the cold water. Speed in transfer is the prime essential. Test both pieces as before. d. Make up a charge with 35 gm. litharge (PbO), using that made in Experiment No. 2, and test the reducing action with galena (PbS) . Pulverize all the materials to 40-mesh, weigh out the required amount of the reagents, add about 2 gm. of fine coke, mix thoroly by rolling together on a rubber mixing cloth, and place the charge in a clay crucible. Fill the balance of the crucible with salt. Heat rapidly from the cold in the oven furnace, remove and pour into a button mold at a bright red heat, first examining the melt to see if it is thoroly molten. Examine the button carefully, break the metallic lead loose with a hammer, brush it free from foreign matter, shape it into a cube with a hammer, and weigh. CAUTION: Always wear goggles when examining or pouring molten materials. Better be safe than sorry. e. Test the reactions 2CuS+O 2 = Cu 2 S+SO 2 +69,36o; - 26,440. Covellite (CuS) on being heated will readily break down into the compound Cu 2 S which will then react with the copper oxide. Figure the amount of Cu 2 S necessary to reduce the oxide, and then figure the amount of covellite required to produce the Cu 2 S according to the first equation. Use about 20 gm. of CuO. Thoroly mix the pulverized materials, adding about 2 gm. REDUCING ATMOSPHERES AND REACTIONS 23 pulverized coke, transfer to a clay crucible, cover with salt, heat strongly in the pot furnace, and pour at a bright white heat. Separate the copper and weigh. /. The most fusible pig iron (alloy of 95.7 per cent iron with 4.3 per cent carbon) melts at 1135 C., while pure iron melts at 1500 C., a temperature beyond the capacity of the gas fur- naces. Hematite can be reduced by coke according to the fol- lowing reaction: Fe2O3+3C > 2Fe+3CO 108,120. The infusible iron will absorb any excess of carbon existing in its neighborhood and be converted into the more fusible pig iron. Therefore, weigh out about 25 gm. hematite, and a com- puted amount of pulverized coke (thru 40 mesh) to effect the reduction according to the above equation, together with excess of coke sufficient to form the most fusible pig iron. Assume the coke to be 90 per cent carbon. Then add approxi- mately 10 gm. powdered borax glass, and mix thoroly by rolling on a mixing cloth. Transfer the charge to a oooo graphite crucible, and cover with a small amount of borax glass. Heat from the cold in the oven furnace, transfer the hot crucible quickly to the pot furnace after pouring e, and heat for thirty minutes to the highest temperature attainable. Pour quickly into a button mold, and carefully remove any material which adheres to the crucible with a sharp knife. Crush, separate the iron from the glass with a magnet, and weigh. g. Re-read this entire experiment carefully, and be sure that you understand the text, have performed all the manipula- tions, and can answer the required queries. h. When the experiment is finished, exhibit the data and results to a laboratory officer. If it is satisfactory, return the special apparatus to the stock-room and clean up your premises. NOTE. These last two instructions have been repeated in Experiments Nos. i, 2, and 3, and will be understood as 24 EXPERIMENTAL GROUP I appearing at the end of each future day's work, even tho they do not again appear in print. Queries, a. Tabulate in five columns the results of pro- cedure b and c with the iron wire. b. Make up a neat tabulation of the results of procedure d, e, and /, showing the weights of the reagents used, the theo- retical weight of the metal to be recovered, and the actual weight found. c. Discuss the causes of any discrepancies between the theoretical and the actual weights of metal recovered. d. What is the function of the salt in procedure d and e? Of the borax in /? e. Do you get any matte in the melt for copper? How do you know? /. Give the computations for procedure / in full. g. What is the theory of valence? State some experimental facts which are satisfactorily explained by this theory. h. What is the pronunciation of tuyere? of gangue? of matte? i. Explain the differences noted in the condition of the iron wire before and after annealing; also before and after quenching. What happens during these operations? j. How many cubic feet of methane (CH-i) at standard condi- tions would be required to reduce 10 kg. of Cu2O to metal? How much heat would be evolved or absorbed during the process? EXPERIMENT NO. 4 REFRACTORIES Object. The object of this experiment is to study the prop- erties of the more common refractories. General Explanation. Since a large proportion of metallurgi- cal operations are conducted with molten materials at quite high temperatures, it is evident that the containers and the furnaces must be constructed of substances which will with- stand such extreme heat without being melted. A substance of high -melting point is said to be a " refractory " material, and bricks, crucibles, muffles, and other shapes made of it are termed " refractories." Some pure substances melt at temperatures attained in the electric arc (3500 C.db). Graphite (C), calcium oxide (CaO), and magnesia (MgO) are common instances. Other substances much used in refractory materials melt in the neighborhood of 2000 C.; such are alumina (Al 2 Oa) at 2000 C., chromite (Fe2O3) x (Cr 2 O 3 ) J/ at 2200 C., fire-clay (Al 2 3 ) z (SiO 2 ) z , in the region of 1800 C., silica (SiO 2 ) at 1600 C., and carborundum (SiC) which dissociates at 2250 C. These temperatures are for the most part approximations. The difficulties attending such measurements may be realized when it is known that the melting-point of silica has been var- iously reported at from 1200 C. to 2000^ C. Very pains- taking and thoro work by Day and Shepherd (American Journal of Science, 1906, p. 273) shows that solid silica changes into solid tridymite with large increase in volume at 775 C. 25, and that solid tridymite melts into an extraordinarily viscous liquid at 1610 .15. The changes take place very slowly; the just molten tridymite can only be distinguished 25 26 EXPERIMENTAL GROUP I from the just solid tridymite by microscopic examination under polarized light. As a matter of fact, the roofs of open-hearth furnaces manufacturing steel are made of silica brick, and are steadily run for weeks at temperatures in excess of 1600 C. without failure of the brickwork. In fact, only silica brick, of all cheap refractories, possesses the required compressive strength at such high temperatures as to be available for use in the construction of the low-arched roofs of these steel furnaces actually working at a higher temperature than the true melting- point of the brickwork itself. (See Mills, " Materials of Con- struction," p. 389.) It has been found that very small percentages of impurities will lower the melting-point of these refractories to a marked degree, even tho the melting-point of the impurity be higher than that of the refractory. Thus, a little lime is used as a binder in the manufacture of silica brick pure silica having no cohesion in itself to enable it to retain its shape after molding. But the amount of lime must be no greater than absolutely neces- sary in order to prevent harmful effects on the refractory proper- ties of the brick itself. For this reason it is evident that only the purest and most uniform rocks can be utilized in the manu- facture of all refractories. Of course other properties than fusi- bility must be considered in the choice of furnace materials, such as its coefficient of expansion, strength and toughness at low and high temperatures, behavior under rapid change in temperature, and conductivity of heat and electricity. For a discussion of such features, the student is referred to Havard's book on " Refractories and Furnaces." A good refractory must not only resist the action of the heat, but must be chemically inactive; that is to say, it should resist the corrosion of any liquid material with which it may be in contact. For instance, a silica or fire-clay brick is rapidly wasted away when in contact with a slag or liquid melt con- taining a large percentage of lime. This slag, however, is per- fectly resisted by a magnesite brick. Such facts as these have led to the following classification: REFRACTORIES 27 Name of refractory. Composition. Basic: Lime CaO Dolomite (CaO)(MgO) Magnesia MgO Alundum A^Os Bauxite A1 2 3 +#SiO 2 Neutral: Chromite Fe and Cr oxides Graphite C Carborundum SiC Acid: Fire-clay (Al 2 O 3 ) n (Si0 2 ) m Silica SiO 2 A magnesite brick typifies the basic refractories, a silica brick, the acid; and chromite is neither. It will be noted that the prominent basic refractories magnesia, dolomite and lime are made of the alkaline earth oxides, each of which is the anhydride of a very active base. On the other hand, the most prominent acid refractory, silica, is made of the anhydride of the various silicic acids. Hence the metallurgist uses the same terms, acid, and base, as are used in elementary chemistry. The student should realize, however, that the ordinary tests for acid and base cannot be applied in metallurgical reactions. He has been accustomed to distinguish an acid from a base by their action on litmus and other indicators and by their behavior during electrolysis. But the chemistry of aqueous solution vanishes at temperatures above 100 C., and the best the metallurgist can say about the chemical action at high tempera- ture is that the substance which is a base in aqueous solution tends to act as a base at high temperatures, and that the acid of aqueous solutions acts as an acid at high temperatures. Bear in mind at all times, however, that the terms acid and base are relative terms only. Thus, iron oxide will act as a base with the very acidic silica forming iron silicates, but will act as an acid with the very basic lime forming lime ferrates. Again, bauxite brick will resist corrosion and union with silica to a very high 28 EXPERIMENTAL GROUP I temperature, and therefore may be called basic; but in the pres- ence of lime, easily fusible compounds will be formed. As an example of the use of such refractories take the case of the so-called " basic open-hearth furnace " (Mills, " Materials of Construction," p. 387 et seq.). Given the fact that this process operates with a slag containing a high percentage of lime, then the lower portion of the furnace must evidently be of the same nature as the slag it holds in order to resist corrosion, and consequently is made with magnesite brick side walls and granular dolomite hearth. Magnesite brick has not the mechan- ical strength at high temperatures to permit the construction of the arched furnace top, which is therefore made of silica brick. At the point of contact of these two different kinds of brick a parting of neutral brick must evidently be placed to prevent the rapid formation of magnesium silicates, and conse- quently a " neutral " course of chromite brick everywhere sepa- rates the two. Special Apparatus. The special apparatus needed is as fol- lows: Glass stirring rod. Stiff wire brush. One loo-mesh screen and pan. One 200 c.c. glazed porcelain pestle and mortar. Fragment of 2-in. round graphite electrode. Hand- saw. o.i -gram trip balance and weights. Brass mold for Seger cones. Spatula. Supplies. The supplies needed are as follows: One 2oo-c.c. beaker, loo-c.c. of cylinder oil. 25 c.c. vaseline. Fragments of the following No. i refractories: Silica brick. Magnesite brick. Bauxite brick. Fire-clay brick. REFRACTORIES 29 Laboratory Equipment. The laboratory equipment needed is as follows: Bucking-board and muller. " Hy-temp " or other electrical resistance furnace with proper electrical control. Optical pyrometer to read 1800 C. Solid reagents in suitable containers: Alumina, Hematite, Fe 2 O 3 Silica, SiO2 Kaolinite (Al 2 O 3 )(SiO 2 )2 Lime, CaO, freshly burned. Procedure. NOTE: The following instructions cover the examination of relatively pure refractories, such as are used for the manufacture of first-class fire-brick, and even the highest temperatures will show little if any effect upon the Seger cones. Various squads, therefore, should add increasing percentages of flux to the brick during the grinding. In this way a series of experiments can be exhibited showing the effect of impuri- ties upon the softening point of the purer refractory. A series of additions up to the following maxima are suggested : For silica brick: 200 per cent A^Os plus 5 per cent Fe 2 Os. For bauxite brick: 10 per cent SiO 2 plus 10 per cent Fe 2 Os. For magnesite brick : 20 per cent kaolinite. For fire-clay brick: 3 per cent CaO plus 6 per cent Fe 2 Os. a. Interview the instructor to find the amount of impuri- ties he wishes you to mix with the pulverized materials. b. Heat the cylinder oil over a Bunsen flame, and dissolve in it about one quarter of its bulk of vaseline. This will be used as a binder. c. Clean the bucking-board and screen as follows: Brush the bucking-board and muller vigorously with the wire brush, covering the entire surface at least twice. Take several frag- 80 EXPERIMENTAL GROUP I ments of the material to be crushed, and grind them down, using the whole surface of the muller, and a considerable area of the bucking-board. Brush the fines into the sieve, screen, and waste the fine materials. Brush off the board and muller again twice vigorously with the wire brush. Clean out the sieve and pan by rubbing with the fingers, and jarring attached particles loose. d. Take a fragment of silica brick, weigh it, and add the necessary amount of impurities (if any is prescribed by the in- structor). Pulverize the whole until it completely passes thru the loo-mesh screen. Avoid spilling the fines in trans- ferring from the bucking-board to sieve, and vice versa. e. Thoroly mix this powdered material in the mortar with a little of the oil-vaseline solution. When mixed properly, with the correct amount of binder, the powder will look dry and granular, but will stick together readily when pressed between the fingers. /. Mold three Seger cones by tamping this granular material into a clean brass mold. A good cone should be free of cracks, have plane sides, and a sharp point. Place these cones on the clean hearth of a cold oven furnace. g. Repeat procedure c to f with each of the other refrac- tories, bauxite, magnesite, and fire-clay. Be particularly care- ful to clean the entire equipment thoroly between moldings, so that the purity of the mixtures may not be impaired. h. When all the cones are properly made, heat the furnace gradually to a maximum heat, and hold at this temperature during the remainder of the period. Observe the temperature of the furnace at the end with an optical pyrometer. (See Experi- ment No. 15.) Cool in the furnace. i. On the following day examine and note the condition of the cones, particularly as to the condition of the tips. Then select the best and sharpest cone of each class for further experi- menting, reserving the others. j. Saw a f-in. wafer from a 2-in. round graphite electrode. With a knife, cut the squad-number on the side of this wafer, REFRACTORIES 31 and notches in one surface so it can be grasped easily by small tongs and lowered into the crucible of the electrical resistance furnace. k. Place the four cones on this wafer in the crucible in such a manner that they are not in contact with their surroundings. Then close the furnace and gradually raise the temperature to 1800 C., carefully maintaining this degree as nearly constant as possible during the balance of the period. One squad member should be delegated to this work, while others proceed to the next experiment. The furnace tender should observe the temperature of the furnace with an optical pyrometer at fifteen-minute intervals, and make the necessary electrical adjustments under the advice of the instructor. /. On the following day, remove the cones, examine and note their condition. Place the wafer containing the cones on the designated shelf alongside the work of the other squads for pur- poses of comparison. Exhibit here also one of the cones of each kind which has been baked in the muffle furnace, but not heated to the high temperature. This will serve as a reference, indica- ting the original condition of the cones. Also attach a card giving the squad-number and personnel, and the composition of each cone. Queries, a. Describe the results of the various squads, sketching the condition of representative cones after exposure to 1800 C. b. Give the percentage composition of each refractory when it first shows appreciable softening at 1800 C. c. What sort of brick should be used for the stack of an iron blast furnace? For the bosh? For the crucible? For the stoves? d. Give a short account of the process of manufacture of refractory brick. EXPERIMENT NO. 5 SLAGS Object. The object of this experiment is to show how to produce fusible mixtures, or slags, from infusible ore substances, or gangue, and to show the effect of chemical composition upon the properties of slags. General Explanation. In the smelting of ores, the operator must recover his valuable metal in as pure a condition as possible. This necessitates the separation of the metal from any impuri- ties which may be present in the ore. Some of these impurities may be volatile; but others will be recognized as highly refrac- tory and would scarcely melt before the furnace itself. Economy of operation demands that the contents be removed from the fur- nace in the fluid state; therefore, the problem the metallurgist must face is, " How can these substances be smelted into a liquid and easily fusible slag? " The impurities in ores are of widely varying composition and character their general nature follows that of the barren " country rock " surrounding the mineralized ore body. The gangue of an ore body occurring in a limestone formation would probably be high in limestone (CaCOs), and therefore basic (see p. 27). On the other hand, ores in igneous rocks would probably be very high in silica (SiO2), and therefore acidic. In many ores, valuable metals may be regarded as impurities; for instance, in smelting an ore for copper, the iron and zinc content is slagged and thrown away with the other impurities. This is necessary because of the fact that in the present state of metallurgical art there is no method known for saving these valuable metals as a commercial by-product. The more important gangue materials of various ores may 32 SLAGS 33 be listed as follows, ranked approximately from basic to most acidic : Basic Alkali Na 2 O and K 2 Baryta BaO Lime CaO Magnesia MgO Iron oxide Manganese oxide Alumina A1 2 03 Boric oxide B 2 Os Titanium oxide Silica Si0 2 Acidic As noted before on page 27, acidity and basicity are relative terms; for instance, alumina acts as a base toward substances below it, but as an acid toward those above. This property is commonly said to be " amphoteric." Silica is the universal acid radical in rock forming minerals, and practically all slags contain it in large quantities; for this reason, most slags may be termed " silicate " slags. They may also contain small amounts of the minor acids, such as P 2 Os, SO 3 , Sb 2 Os, As 2 0s and TiO 2 . The basic oxides are usually present in larger variety three or four of the common oxides are ordinarily present in considerable quantity. As a matter of fact, the gangue usually does not contain the oxides in the free state, with the exception of quartz (Si0 2 ), but the ore is a mixture of minerals of definite composition, such as feld- spar (K^A^SieOie), mica (2KH-2MgFe-2AlFe-Si3Oi 2 ), calcite, CaCOs, etc. Chemical analysis of the ores allows one to separate the constituent oxides in the manner listed above before making a calculation of a furnace charge. (See Appendix A.) In order to remove these more or less infusible materials from the furnace in a liquid state, one avails himself of the facts covered by the general statement that " Given two pure sub- 34 EXPERIMENTAL GROUP I stances, the melting-point of either is lowered by the addition of certain quantities of the other." The logical conclusion of this statement is that there must be one or more mixtures of lower melting-point than either of the constituents. The com- position of the mixtures melting at the lowest points does not correspond to that of a chemical compound; they are the so-called " eu tec tic " mixtures appearing under the microscope as an 1512 1500- 1400 1300 1200 1100 1000 aSiOa Plus Diopside Inversion Points Diopside Inversion Points Mixed j Crystals j of I MgSiO 3 I Diopside mixed la Diopside I Crystal 20 40 60 80 60 40 20 FIG. 2. Equilibrium Diagram, CaO-SiO^ : MgO-SiQj. Reprinted from Hofman " General Metallurgy," by permission of the McGraw-Hill Book Co. intimate mixture of minute crystals of two different substances. (Eutectic means " easily melting.") This general statement holds for alloys of elements, oxides, or more complex compounds. As an illustration, wollastonite, CaOSiO 2 , melts at 1512 C., enstatite, MgOSi0 2 melts at 1524 C. However, a eutectic consisting of 27.5 per cent of MgOSiO 2 and 72.5 per cent of CaOSiO 2 melts at 1350 C., and another eutectic of 67 per cent MgOSiO 2 and 33 per cent CaOSiO 2 melts at 1375 C. The only true chemical compound formed by SLAGS 35 these two silicates is called augite (CaOMgO-2Si02), and it melts at 1381 C., lower than either constituent, but higher than either eutectic. (Hofman, General Metallurgy, p. 444, shows a diagram representing the melting-points of all com- bination of these two silicates, which is reproduced above, Fig. 2.) Quickly cooled slags are glasses, or solid solutions of the oxides composing them and of the compounds which could exist as stable substances at the temperatures in question. On more slowly cooling, the stable compounds tend to separate out of the mother liquor, and may be identified by the expert under microscopic examination. The constitution of slags containing several different oxides is therefore extraordinarily complex, and to a large extent is yet unexplored territory. It is certain, however, that the old fashion of assuming all the bases to range together on the one hand, and the acids joining on the other to make one complex silicate molecule, is based upon no ground- work of fact. However that may be, it is nevertheless conveni- ent to say that the problem of slag-making usually consists of determining the proper amount of basic material to add as a flux in order to make a readily fusible compound, or eutectic mixture, with the excess of acid present in the gangue of the ore, or vice versa, as the case may be. What the relative proportion of acid to base may be, and which oxides shall be regarded as acids and which as bases, are questions ordinarily answered by the metal- lurgist on the basis of his experience obtained in the management of certain furnaces. Fortunately, considerable latitude is allow- able in most processes, and universal agreement or scientific exactitude is therefore not necessary since a furnace will take care of itself and smooth out its own irregularities to a certain extent. Temperatures which are moderate for metallurgical furnaces are not obtainable in the ordinary gas muffle, and therefore special acids and bases, too expensive to use in bulk, are utilized in the laboratory to make more fusible slags. Some of these combinations are molded into the familiar Seger cones of experi- 36 EXPERIMENTAL GROUP I merit No. i, p. 8. Such substances as bicarbonate (NaHCOs), litharge (PbO), and borax glass (Na2B4O?) are extensively used as fluxes in assaying. - Special Apparatus. The special apparatus needed is as fol- lows: One o.i-gm. trip balance and weights. Spatula. Supplies. The supplies needed are as follows: Three scorifiers. Six 5-gm. crucibles. Laboratory Equipment. The laboratory equipment needed is as follows: Optical pyrometer. Charcoal. Two bucking-boards and mullers. Trays of reagents, as follows : Bicarbonate. Litharge. Borax glass. Hematite. Silica. Fresh burned lime. Fluorite. Procedure, a. Fill separate scorifiers half full of each of the following substances: * Laboratory bases. I. Bicarbonate (NaHCO 3 ). II. Litharge (PbO). Laboratory acid: III. Borax glass (Na 2 B 4 7 ). Weigh the scorifier and bicarbonate both before and after heat- ing, in order to note any loss in weight. SLAGS 37 b. Compute the amounts of reagents necessary to make the following monosilicate slags : IV. 2PbO+SiO 2 = (PbO) 2 (SiO 2 ). Use 10 gm. SiO 2 . V. Fe 2 O 3 + C + SiO 2 = (FeO) 2 SiO 2 + CO. Use 8 gm. SiO 2 . c. Replace half the iron oxide in the latter by the more basic and more fusible alkali Na 2 O, thus: (Use 6 gm. SiO 2 .) VI. 4 NaHCO 3 +Fe 2 O 3 + C + 2SiO 2 d. Replace half the silica by the more fusible boric oxide; using 5 gm. silica: VII. Na 2 B 4 O 7 + 2NaHCO 3 + SiO 2 + Fe 2 O 3 + C = (Na 2 O) 2 (FeO) 2 SiO 2 -B 4 O 6 +H 2 O + 2CO 2 + CO. NOTE: In all the above, the amount of the various ingre- dients should be computed previous to the laboratory period, and the calculations should be presented for inspection at the beginning of the afternoon before preceding with the weighing. Use charcoal for the carbon, assuming it to be 100 per cent pure. All materials should be pulverized to 40 mesh, carefully weighed, thoroly mixed, and each mixture placed into a 5-gm. crucible, properly marked. e. Tne most fusible lime-silica mixture contains 37 per cent CaO and 63 per cent SiO 2 . Make up two samples of this mix- ture, each containing about 15 gm. SiO 2 and add to one approxi- mately 10 per cent of pulverized fluorite (CaF 2 ). Mix well and place in separate crucibles, marking them VIII and IX. /., Range these containers in the cold muffle according to the plan, Fig. 3. Heat strongly in a neutral flame, observing conditions at fifteen-minute intervals and being careful not to cool the furnace by keeping the front door open too long. As the contents of the containers become thoroly molten, with- draw each from the furnace and pour into a button mold, observ- ing the approximate temperatures from the color of the muffle 38 EXPERIMENTAL GROUP I (Experiment No. i). Upon request, a laboratory officer will check your estimate by means of an optical pyrometer. CAUTION : Safety goggles must be worn when examining and pouring molten material. g. After VII is removed, close and drive the furnace one Door^ FIG. 3. Location of Slag Containers in Oven Furnace. hour at maximum heat; estimate the temperature, remove all the remaining crucibles, and examine the contents. h. See Procedure h and i of Experiment i. It is important that the instructor be shown all melts at the end of the afternoon. Queries, a. Make up a neat tabulation giving the following information: Name of material, chemical formula, weights of materials mixed, color before heating, color after heating, melt- SLAGS 39 ing-point, viscosity, general appearance and condition of the slag, general appearance and condition of the container. b. Give the equation for the formation of VI and the detailed computations for this mixture. c. What happens to I? Compare your weighings with the theoretical loss. What would happen should water be added to the product? d. What is the result of adding CaF 2 to a calcium silicate? Explain this. How could you determine whether the lime silicate is an acidic or a basic slag? e. Did you get any metal from melt II? Why should such metal appear? /. Give reasons for any corrosion of the fire-clay scorifiers or crucibles. g. What would be the composition in weight per cent of the slag if 27 per cent of the number of basic molecules in IV were replaced by the same number of molecules of MgO? EXPERIMENTAL GROUP II FOREWORD TO THE STUDENT The experiments of Group II are presented in an effort to acquaint the student with the best methods of measuring accu- rately a moderately high temperature. A very large part of modern metallurgical progress can be directly attributed to information gained by a study of the behavior of pure metals and their alloys at high temperatures. It will be apparent that any investigations, either in the works or in the laboratory, must be predicated on strict temperature control. Experiments Nos. 6, 7, and 9, will therefore indicate practi- cal methods whereby excellent pyrometers may be easily con- structed and calibrated. Since the method of calibration involves the determination of the melting-points of two or more metals, a study of the melt- ing-points of various binary alloys is naturally suggested which will directly utilize the work already completed. This is the sub- ject matter of Experiments Nos. 8 and 10, the latter of which constructs an equilibrium diagram of the simple lead-antimony system. This family of alloys, altho of small commercial importance, is of great educational value since all the members have low melting-points, they can easily be handled by novices and their equilibria throw much light on the more important and complex system of alloys between iron and carbon, which includes the various irons and steels. 40 EXPERIMENT NO. 6 THERMO-COUPLE ELEMENTS Object. The object of this experiment is to study the action of thermo-electricity as influenced by the character of the metals used. General Explanation. If two wires of different chemical composition or physical constitution are welded together at one end and the loose ends are connected to a sensitive instrument capable of showing a small electric current, a current will be indicated when the welded junction is heated. The amount of current flowing will be directly proportional to the electromotive force maintained by the heat at the welded ends; and inversely proportional to the resistance of the circuit. This statement follows directly from Ohm's law that the ratio between the difference of electrical potential, or electromotive force E, exist- ing between two points in a circuit, and the current C produced thereby, is a constant and equal to the resistance R of the cir- cuit. In symbols - = R Or, transposing, In the case under discussion the resistance of the circuit will be the sum of the resistance of the two wires welded together (elements), the connecting wires leading to the electrical meter, the meter itself, and the various contacts, wire to wire. The cause of this current is primarily due to the phenomenon known as " contact electromotive force." If two disks of 41 42 EXPERIMENTAL GROUP II unlike metals, such as zinc and copper, are placed side by side and a positively charged plate is suspended above, the plate will deflect toward the copper as soon as the two disks are moved together so that they touch at any point. The contact is a source of electrical energy which projects positive particles of electricity called electrons in one direction and negative electrons in the opposite, thus charging the zinc disk with positive elec- tricity and the other with a like amount of negative. The positive charge on the zinc repels the like charge on the positively charged plate suspended above, while the negative charge on the copper attracts the plate, the result being as noted that the plate swings over the copper disk. It might be thought that the electricity on the zinc, being at a higher potential, would discharge across the junction to neutralize a like amount on the copper plate as soon as a second point comes in contact. The flow of such a current, however, would give an instance of perpetual motion, evidently an absurd- ity when it is considered that all points of contact between the two plates are alike in being sources of electromotive force, and all points are tending to keep the two plates at a different potential. Each point in contact projects electrons like every other point in contact, and no electrical discharge can take place across this generator. A current can be caused to flow, however, in two ways. In the first place, if two plates of unlike metals are touched at their upper ends, and the lower ends immersed in a solution of certain inorganic substances, a current will be produced. Such a current is not an instance of perpetual motion, however, for a certain definite quantity of chemical energy existing in the sub- stances in solution is transformed into electrical energy by the degeneration of compounds of high latent heat into others con- taining lower amounts. Currents produced in this manner are the cause of the action of all primary electric batteries and probably of most of the corrosion of iron and steel. In the second place, a current will be produced from a source of contact electromotive force by joining the ends of two wires THERMO-COUPLE ELEMENTS 43 of unlike metals, and maintaining a difference of temperature between the two junctions of the wires of this so-called " thermo- couple." Currents such as these are the basic principle under- lying the action of most instruments for the measuring of moderately high temperatures. Such instruments are called pyrometers. The value of the contact electromotive force of any junction varies with the temperature, and can therefore be expressed as a function of the absolute temperature of that junction; therefore the measurement of the electromotive force (or the resistance being unchanged, of the current it induces), by suitable electrical instruments gives an indication of the temperature at that time and place. Peltier found that if a current of electricity be passed thru a weld of two metals, such as copper and iron, a certain quantity of heat is absorbed, or a like amount is evolved at the junction should the current flow from the copper to the iron or vice versa. This fact, known as the " Peltier effect," furnishes an explanation of the phenomena in an ordinary thermo-couple. If, there- fore, a circuit is composed of one wire of copper and one of iron welded together at the ends, a current flowing past one junction from the copper to the iron will absorb a certain amount of heat, which will be evolved in like amount at the other junction where the current flows from iron to copper. Or, without impressing an external current, if a difference in temperature be maintained at these two welds, the circuit will act as a heat engine, absorb heat at the hot junction and, transforming it into electrical energy, conduct this electricity along the wires to the cold junction, and there reconvert the electric energy into heat, which is evolved at that point. The action of this engine will cool off the region of the hot terminal and heat the region of the cold until both arrive at the same temperature, at which time the contact electromotive force at the two junc- tions will be equal and opposite, and action will cease. The energy necessary to heat the fire end and to refrigerate the cold furnishes the work necessary to maintain the existence of the electrical current. 44 EXPERIMENTAL GROUP II Naturally the amount of electrical current generated will depend upon the metallic combination as well as upon the differ- ence of temperature between the hot and cold ends. The value of any combination for use as a thermo-couple depends upon several other characteristics which will be discussed later, but primarily the thermo-couple should produce a large electromotive force, and therefore a proportionately large current for a moderate difference in temperature in order that it may be delicate or sensitive. Several different combinations of metals can be compared by measuring the current generated when the ends of the wires are at a constant difference in temperature. The easiest way to maintain these temperatures is to immerse the welded end in boiling sulfur, while the other ends are in an ice bath. Special Apparatus. The special apparatus used is as follows: One electrical meter. One retort clamp. One ring stand. Various wires, i4-gage, 24 in. long, labeled with me- tallic tags as follows: Soft iron. Music wire (hard). Music wire (annealed). Alloys 343, 183, 166, German silver. Copper. Nickel. One piece of lamp cord, 24 in. long. One blast burner. Supplies. The supplies needed are as follows: Ice. Twenty-five gm. sulfur. One 6-in. by i-in. test tube. One f -in. by 3-in. test tube. THERMO-COUPLE ELEMENTS One two-hole cork for i-in. test tube. One 6-in. piece of small glass tubing. One piece of asbestos paper, 2^ in. by 36 in. One piece of asbestos paper, 6 in. square. 45 Laboratory Equipment, is as follows : The laboratory equipment needed Spool of asbestos string. Bare arc, with proper hood. Syrupy sodium silicate. Soldering outfit, including: Soldering tool. Soldering paste. Solder. Piece of cloth for wiping. FIG. 4. Home-Made Arc Welder. Procedure, a. Make up a thermo-couple as follows : Polish both ends of the wires with emery paper, clamp the tip ends of two wires alongside each other in a vise, and twist them tightly together, two or three turns. Fuse the twisted ends by bringing 46 EXPERIMENTAL GROUP II them slowly down to the flame of an electric arc, which may be arranged as in Fig. 4, until a globule of molten metal forms at the tip. Syrupy sodium silicate makes a satisfactory flux. Always wear darkened goggles and look thru the "noviweld" glass plate fixed in the hood to protect the eyes from the harmful effect of the intense light. b. Solder the leads to the couple as follows: Heat the FIG. 5. Apparatus for Boiling Point of Sulfur. soldering tool, plunge it into the soldering paste, wipe off, and rub solder over the end of the heated copper nose until a smooth bright surface results. If the tool is too hot the bright so-called " tinned " surface oxidizes and becomes dull in luster. Clean and " tin " the ends of the wires to be soldered in a like manner, twist tightly together, and fix with a drop of solder. THERMO-COUPLE ELEMENTS 47 c. Set up the apparatus as shown in the illustration Fig. 5 as follows: Wrap the strip of asbestos paper tightly about the large test tube, covering the middle part of its length, A. Cut a small round hole in the center of the square piece of asbes- tos paper B, slip it over the lower end of the test-tube to act as a hood over the flame, and hold it in place with a ring. These precautions are to insure a region at the center of the test-tube which is well insulated against heat transfer so it will main- tain constantly the temperature of boiling sulfur. The cork C in the top is to prevent the vapor catching fire; the glass tube D is required to prevent an increase in pressure in the test-tube which would raise the boiling-point and then burst the appara- tus. The small test-tube is inserted thru the cork to act as a protection tube for the thermo-couple and serves to prevent the wires from sulfidizing. d. Melt the sulfur carefully, using only enough to fill the tube for 3 cm. Adjust the flame so that the vapor rises nearly to the top of the test-tube, where it condenses and runs back down the sides. A point in the center of these fumes will now be main- tained at the boiling-point of sulfur, 444.7 C. e. Insulate the wires from one another at the welded end by running a " figure 8 " with asbestos string for about 6 in. Wrap a bit of asbestos about the wires to form a cork E closing the small glass protection tube, so that the welded end is at the bottom. Immerse the other ends in the ice jar F and connect the free ends of the leads to the binding posts on the meter, taking particular care to make tight connections. Guard against grounds and short circuits. /. Compare all wires against iron, noting the needle deflec- tion for each metallic combination to one-tenth unit. Pay particular attention to the direction in which the current is flowing. g. Substitute boiling water for ice in the pail at the cold end and note the effect on the thermo-couple which gave the highest reading. h. Test the contact electromotive force of a couple made 48 EXPERIMENTAL GROUP II from, hard and annealed music wire. Solder the junction instead of welding in this instance. Queries, a. Make a neat tabulation arranging the metals and alloys tested in a series from maximum to minimum read- ing, showing the numerical values determined by the meter. Such a list might read: Positive Zinc +6 Lead +5 Tin +2 Iron o Copper i Silver 5 Gold -7 Negative b. What is meant by the term absolute temperature? Ex- press 329 F. in absolute temperature, Centigrade. c. Give the names of the units for the measurement of electromotive force, resistance, and current. Define their magnitude. d. Write an equation showing in symbols the amount of current flowing in a thermo-couple and its attachments. e. If the unit on the meter is supposed to be i millivolt, what current would flow in the steel-german silver couple if the total resistance of the circuit be 300 ohms? Which way does the current flow? /. Suppose the leads were lengthened so as to increase the resistance by 10 ohms. What would the current then be? g. What does the meter actually register, current or electro- motive force? What causes the needle to deflect current or electromotive force? Why do you think so? EXPERIMENT NO. 7 THERMO-COUPLE CONSTRUCTION Object. The object of this experiment is to make a thermo- couple to be used as a pyrometer in future laboratory investiga- tions. General Explanation. Note: A good discussion of this subject is given in Chapter IV of Burgess, " Measurement of High Temperatures." As noted in Experiment No. 6, the value of a metallic com- bination for use as a thermo-couple depends upon several char- acteristics, which may be listed as follows: First, the thermo-couple should produce a high electro- motive force for a moderate difference in temperature in order that it may be delicate or sensitive. This requirement was investigated in Experiment No. 6. Second, the wires themselves should be homogeneous, so that parasitic currents may not affect the accuracy of the electrical indications. Parasitic currents may be denned as those uncon- trolled and variable electrical effects arising from any one of several causes. This requirement demands wires which are perfectly uniform in chemical composition thruout their length, and are free from hard or crystalline spots or any other variation in physical constitution. Wires of exactly the same chemical composition will show a contact electromotive force should they possess a different physical structure. This fact was demonstrated in Experiment No. 6 where a thermo-couple made of hard and annealed music wire was tested and showed the production of a considerable electrical current. The soft- ened or annealed wire was made from the hard wire by merely heating it to about 1000 C. and then slowly cooling it. It is important, therefore, that couples for precision work should 49 50 EXPERIMENTAL GROUP II be very carefully constructed and handled a mere bending of the wire may produce sufficient " cold- working hardness " to affect the accuracy of the indication. Such hard spots are sources of parasitic currents. Wires of superior quality may be had from the Electric Alloy Company, Morristown, N. J., and may be tested for homogeneity by methods described by W. P. White, in the " Journal of the American Chemical Society," 1914, p. 2292. Third, the wires should be stable at elevated temperature, stable chemically and stable physically. The most perfectly prepared wire would rapidly become useless if it oxidized readily, or alloyed with the small quantities of metallic vapor existing in most industrial processes. Physical stability requires that crystalline growth at the high temperatures in use be very slow. Wires which change suddenly in any physical property (that is, those which are made of metals possessing " allotropic transformations " at elevated temperatures), are evidently not stable in the sense here used. All physical properties vary with the temperature thus, the hardness becomes less as the temperature increases, the specific gravity becomes smaller with rising temperature, and so on. But a large number of metals and alloys will have sudden and discontinuous changes in their properties at certain fixed temperatures; for instance, pure iron loses its magnetism suddenly when the rising tempera- ture passes 760 C. Evidently, such a profound rearrangement of the configuration of the iron molecule as to cause it to lose its magnetism will doubtless have a considerable effect upon such a property as contact electromotive force. Fourth, the wires should have a high melting-point in order that their useful range may be as extensive as possible. Fifth, the wires should not be so expensive as to prohibit their common use. Sixth, the wires should be small tho rugged, in order that the temperature of the hot end will rapidly attain and closely follow that of the surroundings without being so deli- cate as to require gentle handling. THERMO-COUPLE CONSTRUCTION 51 Considerations such as these point to the fact that an ideal thermo-couple is not a simple matter to construct indeed, has not yet been discovered. The best couples in use are those of pure platinum against platinum alloyed with 10 per cent of rhodium or iridium. These platinum couples (ordinarily desig- nated as Pt Rd or Pt Ir) are sensitive, homogeneous, refrac- tory and rugged, but have the disadvantage of being very ex- pensive. Again, while they are quite stable physically at high temperatures and resist oxidation perfectly, they are rapidly destroyed in a reducing atmosphere by alloying readily with volatile metals and metalloids which may be present. It is important that an excess of oxygen be always present when plati- num couples are in use in order that any volatile substances may be readily oxidized and thus rendered incapable of form- ing an alloy. Altho the noble metals make the best couples, on account of the high cost their use is confined to scientific investigations and the construction of standard thermometers. Cheap couples for use in commercial practice are therefore usually made of wires of base metal alloys. Many useful couples are manu- factured of alloys whose principal metal is nickel (melting- point 1450 C.) combined with various percentages of chromium, iron, copper, manganese, and cobalt. The wires furnished for this experiment are Hoskins' Manufacturing Company nickel- chromium alloys Nos. 183 and 343, i4-gage. This combination gives a high electromotive force. The wires are specially made for thermo-couple use and are homogeneous in constitution; they resist oxidation up to 1000 C. and oxidize but slowly at 1100 C. ; the wire is relatively inexpensive; the gage is small enough to follow a rapid change in temperature, but still not so small as to become rapidly weakened by surface oxidation. The wires composing the thermo-couple must evidently be properly insulated from one another, and also covered carefully to protect them from the corrosive action of their surroundings. In case the couple is to measure the temperature of a bath of metal, the protection tube is necessary to prevent the wires from 52 EXPERIMENTAL GROUP II alloying with the molten metal. The protection tube must in this case also be able to resist the corrosive action of the bath, in order that it may not be rapidly destroyed and at the same tune vitiate the purity of the melt. Base metal wires must further be protected from oxidation or sulfidation. Noble metal couples have very small quartz tubes or beads strung along their length to prevent a short circuit from one wire to the other should they touch, and then an outer tube of porce- lain or quartz covers the whole. Quartz has a very low coef- ficient of expansion, and can be heated or quenched very rapidly porcelain, on the other hand, must be heated and cooled with extreme care. Tubes made of either of these substances are quite brittle, however. Base metal tubes should not be used for noble couples on account of the fact that metals volatilize to a minute extent at temperatures below their melting-point, and this metallic vapor readily alloys with the superficies of the thin elements of the couple, thus destroying the chemical homogeneity of the wires. It is often difficult to procure an adequate thermo-couple protection tube. Gillett recommends that base metal couples may be insulated from the harmful effects of sulfur fumes by a thin glass tube. The wires can be protected from alloying with a lead, aluminum, antimony, or zinc bath by closely wrap- ping them with asbestos string and then covering with retort or high temperature cement. For melts of copper or brass, a nickel-lined tube of siliconized graphite is necessary. An iron protection tube can be used in a lead or cool aluminum bath. A common arrangement for protecting portable couples to be used about the works to measure furnace temperatures is made by inserting the wires, properly separated by porcelain or quartz beads into a seamless protection tube of nickel or steel, screwed into a wooden handle which also contains concealed binding posts for joining the couple to wires leading to the electrical meter. Such tubes, however, are rapidly oxidized under most conditions, and their replacement often amounts to quite an expenditure./ THERMO-COUPLE CONSTRUCTION 53 In laboratory work it is usually possible to maintain the cold end of the thermo-couple at a constant temperature by immersion in an ice bath. Manufacturers of pyrometers have adopted various expedients to eliminate the necessary correction in portable thermo-couples where the cold junction is at a variable temperature. Special Apparatus. The special apparatus used is as follows: One electrical meter. Supplies. The supplies needed are as follows: One piece of Hoskins' alloy 183, 24 in. long. One piece of Hoskins' alloy 343, 24 in. long. One piece of lamp cord, 24-in. long. One piece of okonite tape, 12 in. long. One pint pail filled with kieselguhr. Laboratory Equipment. The laboratory equipment needed is as follows: Core oven at 100 C. Bare arc, with proper hood. Syrupy sodium silicate. Spool of asbestos string. Soldering outfit, including: Soldering tool. Soldering paste. Solder. Piece of cloth for wiping. Can of retort cement. Carpenter's hand-saw. Jig saw. Brace and f-in. bit. Piece of board, ^ in. by 6 in. Vise. Procedure, a. Weld the not junction as instructed in Pro- cedure a of Experiment No. 6, or in the electric butt welder under 54 EXPERIMENTAL GROUP II the direction of a laboratory officer. In any case, present the weld for inspection before proceeding. b. Starting about 18 in. from the weld, wrap asbestos string tightly around one wire, working toward the weld. When this is covered as close up to the joint as possible, lay the bare wire alongside, and continue wrapping out to the bead on the weld. Cover the bead with four short strings put on cross- wise, the ends to be held under the circular wrapping. Work back from the weld, covering both wires for a distance of 12 in.; then the bare wire separately for another 6 in. c. Solder the lead wires to the thermo-couple as directed in procedure b of Experiment No. 6. Wrap the joint tightly with a short piece of electrician's tape. d. Connect the leads to a meter, and heat the fire end gently in a gas flame to test for a possible break or short circuit. During the heating, grasp the welded end and the soldered connections in pliers and work gently back and forth, looking for any oscilla- tion in the needle, which will indicate a faulty electrical circuit. e. Dampen the wrapping and rub on a little retort cement, working it well into the string until the excess cement becomes somewhat dry and granular. Remove this excess, dampen and slick over the insulation with a little water, and dry slowly in a muffle furnace or drying oven held at about 100 C. Put on three thin coats in this manner and lay aside to dry over night. A very good paste for filling in asbestos string is sold by the Johns-Manville Company under the trade name of " Retort Cement." The Quigley Furnace Specialty Company also make a satisfactory cement called " Hytempite." Hoskins recommends a mixture of 10 parts silica thru 2oo-mesh screen, 2 parts burnt fire-clay thru 2OO-mesh screen, 5 parts acid-free sodium silicate, mixed with hot water sufficient to make into a creamy paste. /. The dried couple is to be slowly heated from cold to bright red in a muffle in order to anneal the wire and properly to bake the insulation. g. Construct an ice bath by packing a stout glass jar, 3 or THERMO-COUPLE CONSTRUCTION 55 4 in. in diameter, into a pint pail with some insulating ma- terial such as kieselguhr, mineral wool or asbestos fiber. Cut a suitable piece of wood or asbestos lumber to serve as a cover to the pail, with a f-in. hole in the center for the entrance of the wires. Queries, a. Define the words " metal " and " metalloid." Distinguish between the terms " noble metal " and " base metal." List several chemical elements in these classes. b. What is the diameter of a i4-gage wire? c. Why is it necessary to keep the cold end at o C.? d. How is it possible to arrange a couple for use when the cold end is at some other temperature than zero? e. Would lead wire make a satisfactory thermo-couple element? Why? Would iron? Why? EXPERIMENT NO. 8 THE COOLING CURVE OF A PURE SUBSTANCE Object. The object of this experiment is to obtain the cooling curve of pure lead. General Explanation. One of the most important means of investigating the properties of pure metals and their alloys is by an examination of their heating and cooling curves. Such curves are constructed by taking a small portion of the substance to be studied and observing and recording the temperature of the mass at uniform intervals of time during a uniform heating or cooling. These observations, when plotted in the form of a curve, will show whether the temperature of the mass rises or falls uniformly with the continuous addition or abstraction of heat. The heat which a body absorbs serves either to raise the tem- perature of the mass or change its physical condition. That portion of the heat which results in an increase in temperature of the body is called " sensible heat," inasmuch as such a gain in heat is apparent to the physical senses of the observer. Precise calorimeter experiments have revealed the fact that the relation between temperature and sensible heat can be expressed by a continuous function such as H = (S+kt)t, where H is the total heat required to raise the temperature from oto*, S is the specific heat of the body measured at o, k is a constant, and / is the final temperature of the body. 56 THE COOLING CURVE OF A PURE SUBSTANCE 57 Evidently, if heat were supplied to the body at a uniform rate, the temperature would rise continuously, and if the temperature were plotted against time, a smooth rising curve would result. Or, if sensible heat were abstracted from the body at a uniform rate, a time- temperature curve would again be a smooth falling curve. Such a curve is called a " cooling curve." However, we find that when a body is melting, vaporizing, or otherwise suffering an abrupt change in physical properties, a quantity of heat is absorbed which disappears without chang- ing the temperature of the body. This heat absorbed during a change of state is called " latent heat," because it is transformed into the work necessary to change the configuration and dis- position of the molecules in the body ; but it is again liberated in equal amount when the reverse change takes place. The latent heat of fusion, for instance, is the heat absorbed by the work of driving the molecules far enough apart so that the necessary mobility is gained to change the body from a solid into a liquid; work done against the internal forces tending to hold the mole- cules more rigidly together as a solid. In short, it is the heat required to change a hot body in the solid state into a hot body at the same temperature but in the liquid state. It is easy to comprehend that these large latent heats will notably affect the shape of a cooling curve. From these considerations it would seem that should the cooling curve be continuous and smooth, following closely a~ regular course, all the heat abstracted during cooling is furnished at the expense of a fall in temperature of the body; that is to say, it disappears as " sensible heat." These curves, however, frequently show horizontal portions or " arrests " which denote that at that temperature all of the heat constantly radiating is being supplied by internal changes in the alloy itself; that is, it is being supplied 'by the evolution of a certain amount of " latent heat." The freezing of water is a case in point which falls within ordinary experience. Water exposed to a low temperature will radiate heat into the cold atmosphere, and consequently the 58 EXPERIMENTAL GROUP II temperature of the water will steadily drop until the freezing- point, o C., is reached. At this time, despite the constant radiation of heat into the colder surroundings, the temperature remains at zero until all the water is changed into ice. When the whole mass is entirely solidified, further radiation of heat is again at the expense of the sensible heat of the ice, and the temperature again falls until it finally arrives at the low tem- perature of the refrigerant. During the solidification period, however, the temperature remains stationary, and the radia- tion is at the expense of the so-called " latent heat of fusion " of the ice, which equals an amount of heat which would raise the temperature of an equal amount of water thru an interval of 80 C. The cooling curve of liquid lead or any pure metal near the solidification point is entirely analogous to that traced by freez- ing water. Let a properly insulated thermo-couple be immersed in a crucible of molten lead which is then allowed to cool in still air, and the reading on the dial recorded every thirty seconds, as follows: Time. Scale Reading. Time. Scale Reading. Time. Scale Reading. Min. Sec. Min. Sec. Min. Sec. Beginning 12.0 4 3 5-4 9 4-8 3> 10.7 5 5-4 9 30 4-2 i 9-5 5 3 5-4 IO 3-6 i 30 8-5 6 5-4 10 30 3-0 2 7.6 6 30 5-4 II 2.6 2 30 6.9 7 5-4 II 30 2.1 3 6.2 7 30 5-4 12 1.8 3 30 5-7 8 5-4 12 30 i-5 4 5-5 8 30 5-2 13 1.2 These readings will appear as shown in Fig. 6 when plotted on cross-section paper as ordinates to the scale of i unit=i cm., against times as abscissae to the scale of i minute = i cm., and connected with a smooth curve. Following the reasoning given in the case of water, we know that during the time 4 min. 30 THE COOLING CURVE OF A PURE SUBSTANCE 59 sec. to 8 min., the stationary temperature denotes the pro- gressive solidification of the lead. Therefore, whatever the temperature of the other readings noted may signify, we know that the reading 5.4 corresponds to the melting-point of lead. From a similar series of observations made with a mercury or air thermometer, the melting-point of lead has been determined to be 327 C. In this instance 5.4 units equals Milli voltmeter Readings > M) *. OS 00 t* Z \ \ V niBQ,^ \ o o \ 1 \ \ Fig. 7 COOLING CURVE OF ANTIMONY X >-,,-< w>_ ^s N Fig. 6 COOLING CURVE OF LEAD ^ \ N^ \ ^ x 1-1 12 10 4 6 8 10 Time in Minutes > 10 20 30 40 Time Intervals FIGS. 6 and 7. Cooling Curves of Lead and Antimony. Special Apparatus. The special apparatus used is as follows: One electrical meter. One oooo graphite crucible. Three hundred grams pure lead. Supplies. The supplies needed are as follows: Ice. Charcoal. 60 EXPERIMENTAL GROUP II Laboratory Equipment. The laboratory equipment needed is as follows: Bucking-board and muller. Procedure. CAUTION: New graphite crucibles should be heated -very slowly the first time they are used. Rapid heating will expel moisture from the crucible walls so rapidly that large pieces of graphite will be flaked away. a. With a sharp knife, cut the letters " Pb " in the side of the crucible. b. Melt the lead slowly in a pot furnace, and when it is entirely molten, skim off any floating oxide or slag with a splinter of wood. Then fill the remainder of the crucible with crushed charcoal. Avoid raising the temperature much above the melting-point of the lead or breathing the fumes arising from the pot. c. Place the cover on the pot furnace. Slowly heat the dried and annealed thermo-couple constructed in Experiment No. 7 to about the temperature of the molten lead in the flames issuing thru the opening in the furnace cover. Lower the hot junction into the center of the melt, holding the wires in position with a condenser clamp and ring stand. d. Bend the thermo-couple wires so that the cold junction can be immersed in the ice bath constructed in Experiment No. 7 and connect the leads to the electrical meter. Remove the blast lamp as soon as the hot junction attains the tempera- ture of the bath. e. Read the dial on the meter to one-tenth unit at fifteen second intervals, record and plot the observations in pencil, as in Fig. 7. Continue the observations to about two units below the solidification point. After each reading, the meter may be tapped gently with the fingers to start the needle. /. When the cooling has proceeded far enough, make sure that the couple is frozen solidly in the metal by attempting to lift it out. Then light the blast lamp and heat the crucible slowly at a uniform rate until the lead is again melted. Read THE COOLING CURVE OF A PURE SUBSTANCE 61? the meter at intervals of fifteen seconds, record and plot the heat- ing curve. g. Shut off the flame and observe and plot another cooling curve according to procedure e. h. Present curves and data to a laboratory officer for inspec- tion and approval. i. When a satisfactory check has been secured, melt the couple free from the bath, allow the metal to cool in the crucible, remove the crucible from the furnace and return it to the stock room. Melt away any lead adhering to the insulation by holding it in the flame of the blast lamp. Cover the couple with a fresh layer of cement, and lay away to dry over night. Queries, a. Define specific heat. What is the unit of heat? b. The specific heat of a cubic meter of oxygen is as follows: At o C. = 303 calories. 200 C. = 314 400 C. = 325 6ooC.=3 3 6 8ooC.= 3 47 Derive a formula of the form Specific heat = a+bt to express the data tabulated above. c. What would be the average specific heat for the temperature interval between zero and 1000 C.? d. How much heat would be required to raise 5.3 cubic meters of oxygen from o C. to 1000 C.? e. If heat were supplied to i cu.m. of oxygen at the rate of 1000 calories per second, plot the heating curve which would result. /. Discuss the reasons why the heating curve of lead does not give as satisfactory results as a cooling curve. g. Is it possible to cool a liquid below its true melting- 62 EXPERIMENTAL GROUP II point? Is it possible to heat a solid above its true melting- point ? h. Show how the specific heat of molten and solid lead can be compared by a cooling curve. Can the latent heat of fusion be computed from the curve? i. What effect would a small quantity of alloying impurity have upon the melting-point of the lead? What effect would a small quantity of insoluble impurity have? EXPERIMENT NO. 9 THERMO-COUPLE CALIBRATION Object. The object of this experiment is to calibrate a thermo-couple by known fusion temperatures of metals. General Explanation. It has been shown in Experiment No. 6 that if the resistance of the circuit is a constant, the amount of current flowing in a thermo-couple, and therefore the indication registered on the dial of the electrical instrument, varies directly as the net thermal electromotive force generated in the couple. This resistance is very nearly constant if one uses the same couple, leads, and meter, and if the small variation in resistance in the couple wires, due to varying temperatures at the hot junction, may be disregarded. The net electromotive force of the couple is equal to the contact electromotive force generated at the hot end minus that generated at the cold end. In symbols i. R=E T -E C where R is the net electromotive force of the couple, E T is the contact electromotive force generated at the hot end, and E c is the contact electromotive force generated at the cold end. It was also stated that the value of the contact electro- motive force was a function of the absolute temperature, that is, II. E T =f(T). HI. E c =f(C). where T or C is the absolute temperature of the hot or cold end, respectively. 63 64 EXPERIMENTAL GROUP II Whence, substituting in I, IV. R=f(T)-f(Q. A general equation which exactly expresses the relations be- tween n+i experimental determinations of equal accuracy is . . . A n T n and . . . AC where Ao, AI, A 2 , AS . . . are numerical coefficients, the same in both expansions, because E T is the same function of T that E c is of C. Substituting in Equation IV and collecting V. R = E T -E C = A 1 (T Now substitute for the absolute temperatures T and C their equivalents 2+ 273 and +273, where / and c are expressed in the ordinary centigrade scale. Expanding after this sub- stitution, and collecting like powers of the variables / and c, VI. ^ = /- ' It will be noticed that the quantities multiplying the terms (/ c)-, (t 2 c 2 ) . . . consist of numerical values only, whose sums can be replaced by other constants such as p, q, r. . . . This is really to be expected, inasmuch as the substitution of c+273 for C, etc., merely shifts the origin of coordinates 273 units to the right, and the degree and form of the equation is unaffected by such an operation, as is well known from analytical geometry. Equation VI then becomes VII. R = t THERMO-COUPLE CALIBRATION 65 In laboratory work the cold end is kept in an ice bath, whence c = o, and VIII. R = pt+qt 2 +rP+st* + . . . (cold end at zero). As a matter of fact, calibration curves of useful metallic combinations are so flat as to approach a straight line that is, the coefficient q is very small, and r and s approach zero. For commercial calibration, therefore, the equation IX R = pt+qt 2 (cold end at zero) will be found to express the relation between the electro- motive force registered and the temperature with much more precision than is attained under the conditions of subsequent use. Evidently, then, all that is needed to calibrate a thermo- couple is to discover the numerical value of the constants p and q in equation IX, when the readings corresponding to 100 C., 200 C., 300 C. . . . up to the limiting temperature can be computed, plotted to scale on coordinate paper, and the points connected by a smooth curve. This curve will show the tem- perature in degrees Centigrade corresponding to any observed reading on the dial; or vice versa, the dial reading for any desired temperature. In practice, p and q may be evaluated by a pair of experi- ments which determine the reading on the dial when the hot end of the couple is plunged into a pure metal or salt at its melting-point. The exact reading on the dial corresponding to this temperature is determined from a study of the cooling curve of the metal as discussed fully in Experiment No. 8. The sub- stances chosen for the calibration should have melting-points near the ends of the expected useful range, with a third selected midway for a check. A list of standard points follows, and is taken from page 456, Burgess's " Measurement of High Tem- peratures." 66 EXPERIMENTAL GROUP II Substance. Boiling-point. Freezing- point. Water. IOO Zero Naphthalene. 218 Tin 27T Q Benzophenone 306 Lead ?27 Zinc A.IQ Sulfur . 444- 7 Antimony 6*1 Sodium chloride 800 Silver 06 1 CoDDer. 1083 Lithium metasilicate I2O2 Diopside I 3QI Nickel I4.CO Palladium I ^ CO Platinum. ... I7CC For the purpose of this experiment, pure lead and common salt (NaCl), will be used to determine the coefficients p and q. The cooling curve of lead already determined in Experiment No. 8 will give one set of data, while a corresponding cooling curve of salt gives the relation, in this hypothetical instance, of 12.8 units = 800 C. Two independent relations between the meter reading and a fixed temperature have been discovered in this manner. These values substituted in Equation IX will give For lead, 5-4 = />(327)+?(327 2 ); For salt, i2.8 = X8oo)+?(8oo 2 ); which gives a pair of simultaneous equations in p and q. Solv- ing, the values ^=+0.0169 and c= -0.00000115 are found, and the equation of the calibration curve for this particular couple is completely determined as ^ = o.oi69/ O.OOOOOH5/ 2 . Values of R can now be calculated for each hundred degrees interval from this equation, and plotted as illustrated in Fig. 8. THERMO-COUPLE CALIBRATION 67 Special Apparatus. The special apparatus needed is as follows: One electrical meter. One oooo graphite crucible. One sound lo-gm. clay crucible. Three hundred grams pure salt. Three hundred grams pure antimony. iooo c 900' CALIBRATION CURVE FOR THERMO-COUPLE NO. 13 ALLOYS 183-343 METER 7 C- C" <^7/lAA/yrL/ x X 800 X 700 / x 600 X *1r 500 x x 400 ,-X X 300 X X E.juatiom-R = 0.0169T- .OOOOOllST^ Points used:- Pb =32?C. = 5.4 Mv. Sb =G31C.=10.2Mv. Cheek :- S = 444.?C. Figures 7. 3 Mv. Reads 7.2 Mv. 200 / X 100 X ^^ x x 04 * 00 o OJ ** 2 Mill! voltmeter Readings > FIG. 8. Calibration Curve of Thermo-Couple. Supplies. The supplies needed are as follows: Ice. Charcoal. Laboratory Equipment. The laboratory equipment needed is as follows: Bucking-board and muller. Procedure, a. With a sharp knife, cut the letters " Sb " in the side of the graphite crucible containing the anti- mony. 68 EXPERIMENTAL GROUP II b. Melt the salt contained in the clay crucible slowly in a pot furnace, but avoid heating the crucible much above the melting-point of the salt. At high temperatures the material readily volatilizes and is so labile as to easily work its way thru checks in the crucible walls. c. Follow procedure c to h inclusive of Experiment No. 8, taking care not to cool the crucible more than 100 below the solidification point of the salt, until the work has been inspected and accepted. Subdivide the duties among the squad and plot the cooling curve as in d below. d. Follow procedure b to i inclusive of Experiment No. 8 with the crucible of antimony, and plot as shown in Fig. 7, p. 59, where the time scale has been contracted. In such a plot it is unnecessary and undesirable to connect the points by a curve, as the points themselves indicate its course excellently well. A pencil dot is even better than the small circle. In this work assign one member of the squad to read the meter, another to call time intervals and record the read- ings, the third to plot the readings as they are read with a pencil on cross-section paper (Fig. 7, p. 59), and the fourth to attend the furnace and thermo-couple. Queries, a. Compute the values of the coefficients p and q in Equation IX for your thermo-couple, using the values obtained by the melting-points of lead and salt. Write the equation of the calibration curve as an explicit function otR. b. Using this equation, compute the reading for the melt- ing-point of antimony. c. Transform the equation of the calibration curve into an explicit function of /. Check the value obtained in query b by means of this equation. It should agree with the reading obtained in procedure d within the least count of the meter dial. d. Compute values of R for each hundred degrees, and draw a calibration curve with India ink, following closely the style and form of Fig. 8. THERMO-COUPLE CALIBRATION 69 e. What would be the reading on the meter if the hot end were at 900 C. and the cold end in boiling water? /. If the temperature of the cold end were disregarded, what would be the apparent temperature indicated in e? g. What is the cause of the " supercooling " effect noticed in the. antimony curve? How can it be prevented? EXPERIMENT NO. 10 LEAD-ANTIMONY ALLOYS Object. The object of this experiment is to construct the equilibrium diagram of the lead-antimony system. General Explanation. (Notes on the commercial importance of these alloys may be found in Gulliver's " Metallic Alloys," pp. 46 and 47.) Several different lead-antimony alloys can be melted and their cooling curves obtained as in Experiment No. 8. These alloys may vary in composition from pure lead containing no antimony, thru alloys of lead containing increasing amounts of antimony up to the limit represented by pure antimony itself. As we have already noted (p. 57), a sudden discontinuity or " arrest " in these cooling curves is evidence that a considerable proportion of the constantly radiating energy is supplied by latent (not sensible) heat. This heat may be produced by any one of a variety of physical or chemical changes in the body under observation; and a complete study of an alloy system involves, among other things, the location and explanation of all changes in the state of all possible alloys in that series. The relations between temperature, composition, and constitution obtained in such a study can be represented graphically upon coordinate paper in what is called an " equilibrium diagram." The equilibrium conditions of the lead-antimony series may be simplified by a consideration of the cooling curves of a series of solutions of ordinary table salt, NaCl, in water. We will call this series of alloys the water-salt system. These two systems are typical of an important and common family of alloys alike in that the components lead and antimony or water and salt, respectively form no compounds with each 70 LEAD-ANTIMONY ALLOYS 71 other, and are perfectly soluble in the liquid state but totally insoluble in the solid. In other words, lead and antimony (or water and salt) alloy perfectly when liquid, but upon solid- ification, separate into a cemented mixture of minute crystals of pure lead and of pure antimony (or pure water and pure salt). The cooling curve of pure water will give a horizontal arrest at o C. corresponding to the freezing of water into ice. A cooling curve of a solution consisting of 10 per cent salt and 90 per cent water will show the first break not at o C., but at 4 C. This arrest is not a horizontal portion of a curve as is the case in a pure substance (Experiment No. 8), but merely a change in the direction. Inspection of the solution shows at this same instant the appearance of the first minute crystals of solid, and therefore this change in direction of the cooling curve is interpreted to mean that at that instant and thereafter only a part of the radiating heat is furnished by the sensible heat of the mass, and the balance is furnished at the expense of the latent heat of fusion of the solidifying substance. In fact, the mass does not solidify at a constant temperature; but at 4 C. the first few crystals separate out of the liquid, and prove on analysis to consist of pure water in the form of ice. The presence of the salt in the solution has evidently lowered the freezing-point of the water. This phenomenon might have been predicted from the generalization stated in Experiment No. 5: " Given two pure substances, the melting- point of either is lowered by the addition of certain quantities of the other. " Within limits, the amount of this depression of the freezing- point varies directly with the amount of salt in the solution. As the first crystals of pure water form, there remains behind a liquid (mother liquor) which possesses a correspondingly less amount of liquid water, and therefore contains a higher percentage of salt. Consequently the freezing-point of this richer salt solution is lower than that of the original, and other crystals of ice cannot separate out until further cooling lowers the sensible temperature of the entire mass to the freezing- 72 EXPERIMENTAL GROUP II point of this more concentrated salt solution. By the selective solidification of the water (and consequent removal of these crystals of water from the liquid) the residual mother liquor becomes richer and richer in salt, and the temperature at which crystallization takes place falls further and further. There must be a limit to this short of producing pure salt, for salt solidifies at 800 C. This limit may be considered as being the true " saturated " solution, and is finally reached at the temperature of 22 C., when the whole remaining liquid solidifies at a constant temperature, and this fact is shown on the cooling curve by a horizontal portion. Should the ice crystals forming from such a solution be removed as rapidly as they freeze, the remaining mother liquor which solidifies as a pure substance would be found on analysis to consist of 23.5 per cent salt and 76.5 per cent water. Other salt concentrations will produce similar freezing curves. Assume the solution to be 15 per cent salt and 85 per cent water; the process is strictly analogous, except that the first ice crystals freeze out at a still lower temperature (about 9 C.); which, indeed, is to be expected. The solution re- maining liquid to the last, left behind by the solidification of more and more water, congeals as before at the constant tempera- ture of 22 C., and possesses the same composition, i.e., 23.5 per cent salt and 76.5 per cent water. If a solution originally possess this limiting composition, it has only one freezing-point, which is at 22 C. This cooling curve will look like the cooling curve of a x pure substance, and show a horizontal arrest at that temperature. Solutions which contain more than 23.5 per cent of salt have cooling curves exactly similar to solutions poor in salt; that is, they possess an arrest (or, more correctly, a change in direction) at v c( v k v _/ 1 -^ v -V v V"- '. ~~ I 1 % % 8 K S Data Temperature Time Interval 37 11:36.35 Knc 36.5 M.V. 11:37:25 t>ec ' 36 11:38:05 35.5 11:38:55 TT 35 11:39:40 34.5 11:40:30 iJjJj 33.5 11:42:10 ?*' 33 11:42:55 f, 32.5 11:43:50 ?T 32 11:44:45 31.5 11:45-^0 ,^ 31 11:47:30 ,XX 30.5 11:49:10 *5 30 11:50:25 '" 20.5 11:51^0 29 11:53:11 ?} 23.5 11:57:15 28 11 :58 :05 ?~ 27.5 11:59.00 ? 27 11:59:45 26.5 12:01K)0 '.? 26 12:01:45 . AM- 32.6 M V. =7 75C. \ ^ x< / * X - _ - M.V. "' = 695( . i ** i :=- .f . Fig. 22 INVERSE RATE COOUNG CURVE OF 0.38-C. STEEL I s quad t 50 100 150 200 250 Time Intervals >- FIG. 22. Inverse Rate Cooling Curve of 0.38 C, Steel. temperature corresponding to the beginning of undoubted dis- continuities in the rate of heat absorption. Be careful not to exceed a temperature of 1000 C., on account of the danger of 130 EXPERIMENTAL GROUP IV burning out the heating element of the furnace. Do not change the external resistance during the interval 550 C. to 950 C. e. Shut off the current, and read and plot as in Fig. 22 a similar inverse rate cooling curve thru the same range. Hold the temperature at 550 C. while the curves are discussed with the instructor. /. Repeat procedure d and e at a considerably slower rate, regulating the speed so as to occupy the remainder of the labora- tory period. Place additional kieselguhr over the lid to decrease the cooling speed. g. Present all results to the instructor, and stick properly colored pins into the coordinate diagram on the laboratory wall, locating both the A c and the A T transformations determined by your curves. Queries, a. Construct a neat equilibrium diagram showing the decomposition of austenite, after the style of Fig, 9, page 74. Arrange the scale so that the carbon content will range up to 2 per cent, and the temperatures between 500 C. and 1000 C. Locate the A c points determined by the various squads with small red circles, and connect by a red line. Locate the A r points and lines similarly in blue. Also draw in a heavy black line the equilibrium diagram (A e lines) of Fig. 23, which was adopted by Howe,* after weighing the best information now available on the subject. This will produce a sheet of cross- section paper showing three superimposed diagrams in color two as determined in our laboratory, the third represent- ing the standard adopted by Howe. Label all lines, coor- dinates and areas. b. Draw the heating and cooling curves derived with India ink. Discuss the reasons for the non-concurrence of the corre- sponding A and A T points. Has this lag any relation to the " surfusion " shown in the solidification of antimony? What effect should time have upon the magnitude of this hysteresis? * See Howe, " Metallography of Steel and Cast Iron," page 130. The points located on the diagram are from cooling curves made by Carpenter and Keeling, I, 1904, " Journal of the Iron and Steel Institute," 224. TRANSFORMATION POINTS 131 Fig. 23 HOWE'S DIAGRAM / V 1500 \ v^ Points from Carpenter and Keeling / V ; o ( X x / ^400 \ > \ y i X \ / \ x x Tamperature i Degrees Centigrade > \ \ N x 1 \ \ X \ 1 \ X - \ x / ^, j \j / D 1 / ( 7 / / \ / " \ / \ o o V s \ o ; *LO 2.0 3.0 4.0 5.0 Per cent Carbon &- FIG. 23. Howe's Diagram. 132 EXPERIMENTAL GROUP IV c. In the light of the above general explanation, and also of your study of the lead-antimony diagram, explain the actions proceeding in the metal at all times during the heating and cooling. d. Define allotropy. Give the allotropic modifications of some other substances than iron and sulfur. Give an explana- tion of allotropy based upon the molecular theory, as suggested by the allotropy of oxygen. EXPERIMENT NO. 17 CRYSTALLIZATION OF STEEL Object. The object of this experiment is to show the growth and restoration of the crystalline grain in a piece of steel thru various heat treatments. General Explanation.* Before the analytical chemist came into the steel plant, not so many years ago, eye examination of a bright fracture was the only method at the disposal of iron makers for classifying their product. Indeed, at the present day, the skilled melters operating a modern open-hearth steel furnace, cast, break, and examine the fracture of small pigs of metal at definite intervals of time, in this way gaining accurate information as to the elimination of the impurities in the bath, and, toward the finish, predicting the carbon content of the steel to within a few points. Pig iron is commonly sold by the appear- ance of the fracture, which, by the way, is the universal reliance of the " practical " foundryman. The fracture of " spies " withdrawn from cementation boxes indicates the progress of the carburization (v. " Mills, Materials of Construction," page 364), and each cemented bar is broken, examined, and classified as to carbon content before being melted into crucible steel. Test lugs from case-carburizing and malleableizing boxes (op. cit. pp. 421, 336) are regularly broken to indicate quickly the quality of the product. In this way the skilled operator is enabled to determine the homogeneity and approximate chemical composition of an iron-carbon alloy, if the previous heat treatment is known. On the other hand, a skilled inspector can approximately determine * A detailed discussion of the points covered by this experiment may be had in Stoughton, " Metallurgy of Iron and Steel," pp. 357-368. 133 134 EXPERIMENTAL GROUP IV its past heat treatment, and probable strength and brittleness by the fracture, if the carbon content of a steel is known. Pro- gressive manufacturers are controlling, supplementing, and checking their fracture indications by chemical analyses, while wide-awake engineers are demanding more thorogoing frac- ture tests, supplemented by metallographical analysis, to con- trol their acceptances. At the present time, railway engi- neers are demanding that top rails from each ingot be tested under the drop-hammer (see Fig. 24), while rail manufacturers are insisting that the present practice of testing random rails FIG. 24. Undesirable Rails. Reproduced from " The Engineering News-Record," by permission of the McGraw-Hill Publishing Co. from each heat is sufficient. Bridge and mechanical engineers are coming to the conclusion than many disastrous failures attributed to " crystallization under alternating stresses " are in reality due to the brittleness induced by the coarse crys- tallization resulting from improper heat treatment during fabrication, and are relying upon their inspectors to reject such pieces at the steel mills. Microscopic examination has confirmed the supposition that a piece of metal is essentially an aggregate of small crystalline grains cemented together with a thin film of non-crystalline (or amorphous) material. (See Experiment No. n). It seems that the amorphous cement is stronger and tougher than the crystalline aggregate; therefore, when a piece of such material CRYSTALLIZATION OF STEEL 135 fails, the fracture passes thru the crystals themselves, along well-defined cleavage or parting planes, rather than around the rougher superficies of the individual crystals. Consequently, the fresh fracture will show bright flashes of light reflected from the small, flat, parting planes. The larger the crystalline grain, the larger the parting planes, and the coarser and more fiery the fracture. The finer the crystals, the greater the percentage of amorphous material, the stronger and tougher the steel, and the more silky the fracture. If cast steel (Fig. 25) is held at a high temperature, the crys- talline grain does not seem to grow coarser. Once the crystals of this casting have been broken up and intermixed by hammer- ing, rolling, or other mechanical kneading (Fig. 27), the grain size is strictly dependent upon the heat treatment subsequently imparted to the bar. (Fig. 28). It seems that even a moderate degree of heat (500 C.), if continuously applied, is sufficient to cause growth in the grain size, ultimately inducing weakness and failure thru a phenomenon called " Stead's brittleness." Furnace buckstays, crane chains, and other members repeatedly heated to such temperatures should, therefore, be annealed at intervals to restore their original properties. A very short time at an extreme temperature will cause the same grain growth, and the finest steel can be absolutely ruined by long exposure at high temperature if followed by no subse- quent working or heat treatment. In welding and forging practice, the blacksmith guards against this danger by. contin- ually hammering the metal until it has cooled to a red heat, less than 900 C., in order constantly to break up the growing crys- tals. Below Arz, mechanical work is abandoned, because the rate of growth is then quite slow, and the metal has lost the plasticity which is required for hot working. Fortunately, unless the steel has been absolutely ruined by overheating to the point of incipient fusion, and therefore " burned " (Fig. 29), all previous crystalline structure seems to be obliterated by a reheating thru Ac\, at which transforma- tion range a new and independent accumulation of crystalline FIG. 25. Cast Ingot. FIG. 26. Forged and Reheated Nearly to Ac z . FIG. 27. Reheated Considerably above Ac$. FIG. 28. Reheated Much above Ac 3 . FIG. 29. Reheated Past Burning Point. (Photomicrographs by E. P. Stenger, of 0.35 Carbon Steel. All at 75 diam.) 136 CRYSTALLIZATION OF STEEL 137 nucleii seem to come into existence. Growth of these new crystals begins at this range, and continues with increasing rate at higher temperatures. Annealing for restoring the grain, and for inducing the maximum toughness possible for a given carbon content, should therefore be carried no higher than to make sure that the whole mass of the steel has definitely passed Aci. However, a certain qualification is necessary for low-carbon steels, as follows: the primary ferrite crystals are not entirely absorbed into the solid solution until Ac% has been passed, nor are all traces of the old structure entirely eliminated until enough time has been given above Acs for a thoro diffusion of the iron molecules into the surrounding austenite. (Fig. 26.) During this time, however, the new crystalline orientation is growing apace; one is, therefore, between the horns of a dilemma, but it will generally be found that for hypo-eutectoid steels a thoro mechanical working (producing a very small uniform grain), followed by annealing slightly above Acs, will give the best physical properties to the piece. Special Apparatus. The special apparatus required is as follows : One electrical meter. Supplies. The supplies needed are as follows: One piece |-in. wrought-iron gas pipe, 12 in. long with closed end. Four bars from Experiment No. 16, two of which will be further useful. Box of gummed labels. Laboratory Equipment. The laboratory equipment needed is as follows: Ice. Crucible furnace, at 1250 C. Set of alphabet punches. Anvil. Three-pound blacksmith's hammer. 138 EXPERIMENTAL GROUP IV Scleroscope. Emery wheel. Impact machine. Optical pyrometer. Five-gallon water pail for quenching bath. Vises. Procedure. NOTE. High-carbon steels are best to show crys- talline growth at high temperatures. Medium carbon steels can also be used. Lower carbon steels should be nicked deeper, but even so, some will not break without large bending, which spoils the appearance of the fracture. These steels should, therefore, be sawed off, and studied microscopically by the better squads, and a series of specimens mounted, at equal magnifica- tions, for inspection by the whole class. a. Saw about three-eighths of the way thru two of the rods of Experiment No. 15 at f-in. intervals. Mark the sec- tions consecutively with alphabetical punches. Put one of these nicked bars, together with another of the plain bars from Experi- ment No. 1 6, in the hot crucible furnace. Delegate one mem- ber of the squad to hold this furnace as close to 1250 C. as pos- sible, maintaining a reducing flame at all times. Allow the notched piece to remain in the furnace thirty minutes and then remove it and allow it to cool in the air. The plain bar is to remain in the hot crucible two hours, cooled in air, and then used in procedure/. . b. Break one section from the ends of both notched bars by placing them in the round hole in the anvil and striking a sharp blow with a 3-lb. hammer. Vises must not be used for breaking specimens. Examine the fractures, and save the small pieces for further comparison. Paste a small gummed label around each fragment, giving its carbon content and heat treatment. c. Place the two notched pieces alongside each other in the oven furnace and heat (anneal) successively to the following temperatures: Aci $o -, ^1 + 25; Ac 3 -2$; ^3 + 25; and then by 100 intervals to the limit of the furnace. Measure the CRYSTALLIZATION OF STEEL 139 temperature as in procedure d. After each temperature has been reached, remove the bars from the furnace, cool in air, break off a section, examine, and label as before noted. d. Measure the temperatures as follows: Place your cali- brated thermo-couple, properly covered with asbestos string (Experiment No. 7) in a i2-in. protection tube of f-in. wrought- iron pipe. Insert this tube thru a suitable hole in the side or back of the furnace in such a manner that the hot end will be directly above the notched bars. A reducing flame should be maintained at all times. Adjust the couple and bars correctly, close the door, and raise the temperature very gradually up to the first annealing temperature noted, taking at least fifteen min- utes to attain that degree, and making sure that the bars and furnace do not pass that heat. The door of the furnace may be left open to cool the muffle slightly, so that another fifteen min- utes will be consumed in heating to the next temperature, and so on. Remove the thermo-couple, and use an optical pyrometer for temperatures above 1000 C. e. Remove only one bar from the muffle at the highest tem- perature, and try to refine the grain of the other by allowing it to cool in the furnace to just above Ac%, and then quickly quench by plunging it endwise into a pail of tap water. Dry the bar and inspect the grain size by breaking, as in procedure b. f. Take one of the rods of Experiment No. 16 and anneal it carefully up to the temperature which, according to your experi- mental results, produces the finest grain. Allow it to cool in air from that temperature. Both this bar, and the bar overheated for two hours in procedure a should be tested for relative tough- ness in the impact machine, as follows: Mount the bars in the vise so that about one inch will protrude, and strike this pro- jecting end successive blows with the drop-hammer, starting at i-in. fall, then 2-in., then 3-in. and so on until failure ensues. g. Grind one end of each piece flat, and test the hardness with the scleroscope. Queries, a. Take a piece of heavy cardboard, 8| in.Xn in., letter it with the squad number, personnel and composition of 140 EXPERIMENTAL GROUP IV the steel treated. Mount all the steel fragments in an orderly manner by thrusting them thru small holes cut in the paste- board. Under each specimen place its heat treatment, its hard- ness, and its toughness, if determined on the impact machine. Place this sheet on the bulletin board for inspection of the other squads. b. Make up an equilibrium diagram for steel on coordinate paper in light black lines, following Fig. 23, page 131. Plot upon this diagram the hardness numeral of all the fragments tested by the laboratory squads as posted on the bulletin board. Draw contour lines of equal hardness in red. c. By examination of the work of the various squads, place a heavy black line on this diagram showing the annealing temper- ature which produces the finest grain. d. What effect does time have upon the grain size ; if the tem- perature is above Acs? below Ac\? e. Distinguish between overheated and " burnt " steel. Can the latter be again restored to its normal condition by a heat treatment? /. Judging from the experimental data, will there be any change in the grain size during tempering? g. Is there a relation" shown between grain size and the hard- ness and ductility of steel? Cite figures. h. How would one measure the size of the crystalline grain? i. Give experimental support to the statement on page 125 that molecular rearrangement is not only possible in the solid state but molecular migration actually covers considerable dis- tances. EXPERIMENT NO 18 HARDENING OF STEEL Object. The object of this experiment is to study the hardness of steel as affected by quenching temperature and carbon content. General Explanation. It has been known since prehistoric times that certain irons, if quickly cooled from a bright red heat became very hard indeed, and could then be formed into most useful and durable cutting tools. This discovery marked the last milestone of the evolution of man, and foreshadowed modern civilization. The curious phenomena of hardening are yet controlled by rule of thumb methods which are closely guarded trade secrets. During the last century, the advent of chemical analysis indicated a close correspondence between the hardening power of irons and their carbon content; thus, wrought iron containing practically no carbon could not be usefully hardened, while steels containing an increasing amount of carbon were capable of taking a harder and harder edge. The exact cause of hardening is still an unsolved mystery, despite the enormous amount of scientific research which has been expended upon the problem in the last twenty years, even applying, as it has, the utmost resources of such powerful investigative weapons as physical chemistry and metallography. Several hypotheses are under discussion as among the possibilities, each of which has its ardent supporters; and for the most recent exposition of our knowledge on this matter the student is referred to Howe, " The Metallography of Steel and Cast Iron," pages 173 to 196. From the point of view of practice, the hardening operation consists of two parts; first, heating the metal uniformly to the 141 142 EXPERIMENTAL GROUP IV proper temperature, and second, cooling it quickly and uniformly. The first operation presupposes a well-designed furnace, uniform in temperature from top to bottom and end to end, operating steadily with a neutral flame (" hazy heat ") Modern produc- tion of quantity work demands a heat control more delicate than the unaided eye, and pyrometers are installed in all furnaces where quality and uniformity are prerequisite. The metal must be in the furnace long enough to attain the furnace temperature, which time, of course, varies with the mass of the piece. It has been found that, owing to the large heat conductivity of steel, the surface of a metallic object is con- stantly somewhat hotter than its center, but colder than the furnace atmosphere. Both differentials decrease as the steel nears the furnace temperature; therefore, when a thermo-couple in contact with the surface of the metal registers the same degree as the furnace itself, the piece is heated uniformly and is ready for quenching. At such a time the piece can be seen only with difficulty, as it radiates heat at the same rate as the furnace walls themselves. As a matter of fact, it is quite a difficult thing to construct a furnace which will operate at a uniform temperature and a neutral flame. It also requires good judgment so to place the metallic pieces that a good flow of heat can circulate all about them, warming all parts at a uniform rate. In many places baths of molten salt are installed which have many obvious advantages over a gas- or coal-fired furnace. Salt baths are better than molten lead baths, because they do not alloy with the metal nor do they oxidize, the steel pieces readily sink in the hot liquid and do not have to be held below the surface. On withdrawal for quenching, the objects are protected from surface oxidation or decarbonization by a thin film of adherent salt, which can easily be washed off. Warping may occur in the heating furnace due to a non- uniform heating, or it, with its partners cracking and internal strain, may appear after the quenching operation, but for the self- same reason variable heat-transfer rates. The ordinary dif- HARDENING OF STEEL 143 ferences in volume due to expansion with increasing tempera- ture are accentuated by the fact that there is a large contraction in passing thru the A 3 range and a corresponding expansion at Ar s transformation. A large forging, for instance, rapidly quenched from the austenitic condition, will act somewhat as follows: The surface will be cooled very rapidly thru the Ar ranges, so rapidly that the normal expansion occurring here will be suppressed. The drop in temperature will also cause a further large contraction, and the net result is a cold muff shrunk upon a hotter core. This core will cool much more slowly and the Ar transformations, with its accompanying expansion, will proceed (at least in part), further stressing the outer regions. It is small wonder that quenched forgings warp, split, and even explode violently, under the excessive stresses thus produced. Large pieces should, therefore, be counterbored, mildly quenched, and annealed immediately. Boring a hole in the center of forg- ings removes metal (which is usually below grade because of piping and segregation) from that portion of the piece where even good metal would be least effective, and at the same time provides for a more uniform cooling rate, edge to center. It is, therefore, seldom that even a piece with a very large counter- bore would not be a better, stronger, and safer piece than originally. Consequently, such practice should be insisted upon in all large pieces unless absolutely impossible of attain- ment. From the above discussion, it is evident that for hand work, symmetrical sections should be quenched vertically in the direction of their greatest length. Hollow sections should, in addition, have a stream of the quenching fluid forced up thru the interior opening. For quantity work, certain mechanical features are useful, as noted in Bullens' " Steel and its Heat Treatment," ist Edition, pages 86 to 95. Special Apparatus. The special apparatus needed is as fol- lows: One electrical meter. Five-gallon pail for quenching bath. 144 EXPERIMENTAL GROUP IV Supplies. The supplies needed are as follows: Ten steel rods from Experiment No. 16. Two steel rods from Experiment No. 17. Handful of waste. Laboratory Equipment. The laboratory equipment needed is as follows: Vises. Numerical punches. Ice. Tank of quenching oil. Impact testing machine. Scleroscope. Procedure, a. Take twelve pieces of steel from Experiments Nos. 16 and 17, and saw one notch f in. from the end of each piece. These notches must be gaged and cut exactly the same depth in all bars. Punch-mark the short ends corresponding to the quenching temperatures of procedure e. b. Place the bars crosswise in the oven furnace, separated somewhat from each other, with that one bearing the lowest number at the front. Insert your calibrated thermo-couple in the wrought-iron protection tube; and thrust this pipe thru the pyrometer hole in the furnace in such a manner that the hot end will be directly above and preferably resting upon the first of the bars. c. Heat the furnace slowly with a reducing flame, taking -about forty-five minutes to attain a temperature of 600 C. When this degree is reached, one squad member should open the door only long enough for another to grasp the end of the first bar with tongs, plunging it immediately end on into a 5-gallon pail of quenching oil, placed close up to the furnace, so that no steps are necessary. Speed in transfer is essential. Move the bar back and forth sideways in the bath until it is cold, keeping it constantly submerged in a vertical position. Remove, wipe off the oil, and test for toughness and hardness according to pro- HARDENING OF STEEL 145 cedures/ and g of Experiment No. 17. Paste a small gummed label around each fragment, giving its carbon content, heat treatment, hardness, and toughness. d. During the testing, the furnace tender should adjust the next bar and the thermo-couple to juxtaposition, and increase the temperature slowly so that at least ten minutes shall inter- vene before reaching 650 C., when procedure c should be repeated. Great care is necessary to increase the temperature of the furnace slowly, and to get the bar out of the furnace at the exact temperature neither more nor less. e. Continue in this manner, quenching bars from the follow- ing temperatures: 600 C. 725 C. 800 C 900 C. 650 750 825 1000 700 775 850 noo /. Shut off the gas from the furnace after the last bar has been removed, and cool it rapidly by leaving the air blast on. When it has dropped to a black heat, range all the long bar-ends properly in the furnace, and anneal carefully at the temperature which gives the minimum grain size. Cool over night in the furnace. g. Plot on a sheet of coordinate paper one curve showing the relation of hardness to quenching temperature, and another showing a like relation for toughness to quenching temperature. Use degrees Centigrade rather than dial readings for abscissae and locate each quenching by a small circle. Letter on the sheet the squad number, personnel and the carbon content of the steel. Submit this sheet of curves to a laboratory officer for inspection and O.K., and then pin the sheet on the proper bulletin board for reference by the rest of the class. Queries, a. Each student should make in India ink the curves described in procedure g above. Locate in an appropriate manner the position of the Ac transformation ranges. Discuss the relation of hardness to transformation ranges as evidenced by this experiment. 146 EXPERIMENTAL GROUP IV b. Compare the crystalline grain shown by the fracture of the quenched pieces from this experiment to those from Experi- ment No. 17. What general difference, if any, is exhibited by quenched pieces from those annealed at the same temperature. c. Make up an equilibrium diagram for steel on coordinate paper in light black lines, following Fig. 23, page 131. Plot on this diagram the hardness numeral of all the fragments tested by the laboratory squads as posted on the bulletin board. Draw contour lines of equal hardness in red. d. What is the constitution of a .05 carbon steel (five-point steel) at 740 C., 850 C., and 1000 C.? Are these states pre- served in the cold after quenching in oil? What conclusions can be drawn from the curves of query c as to the hardness of wrought iron if quenched from these temperatures? e. Why is the lowest heat giving the hardening effect the best heat to use? EXPERIMENT NO. 19 QUENCHING MEDIA Object. The object of this experiment is to study the hardening power of various liquids. General Explanation. When a piece of hot iron is plunged into a cold liquid, the latter dances against the surface of the metal in much the same manner as drops of water dance on a hot stove. The quenching fluid comes into contact with the heated surface perhaps momentarily, but it immediately vaporizes, and the liquid for the most part is held away by a thin film of gas. The actual transfer of heat (see page 105, Experiment No. 14) from the hot metal to the cold liquid takes place in the first place by convection currents, where gaseous bubbles are driven away into the colder surroundings, there to escape or be con- densed; secondly, by radiation thru this gaseous envelope, and thirdly, by conduction across the following system metal: gas interface, gas film, gas: liquid interface. The relative impor- tance of the three methods are about in the order as stated the first (convection) probably removing a large part of the heat, the third (conduction) removing comparatively little, owing to the great resistivity of the system due to the large coefficients of internal transfer at the two interfaces. The heat loss by convection is obviously accelerated by moving the quenched metal about in the bath. We may say, therefore, that the quenching power of a fluid depends, to a large extent, upon its total heat to the boiling point, its latent heat of vaporization, and its coefficient of emissivity (page 112). A bath which absorbs much heat per unit volume to bring the contact firm to the boiling point, then a further large amount to vaporize this hot liquid, and is thin or mobile enough to allow free passage of gas bubbles away from 147 148 EXPERIMENTAL GROUP IV the hot surface into the colder surroundings, will cool heated articles quickly. Thousands of different fluids have been tried as quenching baths. Those most commonly used are water, various salt solutions, different kinds of mineral or organic oils, and finally, molten metal or salt baths. A tank of ordinary tap water is one of the quickest quenching baths known, being exceeded only by salt solutions and water jets; in fact, it produces an action so drastic that it is unsafe to use on work with more than twenty points of carbon except in the hands of experts. Oil baths are much milder, and their cooling speed is less dependent upon the temperature of the bath. In fact, some sticky oils quench more rapidly when hot than when cold, owing to the greater viscosity at low temperatures. Molten metals or salts cool hot steel quite slowly, but are invaluable in the heat treatment of modern high-speed steel. Many secret compounds on the market are said, especially by their salesmen, to be capable of curing any hardening room difficulty. As is the case of patent cements (Experiment No. 23), most, if not all, such panaceas are worth- less; educated people shun " cure-alls." Good quenching baths should have arrangements to main- tain a uniform temperature. This may be effected by coiling cold-water pipes around the sides of the tanks; or, better, by circulating the quenching medium itself. Stirring the oil by means of compressed air should be avoided, on account of the oxidation of the oil which usually proceeds, and the danger of a stream of air bubbles bathing the side of a hot piece during quenching, with the consequent formation of a soft spot. Special Apparatus . The special apparatus needed is as follows : Five 5-gallon pails. One large gas burner. One large tripod. Two pieces of gas tubing, 10 feet long. One 300 C. mercury thermometer, metal cased. One electrical meter. QUENCHING MEDIA 149 Supplies. The supplies needed are as follows: Four pounds of salt (NaCl). Seven bars from Experiment No. 18. Laboratory Equipment. The laboratory equipment needed is as follows: Vises. Alphabet punches. Ice. Tank of annealing oil. Lead pot at 400 C. with pyrometer equipment. Procedure, a. Call an instructor's attention to the curve constructed in Experiment No. 18, procedure g, and decide with him the correct quenching temperature for that particular steel. Notch seven pieces of steel from Experiment No. 18, three- quarters of an inch from the end, and to exactly the same depth in each bar. Punch-mark the short ends corresponding to the quenching media used. b. Arrange the thermo-couple in the furnace as directed in procedure d of Experiment No. 17, and heat the oven furnace to the proper quenching temperature. Hold the temperature at this degree for twenty minutes with a reducing flame, before putting the bars into the furnace. One member of the squad should give his entire attention to the furnace control. c. Place the seven bars endwise in the furnace, notches to the rear, so that they may be equally heated in all parts, and allow to remain thirty minutes, or longer if necessary to attain the temperature of the furnace. The flame need not be adjusted during the heating if the furnace was at equilibrium before the bars were introduced. d. During the heating, arrange nearby quenching baths as follows, each in a 5-gallon pail: Ten per cent salt solution. Ice water. 150 EXPERIMENTAL GROUP IV Tap water. Boiling water. Annealing oil at 200 C. Also see that the lead pot is ready for use, and at a temperature of 400 C. e. Quench one bar in each medium after procedure c of Experiment No. 18. Close the furnace between times long enough for the pyrometer to recover the correct temperature. Test each piece for hardness and toughness according to pro- cedures / and g of Experiment No. 17. In preparing for the hardness test, grind the pieces very slowly, keeping the bars cold. Paste a small gummed label around each fragment, giving its carbon content, heat treatment, hardness, and toughness. Preserve all the pieces for reference. /. Heat the seventh bar in the furnace to the maximum temperature attainable, and quench in iced brine. Test as in procedure e, above. g. Shut off the gas after the last bar has been removed, and cool the furnace rapidly by leaving the air blast on. When it has dropped to a black heat, range all the long bar-ends properly in the furnace, and anneal carefully at the temperature which gives the minimum grain size. Cool overnight in the furnace. Queries, a. Construct a neat tabulation of the results, including the hardness and toughness of oil-quenched steel from Experiment No. 17. b. Why should a quenching in brine give a greater hardness than a quenching in water of the same temperature? c. Define " toughness." Is there any relation between hardness and toughness in carbon steels? d. Give two reasons why the hardness resulting from water quenching should vary inversely as the temperature of the quenching bath. e. Why should the hardness of a high-carbon bar quenched from 1200 C. in iced brine be less than that of the same bar quenched from 850 C. in cold water? QUENCHING MEDIA 151 /. Outline a system for circulating and cooling the quench- ing oil for a battery of hardening ovens. g. What is modern high-speed steel? Outline the heat treatment of such tools recommended by Frederick W. Taylor. h. Why does procedure e, above, specify "grind cold?" EXPERIMENT NO. 20 TEMPERING AND TOUGHENING Object. The object of this experiment is to study the effect of reheating a hardened steel. General Explanation. When carbon steels are heated above the transformation range into the austenitic area, the various aggregations of ferrite and cementite crystals normally present in slowly cooled steels are converted into a uniform solid solution, austenite. The exact reverse of this happens on slow cooling, a? was indicated in Experiment No. 16. All previous structure is consequently obliterated by a sufficient heating, and upon sub- sequent rapid cooling by quenching, the usual conversion of austenite (stable only at high temperatures), back into the aggregate of ferrite and cementite is prevented, at least in part, because of lack of time afforded at the transformation ranges. Time is an essential to this decomposition, for the gamma iron existing in the austenite must change thru the beta into the alpha modification, and the ferrite and cementite must further separate and coagulate into the state recognized as pearlite. Molecular rearrangement and migration in the solid is possible only if a sufficiently elevated temperature allows the molecules the requisite freedom of movement. Should cooling be so rapid as to lock the molecules tightly together in the cold steel before any change proceeds, pure austenite may be pre- served and examined in the cold (Fig. 30). In the ordinary practice of hardening steels as described in Experiments Nos. 1 8 and 19, the quenching is not so drastic, and the transforma- tion of austenite back to ferrite and cementite is more or less completely effected, giving rise to certain transitory forms which are known as " martensite," " troostite," " sorbite," and, finally, 152 TEMPERING AND TOUGHENING 153 the various kinds of pearlite. (See Mills, " Materials of Con- struction," pages 430 to 438.) The phenomena of hardening plain carbon steels by rapid cooling and the softening which occurs by a subsequent moderate reheating, undoubtedly are intimately associated with this delayed transformation. A complete explanation of hardening must account for other facts, such as these: The hardness of certain alloy steels high in carbon, manganese, or nickel, is con- siderable even after slow cooling. Some manganese steels are FIG. 30. Patch of Austenite from Eutectoid Steel. 22oX. Quenched from 800 C. in Iced Brine. hard if cooled slowly, but softer if quenched. Another alloy steel may be hardened in three ways : Either by cooling in liquid air, by a moderate tempering, or by attempting to machine it. Pure iron and low-carbon steels are hardened materially by cold working. A considerable analogy exists between hardening and cold working (which latter causes internal stresses and strains), inasmuch as both increase hardness, brittleness, and strength. Some scientists have, therefore, attempted to explain quenching phenomena by attributing the resulting hardness to the internal stresses caused by the shrinkage of the shell and the dilation of 154 EXPERIMENTAL GROUP IV gamma into beta iron (Andre Le Chatelier). McCance thinks that supercooled gamma iron is held in its original octahedral crystalline orientation under great intermolecular stress; while Humphrey believes that the transformation of iron has actually proceeded far enough to lose its octahedral symmetry without yet attaining the cubic orientation of alpha iron. The resulting iron is, therefore, in an amorphous condition, which we know to be harder and stronger than a crystalline state. These various hypotheses are incapable of explaining the hardness of slowly cooled and annealed alloy steels, altho internal strain, doubtless, produces considerable hardness under certain conditions. Likewise, hypotheses based on the presence of carbon or some hard carbide are insufficient to explain the great hardness of low-carbon alloy steels or the change in hard- ness on tempering high-carbon steels. As a matter of fact, a massive carbide FesC has been prepared and found to possess only moderate hardness. Osmond's allotropic theory thus seems most nearly to cover the facts. It rests on the recognition of three unique states of iron, viz.: gamma iron, dense; beta iron, hard; and alpha iron, magnetic. When the steel cools from the gamma condition, the allotropic changes may be regarded as taking place in two steps; the first characterized by a dilation and hardening from the gamma state into beta iron; and the second by softening and magnetization from beta into alpha iron. These steps overlap to various degrees in quenching at different speeds, and cause independent variations in magnetism against hardness. The latter, in fact, may be anything from the Brinell hardness 125 of gamma iron, up to the 800 of beta, or down on the other side of the maximum to the 75 of alpha iron. According to this theory, a soft steel may, therefore, be austenitic, containing gamma iron, or pearlitic, with alpha iron. Alloy steels are often austenitic, with their A e ranges depressed to nearly room temperatures. Such a steel on moderately slow cooling ordinarily would be in a slightly supercooled austenitic condition, and could be hardened either by further cooling in a TEMPERING AND TOUGHENING 155 refrigerant, causing the beta iron to form (martensite), or by a moderate reheating to a temperature below the transformation range where the time and molecular mobility would be sufficient to effect the same conversion, or, lastly, by the molecular activity induced by the overstraining of machining, forging, or other cold working operations. Osmond's theory is also competent to explain annealing and toughening practice. A quickly quenched carbon steel is mostly martensitic (Fig. 31), which metaral is a solid solution of beta iron FIG. 31. Patch of Martensite from Eutectoid Steel. 2ooX. Quenched from 800 C. in Iced Brine. and cementite, hard and brittle. Moderate reheating or anneal- ing changes this structure largely into troostite (Figs. 32 and 33), which is a partly transformed martensite, possessing much of the hardness of martensite, but with a largely increased toughness and shock resistance. This toughness is the chief characteristic of the next metaral in the transformation series, sorbite (Figs. 34 and 35), which is merely martensite wholly transformed into a mixture of ultramicroscopic crystals of ferrite (alpha iron) and cementite (FeaC). The word tempering should be restricted to denote a moderate reheating, up to about 350 C., forming troostitic steel, while 156 EXPERIMENTAL GROUP IV FIG. 32. Martensite (light areas) pass- ing into Troostite (dark areas). 22oX. Eutectoid Steel, Quenched from 800 C., tempered at 275 C. FIG. 33. Martensite (light needles) passing into Troostite (dark patches). i3oX. From a piece of Eutectoid Steel Electrically Welded. FIG. 34. Sorbite (dark patches) pass- FIG. 35.* Carbon Steel, i.co per ing into Pearlite (wavy striations). Light Areas are Patches of Ferrite. 22oX. From a piece of Hypoeutec- toid Steel Electrically Welded. cent Carbon. 15 oX. Osmond. Pearlite, laminated, passing into sorbite, dark and formless. * Reproduced by permisison from Saveur, " Metallography and Heat Treatment of Iron and Steel." TEMPERING AND TOUGHENING 157 " toughening " represents the practice of reheating hardened carbon steels from 350 C. up "to just below Ac\, and forms sorbitic steel; while " annealing " refers to a heating for grain size at or above the transformation ranges, followed by a slow cooling. Any of these operations not only allows the trans- formations from austenite to pearlite to proceed, but also relieves internal stresses in the steel. Tempering heats have been gaged since time immemorial by the distinctive colors of the thin oxide films which form on bright steel as it is heated from 200 C. to 350 C. A skilled artisan can estimate these colors to a tolerance of a few degrees, which, indeed, is necessary for the production of uniform results. The disadvantages of this handy method are obvious, however. Leaving aside the effect of the skill of the operator, that is, the " personal equation," and the effect of the varying lights due to shop and weather conditions, it should be borne in mind that tempering by colors places reliance on what is strictly a surface phenomenon. The rate of heating, the time at the maximum heat, the mass and configuration of the steel object all affect the color indications, and consequently the uniformity and homogeneity of the results. Ordinary usage is to heat the bar rather rapidly in a furnace which is considerably above the desired degree. In order to make sure that the temperature does not mount too high, it is then necessary to quench the piece after the desired color appears on the surface. This quenching induces, in part, the internal stresses which a proper tempering followed by a slow cooling would relieve. Steam plates, sand baths, molten salt, and metal, are there- fore used in the most modern installations, adequately con- trolled by pyrometers and operating uniformly at the proper temperature. With such equipment it is easy thoroly and properly to heat metal pieces even of irregular cross-section. The subsequent cooling can then be as slow as desired. Special Apparatus. The special apparatus needed is as fol- lows : 158 EXPERIMENTAL GROUP IV Two riddles; one i-in.; one J-in. Hatchet. One 5-gallon pail for quenching bath. One deep metal pan for annealing bath. Wire basket for holding metal. One 300 C. mercury thermometer, metal cased. One electrical meter. Supplies. The supplies needed are as follows: Eleven bars from Experiments 17, 1 8 and 19. Laboratory Equipment. The laboratory equipment needed is as follows: Ice. Coke. Bucking-board and muller. Kindling wood. Anvil and blacksmith's tools. Emery wheel. Vises. Numerical punches. Scleroscope. Impact machine. Drop-hammer, with anvil of mild steel. Annealing oil. Lead pot and pyrometer equipment. Portable rivet forge. Procedure, a. Make a coke fire in the portable forge as follows: Crush about i cu. ft. of coke thru the -in. riddle, screening out the fines with a J-in. riddle. Wet the fines and bank them around the sides of the forge, forming a saucer- shaped depression with the tuyere at the bottom. Cut fine kindlings and build up a good bed of hot wood embers with a gentle blast, gradually adding more and more of the f-in. coke with the wood until a solid bed of glowing coke remains. b. Insert eleven steel bars into the hot bed of coals, one at a time, heating about 2 in. at the end to a bright red, and then TEMPERING AND TOUGHENING 159 forge a round point on the bar. Guard against overheating and burning the bar, and continue the hammering to a low heat. Do not start the heating of the second bar until the first is ham- mered cold. Finish the point on a coarse emery wheel. c. Discuss the results of Experiments Nos. 18 and 19 with an instructor, and decide with him the correct quenching tempera- ture and medium. Notch each bar f in. from the square end to exactly the same depth, punch mark the ends and bodies of the bars with the reheating temperatures used in this experiment, heat the bars according to procedures b to d of Experiment No. 19, and quench according to procedure c of Experiment No. 18. Grind the square end carefully, and test each bar for hardness. Should the hardness of the different bars vary more than 15 per cent the hardening must be repeated until this tolerance is attained. d. Arrange a pan of boiling water over a gas burner, place one bar in the wire basket, and plunge it into the liquid. The bar being reheated must not touch the bottom or sides of the tank, and the annealing fluid must be in constant circulation on all sides of the piece. After fifteen minutes, remove and cool in the air. e. Grind a sharp point with a fine emery wheel, and test its condition by holding the bar upright under the drop-hammer, with the bottom block replaced by a block of mild steel. Raise the hammer i in., release, and examine the condition of the point after the blow. Repeat the drop from 2 in., 3 in., 4 in., etc., until the point breaks, turns over, flattens, or fails in some other manner. Test the square end with the impact machine and scleroscope according to procedures/ and g of Experiment No. 17. Paste a small gummed label around both fragments, giving the carbon content, heat treatment, hardness and toughness. Pre- serve all pieces for reference. /. Reheat four other pieces in a pan of oil, one to the follow- ing temperatures, 200 C. 250 C. 300 C. 160 EXPERIMENTAL GROUP IV stirring the oil constantly, and making sure that the piece is not overheated. These pieces may be withdrawn from a rising temperature at the proper point if at least twenty minutes is consumed in covering the intervals between. Cool each piece in air, and test as in procedure e above. The fourth bar should be held in the hot oil bath at 300 C. for one hour (being extremely careful to maintain a uniform temperature) and then air cooled and tested as the others. g. Reheat six other pieces in the lead pot, one to each of the following temperatures: 400 C. 500 C. Use the precautions noted in procedure / and test each bar as usual. h. Draw curves on coordinate paper showing the variation in hardness and toughness with the reheating temperature as determined by the results of this experiment. Post this on the bulletin board, and exhibit the results of the tests to an instructor. Queries, a. Draw curves in India ink on coordinate paper showing the variation in hardness and toughness with the reheating temperature as determined by the results of this experiment. Show in color on the same sheet a curve of hard- ness against annealing temperature derived from Experiment No. 17. Discuss reasons for similarities or differences in the results of Experiments Nos. 17 and 20. b. Make up an equilibrium diagram for steel on coordinate paper in light black lines, following Fig. 23, page 131. Plot on this diagram the toughness of all the fragments tested by the laboratory squads as posted on the bulletin board. Draw con- tour lines of equal toughness in red TEMPERING AND TOUGHENING 161 c. What effect will the speed of heating from the cold to the tempering or toughening temperature have upon the results? d. What effect will the speed of cooling have? e. How will the length of time at the reheating temperature affect the results of the operation? /. Assuming that the hardened steel bars of eutectoid com- position have a structure essentially of martensite with but a small amount of austenite, what will be the structure persisting after each reheating? EXPERIMENT NO. 21 TOOL MAKING Object. The object of this experiment is to apply the knowledge gained in previous experiments in making a center punch and a cold chisel. General Explanation. Important as are applications of the complex alloys known as " high-speed steels," there are now, and doubtless always will be, enormous numbers of tools, machine parts, and structural members which are made of plain carbon steels. Even some very heavy-duty tools will continue to be made of the simpler carbon steels, since they can be hardened to a much higher degree than can the modern high-speed tools, which latter are pre-eminent for most metal cutting purposes not because of their intrinsic hardness, but because of their ability to retain their moderate hardness when cutting so fast that the sharp edge is heated to an annealing temperature. Among makers and vendors, simple carbon tool-steels are classed by " grade," and " temper." The word grade is qual- ified by many adjectives, some with more or less special or cryptic meanings, but, in general, it has to do with the process and care with which the steel is made. The more important grades may, therefore, be listed as Crucible steel, Open-hearth steel, Bessemer steel. The grade adopted for a particular tool depends upon the pre- cision and life expected of the intrument, and the cost of the tool-maker's labor. Crucible steels are used for such things as fine-edged tools and saws, fine springs, rock drills, precision tool parts, and high-speed weaving machinery parts. Car and wagon 162 TOOL MAKING 163 springs, heavy machinery and locomotive parts can be made amply strong of good open-hearth steel, while sledges, picks and other coarse, battering hand-tools can be satisfactorily made of Bessemer billets. Just why a crucible steel should be better than an open- hearth steel of the same analysis has been productive of much argument, and it is still a contested point, especially by open- hearth steel makers. Metcalf thinks it proven that the lower oxygen and nitrogen content of crucible steels makes them superior to the other grades. The temper of a steel refers to the carbon content of the material. This should preferably be noted by " points," but, unfortunately, a 53-point steel (containing 0.53 per cent of car- bon) may locally be called something like " No. 3 temper." A list of the approximate carbon content and tempering heats favored for many tools and machine parts is appended, taken largely from Bullens, " Steel and its Heat Treatment," Chapter XVI. Carbon Per Cent. Tools. Tempering Heat. Possessing extreme hardness in cutting edge. Toughness a slight factor: i . 50 Lathe tools for tempered gun f orgings i . 40 Lathe tools for chilled rolls Graver tools 215 Brass- working tools i . 30 General lathe tools General slotter tools General planer tools 215 Razors Drawing, trimming and cutting dies 240 Mandrels Granite points Scale pivots Bush hammers 215 Peen hammers 215 Files Ball races Great hardness, combined with some toughness: i . 20 Twist drills 250 Small taps 225 Screw and threading dies 225 164 EXPERIMENTAL GROUP IV Carbon Per Cent. Tools. Tempering Heat. Cutlery 225 Cold stamping dies 240 Leather-cutting dies 225 Cloth and glove dies 240 Nail dies 240 Jeweller's rolls and dies 240 i . 10 Milling and circular cutters 225 Wood-working and -forming tools. Small punches 240 Taps 225 Cup and cone steel 250 Small springs 300 Anvils Toughness and cutting edge about equal considerations: i . oo Reamers 240 Drifts 265 Broaches 275 Large milling cullers 225 Saw swages Springs 300 Rock and channeling drills 225 Large cutting and trimming dies 240 Good cutting edge, but toughness an important factor: o. 90 Hand chisels 270 Chipping chisels 270 Punches 240 Blanking punches and dies 240 Drop dies for cold work 240 Small shear knives 225 Tough tools for withstanding shocks: o. 80 Large shear knives 225 Large chisels 270 Hammers 215 Sledges Cold sets 270 Forging dies 240 Hammer dies 240 Boiler-maker's tools 270 Mason's tools 240 Churn drills 225 o. 70 Track-layer's tools Cupping tools 250 Hot sets Set screws . . TOOL MAKING 165 Carbon Per Cent. Tools. Tempering Heat. Great toughness required, but still suitable for hardening and tempering: o. 60 Hot work battering tools 280 Bolt and rivet headers Hot drop-forging dies Rivet sets Fullers Wedges Toughness the prime consideration : o. 50 Machinery parts Hot dies for bolt-making machines Of course it should be remembered that the above classifi- cation is by no means final the carbon content may vary con- siderably with the presence of some other alloying element; while the tempering heat especially should be carefully adjusted to fit the needs of each class and size of tools or machine, as well as varied to conform with the hardness and other physical and chemical properties of the materials worked upon. Supplies. The supplies needed are as follows: One 6-in. piece of f- or f-in. eutectoid steel. One 6-in. piece of f-in. eutectoid steel. Laboratory Equipment. The laboratory equipment needed is as follows: Ice. Quenching media. Scleroscope. Annealing Oil. Coke. Bucking-board and muller. Anvil and blacksmith's tools. Emery wheel. Forge. Procedure, a. After the student has performed the pre- ceding experiments, he should be able to draw up the proper procedure to make a center punch and a cold chisel. Make up a 166 EXPERIMENTAL GROUP IV brief outline of the various steps in the process and submit it to an instructor for O.K. b. Plan the work and make out two slips, one calling for the complete list of special apparatus required, and the second calling for all the supplies desired. Procure these things of the stock keeper, signing and leaving your slip as a receipt. c. Make good, sharp, tools, of finished appearance, and with the correct hardness and toughness. Submit them to the instructor for testing and O.K. Queries, a. Write up the experiment after the style of the others in this book, beginning with the Special Apparatus, and finishing with detailed Procedure, giving step by step instructions for making this kind of tools. Show how the proper hardening and toughening temperatures are to be found. b. When is it necessary to anneal a tool after forging? c. If either of two analyses would prove satisfactory for a tool, but one required a toughening treatment some 200 C. higher than the other, which steel should be used? Give cogent reasons. d. Is it better to quench and temper moderately, or to quench the same steel quickly, and temper at a higher heat, to give equal hardness? Why? e. Outline a heat treatment for a sledge which would provide a very hard face, underlain by a tough body. EXPERIMENT NO. 22 METALLOGRAPHY OF STEELS Object. The object of this experiment is to produce, examine, and test the various metarals existing in hardened and annealed steels. General Explanation. An excellent paper giving notes on the historical development of metallography has been written by Bradley S tough ton, and was published as " Notes on the Metallography of Steel," in Vol. LIV, Part E, Transactions of the American Society of Civil Engineers, pages 357-421. The names, production and appearances in eutectoid steel of the various decomposition products ranging from austenite to pear lite have been given in former experiments. It should be borne in mind that the structure, as illustrated, is that developed by proper etching agents, as a freshly polished piece of steel appears brightly mirrored at all parts, unless it is a very poor piece of material, containing blowholes, or specks of slag. It is only natural that different etching agents should attack the constituents in various manners and rates. The reagent in universal use for steels is an alcoholic solution of picric acid, made up of Picric acid crystals 5 gm. Absolute alcohol 100 cu.cm. Fay (" Microscopic Examination of Steel," page 17) recom- mends nitric acid as being superior to picric acid for use on hardened steels. He uses the following solution: Nitric acid, sp.gr. 1.42 4 cu.cm. Absolute alcohol 96 cu.cm. 167 168 EXPERIMENTAL GROUP IV It is impossible to distinguish excess cementite from excess ferrite by these reagents, as both of them attack massive crystals but slowly. If the specimen is boiled for five to ten minutes in a solution of sodium picrate, cementite will be colored black. The reagent is made up by dissolving 250 gm. sodium hydrate (NaOH) in 750 cu.cm. of water, and then dissolving 15 gm. picric acid crystals in the solution. It will, perhaps, be well to recapitulate and amplify somewhat the data already presented on steel metarals, in order that the FIG. 30. Patch of Austenite from Eutectoid Steel. 200 X. Quenched from 800 C. in Iced Brine. student may recognize the entities he views in the microscopic field. Austenite has been defined as a solid solution of cementite (FesC) in gamma iron. It is stable at various temperatures dependent upon its carbon content, which may be any amount up to the saturated solution containing 1.7 per cent. Austenite is not nearly as hard as martensite, owing to its content of the soft gamma iron. In picric acid, it etches slowly and irregu- larly sometimes faster and sometimes slower than martensite, and will at various times appear the darker or the lighter of the two. Fig. 30 (reproduced above) shows austenite to possess the METALLOGRAPHY OF STEELS 169 typical appearance of any pure, crystallized substance, cut at an angle to its cleavage. Altho none of the large crystal bound- aries appear in the field, the long-continued action of the etching acid has been to dissolve small particles from the corners of the regularly oriented small particles called crystallites, which together make up the larger crystal. Hence the lace-like regularity of the markings. In the most quickly quenched high carbon steels, austenite commonly forms the ground mass which is interspersed with FIG. 31. Patch of Martensite from Eutectoid Steel. 200 X. Quenched from 800 C. in Iced Brine. martensite, a large field of which is illustrated in Fig. 31 (repro- duced above). Martensite is usually considered to be a solid solution of cementite in beta iron. It is not in equilibrium in any part of the diagram, but represents an unstable condition in which the metal is caught during rapid cooling. It is very hard, due to its content of the hard beta modification, and is the chief constituent of hardened high-carbon steels, and of medium- carbon nickel-steel and manganese-steel. In picric acid, it usually etches lighter than austenite, and always lighter than troostite. The structure of uniform martensite, as shown in Fig. 31, after six minutes etching in picric acid, has the char- 170 EXPERIMENTAL GROUP IV FIG. 32. Martensite (light areas) pass- ing into Troosite (dark areas) . 200 X . Eutectoid Steel, Quenched from 800 C., tempered at 275 C. FIG. 33. Martensite (light needles; passing into Troosite (dark patches) . i3oX. From a piece of Eutectoid Steel Electrically Welded. FIG. 34. Sorbite (dark patches) pass- ing into Pearlite (wavy striations). Light Areas are Patches of Ferrite. 22oX. From a piece of Hypoeutec- toid Steel Electrically Welded. FIG. 35.* Carbon Steel, i.oo per cent Carbon. i5ooX. Osmond. Pearlite, laminated, passing into sorbite, dark and formless. * Reproduced by permission from Saveur, " Metallography and Heat Treatment of Iron and Steel." METALLOGRAPHY OF STEELS 171 acteristics of a pure substance, with the etching lines developed along an octahedral cleavage, appearing roughly parallel to the sides of an equilateral triangle. Indeed, the characteristic appearance of martensite in a field comprised of various metarals consists of these needles, more or less plainly marked, and inter- secting at an angle of 60, such as is shown in Fig. 33 (see page 170). Troostite is a metaral of doubtful composition, but possibly is an unstable mixture of untransformed martensite with sorbite (q. v.). It contains more or less untransformed material, as it is too hard to be composed entirely of the soft alpha modification, and it can also be tempered more or less without changing in appearance. It etches most rapidly a few seconds in picric acid being all that is required to darken the area. Its normal appearance as rounded grains is given in Fig. 33 ; larger patches show practically no relief in their structure, and a photo- graph merely shows a dark, structureless area. (See Fig. 32, page 170.) Sorbite is believed to be an early stage in the formation of pearlite, when the iron and iron carbide originally constituting the solid solution (austenite) have had an opportunity to separate from each other, and the iron has entirely passed into the alpha modification, but the particles are yet too small to be distinguish- able under the microscope. It also, possibly, contains some incompletely transformed matter. Sorbite is softer, tougher, and etches less quickly than troostite, and is habitually asso- ciated with pearlite. As sorbite is merely a mode of aggregation, it has no place on the equilibrium diagram. Its components are tending to coagulate into pearlite, and will do so in a fairly short time at temperatures near Aci, which heat will furnish the necessary molecular freedom. The normal appearance of the substance, however, is the cloudy mass shown in Fig. 34 (page 170), where the metaral has been partially transformed into the stable pearlite. Pearlite is a definite conglomerate of ferrite and cementite containing about six parts of the former to one of the latter. 172 EXPERIMENTAL GROUP IV When pure, it has a carbon content of about 0.95 per cent. It represents the complete transformation of the eutectoid austenite accomplished by slow-cooling of an iron-carbon alloy thru the transformation range. If the steel is held at tem- peratures just below Aci for some time, the thin laminations of cementite agglomerate into fine globules held in a matrix of ferrite. This appearance is called " granular pearlite," or " spheroidal cementite." This condition, which is shown in Fig. 36, seems to be a more stable form of aggregation than the FIG. 36. Spheroidal Cementite, or Granular Pearlite. From a Piece of Eutectoid Tool Steel. familiar lamellar aggregate usually called pearlite, and illus- trated in Fig. 34, page 170. The micrographs thus far shown have been made from eutec- toid steels whose total mass could resolve itself into pearlite if the opportunity presented. Under the microscope, a slowly cooled, lower carbon steel shows more or less primary ferrite, soft, white areas of alpha iron precipitated from the cooling, under- saturated austenite. The balance of the field consists of darker patches of pearlite, material of eutectoid composition and constitution. Fig. 37 shows a coarse-grained, medium-carbon steel magnified sufficiently to show the structure of the dark METALLOGRAPHY OF STEELS 173 etching eutectoid areas. The pearlitic patches seldom have developed striations, however, usually seeming to be merely the dark, formless areas shown in Figs. 38 and 39. This pair of photographs, it may be mentioned in passing, is an excel- lent example of the utility of the microscope for determining the condition of the component crystals of a metal a thing entirely beyond the province of analytical chemistry. In a similar way, hyper-eutectoid steels, if cooled slowly, will show bright primary cementite crystals bordering the dark pearlitic areas. As is also the case with lower car- bon steels, various quenching and annealing practices will suppress the precipitation of the excess material, and retain the whole mass in the various unstable forms noted and illustrated above. Primary cemen- tite crystals, as shown in Figs. 40 and 41, by Mr. E. P. Stenger, need not be confused with excess ferrite. The experienced eye recognizes a marked difference in the " habit" (compare Figs. 37, 40 and 41). A needle will scratch ferrite deeply, but leave cementite untouched. Lastly, sodium picrate will reverse the coloration given cemen- tite by picric or nitric acid. The preparation of specimens for microscopic examination has been described in Experiment No. n, and the general instruc- tions need little addition. A portable motor with vertical shaft carrying a round plate (Fig. 42, page 175), has been found to be a suitable, convenient and inexpensive device. The coarse grind- ing is done by thin emery wheel disks of decreasing coarseness laid on the horizontal plate. These are followed by cloth-covered disks, carrying finer and finer abrasives, and the specimen fin- FIG. 37. Medium Carbon Steel. 220X. 174 EXPERIMENTAL GROUP IV ished on a broadcloth-covered disk with rouge or levigated alumina. The beginner should be very careful to wash the specimen FIG. 38. Low Carbon Boiler Plate. As Rolled. X4o. FIG. 39. Same as Fig. 38. X40. From a Cold Flanged Corner. FIG. 40. High Carbon Ingot. X75- FIG. 41. High Carbon Muff on Case- carburized Disk. X 100. C = i .40%. and his hands after removing one disk and before touching its successor. Disks, when not in use, should be kept dust-free and moist in a compression top can, or under a bell jar. Polished specimens should be kept in a desiccator over night METALLOGRAPHY OF STEELS 175 until their examination is entirely completed. At that time, they should be stored in a fairly tight drawer, enclosed in small pasteboard pill boxes, wrapped in tissue. In this condition, they will remain for years with no appreciable rusting. FIG. 42, Portable Grinding Machine. Special Apparatus. The special apparatus needed is as follows : Five-gallon pail for quenching bath. Four-inch watch glass. Nichrome-tipped forceps. Warm air blower. Soft hand towel. Supplies. The supplies needed are as follows: One piece f-in. eutectoid steel, 4 in. long. Fragments of eutectoid steel from Experiments Nos. 19 and 20. Six tension test bars of 4-in. eutectoid steel, machined. 176 EXPERIMENTAL GROUP IV Metallography Room Equipment. The metallography equip- ment needed is as follows: Electric butt welder. Ice. Salt. Emery wheel. Polishing machine, including Motor, Two fine emery wheels, Three canvas-covered disks, One broadcloth-covered disk, Four bell jars, Four corked Erlenmeyer flasks with suitable abrasive. Etching solutions of picric and nitric acid. Microscopic set, including Brass mounting cup, Can of BB lead shot, Cover glass, Candle lamp with condensing lens complete. Metallographic microscopic set, with Plane glass reflector, Oculars, 5x, lox, Objectives, 8 mm., 16 mm. Scleroscope. Procedure, a. Clamp the bar of f-inch eutectoid steel in the jaws of the electric welder, as directed by the instructor. Arrange a quenching bath of iced brine directly in front of the machine. Turn on the electric current, and when the steel between the jaws of the welder becomes white hot and plastic, move the jaws slowly together. Then quickly interrupt the current, release the jaw grips, and drop the hot bar into the quenching bath. b. Saw off that portion of the steel bar which was enclosed in the welding machine grips, and grind a flat surface J in. wide on METALLOGRAPHY OF STEELS 177 the side of the cylinder remaining. Grind slowly on a wet stone, being very careful to keep the ground surface cold at all times. c. Prepare this surface on the polishing machine according to procedures d and e of Experiment No. n, page 82, noting par- ticularly the precautions regarding cleanliness mentioned in the General Explanation, above. Mount the specimen under the microscope as in procedures/ and g, Experiment No. n, to judge the excellence of the polishing, and to examine for blow- holes or slag specks. d. This specimen will show austenite at the center of the weld, and definite zones of martensite, troostite, and sorbite, leading into the unaffected stock. Place a few cubic centimeters of the alcoholic nitric acid solution in a watch glass, and immerse the polished surface for five seconds. Hold the specimen with forceps and wash quickly in running water, shake off excess, dab (not rub) lightly with a soft towel, and dry with a warm air blast. Examine the piece carefully under the microscope. When the band of troostite, which will develop first, is definitely shown, exhibit the work to the instructor. Photograph the area, if required by the instructor, following the methods of Experiment No. 12. e. Continue the process by etching and examining until the entire surface has been developed to a point where the various components have singly been sharply exhibited and carefully examined. /. Explore the whole surface with a scleroscope. g. Procure the fragment of eutectoid steel resulting from pro- cedure /, Experiment No. 19. Polish, etch, and examine the broken end, to determine the constitution. h. Polish, etch, and examine the broken ends of all the eutectoid bars from Experiment No. 20, obtained after all reheatings up to A c 1 + 25. i. From the above work, decide upon the heat treatments which will probably produce bars entirely of austenite, marten- site, troostite, sorbite, pearlite, and coagulated cementite. Pre- pare an outline of the necessary procedure to produce these test 178 EXPERIMENTAL GROUP IV bars, and submit it to the instructor for O.K. Make up a complete list of all the special apparatus and supplies needed, and draw them from the stock-room, leaving the signed list as a receipt. Prepare the test bars, polishing and examining the end of each bar to determine whether it has responded to heat treat- ment as predicted. j. Test these bars in the tension machine in the testing laboratory, either drawing a stress-strain curve autographically, or observing the necessary elongations with an extensometer. Determine also the hardness and impact strength of each bar- end. Queries, a. Sketch the appearance of the polished and etched weld, with proper notations recording the variations in hardness. b. Make up a neat tabulation of all the metarals of steels, showing the following properties of each, as far as determined or known : Metaral. Equilibrium range. Constitution. Composition. Etching characteristics. Microscopic appearance. Physical properties: Hardness, Elastic limit, Ultimate strength, Elongation. Mode of production in small pieces of eutectoid steel On quenching, On annealing or tempering. c. Figure the composition of pearlite, in percentage of ferrite and cementite. d. Why does a steel containing 0.950.50 per cent carbon show a larger area of primary crystals than does another con- taining 0.95+0.50 per cent carbon? METALLOGRAPHY OF STEELS 179 e. Why should an eutectic structure be fine grained? /. Double quenching is sometimes used in coarsely crystalline bars to raise the elastic limit and tensile strength. Outline a heat treatment program by which the same result could be effected without the danger of hardening cracks, which would almost assuredly appear on double quenching in any but low or medium carbon steels. g. Give a full explanation of the changes which take place in pieces of carbon steels of the following composition during heating and cooling: 0.03 per cent carbon. 0.60 -95 1. 10 EXPERIMENT NO. 23 CASE-CARBURIZING Object. The object of this experiment is to study the cemen- tation of a mild steel bar. General Explanation.* " Cementation processes " include those which enrich the surface of wrought iron or low-carbon steel to various depths by the addition of cementite (FesC). The method consists of heating a carbonaceous substance in close contact with the iron to be carburized, in a suitable con- tainer. From the iron-carbon equilibrium diagram, it is evident that Ac i is the minimum temperature at which true cementation can proceed; graphitic carbon will diffuse into iron at lower tem- peratures than this by reason of the forces due to a difference in concentration, and if it does combine with the iron, the carbide produced cannot enter into solid solution until Ac\ is exceeded. Otherwise the properties of a more highly carburized steel are lacking the added cementite does not form increasingly large areas of pearlite, but remains as a fine meshwork of free crystals wherever formed. Apparently, the most obvious variation in cemented metal is the depth to which the carburization extends. On this basis has grown a distinction between " total " cementation, where the piece is intended to be wholly transformed into a high-carbon steel, and " partial " cementation, where the process is frankly limited to the production of a thin, hard case overlying the orig- inal low-carbon, tough material. As a matter of fact, the term * This explanation is condensed from a longer article by the author appear- ing in XVI, "Metallurgical and Chemical Engineering," 385. The great au- thority, on the matter of case-carburizing is Giolitti, " The Cementation of Iron and Steel." 180 CASE-CARBURIZING 181 " total cementation " is a misnomer, inasmuch as the process is seldom if ever continued for a time when there is a considerable increase in the carbon content at the very center of the metal. A description of the methods of cementation for the produc- tion of blister steel is beside the scope of the present discussion, and the student is referred to the article noted above, or to Mills, " Materials of Construction," pages 363 to 365. Altho the manufacture of blister steel may be on the decline, the advances in machine design call for an increasingly large number of parts which are hard enough to resist wear, but are tough enough to withstand shocks. It is well known that a high hardness numeral is had at the expense of ductility, but the combination of these desirable but contradictory qualities may happily be had by transforming the outer surface of a tough, low- carbon piece into a hard, high-carbon steel. Again, case- hardening operations are the favorite method for producing cheap metallic pieces for taking a high polish. Or, lastly, the ends of such articles as axles may be carburized for the wear against bearings, while the rest of the bar can be entirely unaltered. The methods for this most important operation of case-car- burization are various, depending first of all upon the cement used, whether solid, liquid or gaseous, or some combination of these three. As a matter of fact, most case-carburization is still done by packing the pieces in crushed charcoal or other car- bonaceous material in closed metallic boxes, and then heating the combination to a proper degree, and for a proper time. A superficial consideration would suggest that the carbur- izing action is due to the diffusion of the solid carbon of the cement into the carbon-poor iron. Doubt is thrown upon this view when it is known that charcoal, as a cement, is rapidly " exhausted "; that is to say, its efficiency is continuously im- paired upon re-use. Careful experimenting has shown that actual physical contact of the iron and carbon is the prime requisite of solid diffusion. When this requirement is rigorously met, a three and one-half hour cementation in pulverized carbon at 1000 C. will increase the pearlitic areas very slightly to a depth 182 EXPERIMENTAL GROUP IV of 0.15 mm., while any portion more than this distance from a carbon contact will be absolutely unaltered. Evidently the deeper carbonized layers so easily produced in practice must be explained in some other way. Without dwelling at length upon the various other possibili- ties, we may accept it as definitely proven that carbon monoxide (CO) is the most active cementing material existing in the ordinary carbonizing box, and fortunately the most easily con- trolled. Interaction between the oxygen of the air occluded in the charcoal pores and entrapped in the cementation box with the heated carbon produces quantities of carbon dioxide (CO2) at low temperatures. At the temperature of ordinary cementation over 1000 C. most of the carbon dioxide gas decomposes, if in contact with an excess of carbon, forming carbon monoxide (CO). This is indicated by the reversible reaction The resulting gas has a very small proportion of carbon dioxide, the actual amounts of each gas in equilibrium varying with the temperature and pressure. These conditions of equilibrium may be noted, in a short-hand way, by saying that there exists an equilibrium in the system C : CO : C02. The gas resulting from this equilibrium, particularly the carbon monoxide, easily diffuses into the hot iron and reacts with it, forming iron car- bide, which is immediately absorbed into the solid solution austenite. The reaction may be The dioxide formed at this point diffuses outward to a region of lower concentration, while the CO : CO2 diffusing inward will form more iron carbide on its way to the center of the piece of metal if demanded by the state of the system, Austenite: CO : CO2, existing at that particular locality. The carbon dioxide escaping from the steel is immediately regenerated by CASE-CARBURIZING 183 the excess of hot carbon to a condition represented by the equilib- rium of the system C : CO : CO2- It is clearly seen that carburization will cease when the relative concentration of the gases in the packing represented by C : CO : CO2 equals that existing in the steel, which is repre- sented by Austenite : CO : C02- As a matter of fact the maximum carbon concentration in a case-hardened article never reaches this theoretical amount, simply because the proc- ess is interrupted long before equilibrium is established. Of all the solid cements, wood charcoal is the best; first, because it is pure, and therefore maintains the purity of the product; and second, it is of simple composition, and therefore the carburizing reactions are not complicated by variable and unknown factors. This allows its use with certainty of control; the results can be predicted with considerable accuracy. The greatest disadvantages are the slow speed of the cement and the phenomenon of " exhaustion," both of which now seem to be due to the limited amount of carbon monoxide formed in the box. In order to increase the speed of the cement, it is necessary to add one-third to one-half unused charcoal to every packing, or better, to mix into the charcoal in the first place some 40 per cent of powdered witherite (BaCOs). The latter procedure, recom- mended by Caron in 1861, is especially valuable in increasing the efficacy of the cement by bringing a large amount of gas to the cementation chamber in the following manner: BaCO 3 <=BaO + CO 2 , The first reaction goes toward the left at lower temperatures CO 2 being rapidly absorbed upon exposure to the air. Such a combination has the added advantage, therefore, of being inex- haustible; it is largely used today, and all things considered, is the best solid cement for " partial " cementation. Innumerable preparations have been recommended and are on the market for use as solid cements, but " there are only some types of simple cements, at present well known, whose efficacy 184 EXPERIMENTAL GROUP IV is a maximum from all points of view, and the addition of other ingredients in no wise increase their efficacy " (Giolitti, " The Ce- mentation of Iron and Steel," page 261). Unfortunately, Caron's cement requires considerable time to produce its deep- seated carbonizations. More " rapid " cements are made of many mixtures of sawdust, charcoal, bone, lampblack, tar and various hydrocarbons, alkaline and alkaline earth carbonates and cyanides, ferro- and ferri-cyanides; together with homeopathic proportions of numberless other substances, all more or less inert. Such compounds act by virtue of the large quantities of hydro- carbons or cyanides which are evolved on heating; they exhaust rapidly, give very irregular and non-uniform carbon penetra- tions with a rapid, discontinuous rise from the non-cemented core to the thin, excessively high-carbon concentrations at the outside. The variable quantities of complex gaseous mixtures evolved in the cementation box preclude even a moderately close control of the process, and result 'in a non-uniform product containing a hard, brittle case badly addicted to exfoliation. In cementation as well as other processes, the simplest combina- tions are managed with difficulty, while any such variable and complex mixtures as are produced by the destructive distillation of organic materials are absolutely uncontrollable. The use of a molten bath of carburizing materials presents a great many apparent advantages as to ease of manipulation and cheapness of product. The materials composing these baths are usually cyanides or cyanogen compounds, all highly poisonous. The chemical reactions occurring in such baths have not been studied with precision, but it seems that the cyanogen (CN) 2 decomposes in contact with the iron yielding carbon, which is absorbed into the surface layers of the metal. The time of cementation is so short, however, that the carbide thus formed has little chance to penetrate into the metal, and the result is a very thin zone (0.03 to o.i mm.) of hypereutectoid steel. Deeper zones result in very brittle and higher carbon muffs which are liable to split off when quenched, i.e., " exfoliate." Since carbon monoxide gas is the active cementation agent CASE-CARBURIZING 185 even when the so-called solid cements are used, it would appear that the best course would be frankly to adopt this gas as a carburizing agent. Theoretically, it would seem that a high degree of uniformity in the case-carburized bars could be pro- duced with certainty by feeding to the container a gas of the proper composition (xCO+yCO2) which, at the temperature of the furnace, would be in equilibrium with the system *Austenite : CO : CO 2 . where x Austenite represents the per cent of carbon desired in the muff. In such a case, the time element would be absolutely powerless to increase the maximum carbon content of the case, it would only produce a deeper and deeper zone of uniform compo- sition. Practically, however, the equilibrium represented by # Austen- ite is approached in the outer layer of the steel with extreme slowness, owing to the counter effect of the carbon dioxide dif- fusing from the lower carbon austenite at the center of the bar outward toward the gaseous atmosphere. For, as the action of carbon monoxide going inward is a carburization, for like reasons a stream of carbon dioxide coming out will decarburize. Even so, the carbon content of the finished bar carburized in carbon monoxide diminishes from the case inward with absolute uni- formity and even this slow process would be used where a zone of this characteristic (called Type I by Giolitti) would be desir- able. Cemented zones belonging to Giolitti 's Type II are pro- duced by the use of ethylene, methane, acetylene, or other hydro- carbons which decompose at elevated temperatures when in contact with iron, depositing a larger or smaller quantity of the finest carbon on the outside of the specimen. The distinctive feature of cementations of Type II is the possession of a hyper- eutectoid case, whereas Type I always produces hypoeutectoid zones. In addition to this characteristic, a slowly cooled zone of Type II shows the existence of three distinct zones shown plainly in the micrograph, Fig. 43 (reproduced from Giolitti, c% 1.4 1.2 1.0 0.8 O.G 0.4 0.2 "v^ >'. X L- :v \ V \ \ 0.5 1 1.5 2 2.5 3 MM. FIG. 44. Concentration Depth Diagram of Bar Carburized with Ethylene. 5 Hours at 1050 C. NOTE Figs. 43 to 46 inclusive have been reproduced from Giolotti. " Cementation of Iron and Steel," by permission of the McGraw- Hill Book Co. FIG. 43. Microsection of Edge of Bar Carburized with Ethylene. 5 Hours at 1050 C. 186 CASE-CARBURIZING 187 Fig. 38), and the corresponding concentration-depth diagram, Fig. 44 (reproduced from Giolitti, Fig. 37), which were produced by a four-hour cementation in ethylene, at 1050 C. The mechanism of this action is not quite clear, but it is possible that the hydrocarbons decompose so rapidly upon entering the steel that a deposit of carbon is formed on the out- side of the steel much more rapidly than it can diffuse into the metal. This lavish deposition of carbon evidently continues as the gases penetrate further, the direct action of the decomposing gases being intensified by the presence of the excess of carbon deposited on the surface of the sample as well as being possibly reinforced by the diffusion of the carbon bodily into the steel (perhaps a considerable factor in this instance owing to the fine subdivision and intimate contact of the sooty deposit). For these reasons a high carbon concentration is obtained with a zone of zAustenite of considerable thickness (i mm.-|-). It is therefore evident that the best results (from a practical standpoint) may be attained by a proper mixture of carbon mon- oxide and one of the gases of Type II. These actually do pro- duce most desirable results, called Type III, giving rise to car- burized cases of varying thicknesses approximating eutectoid composition at the exterior, underlain by steel of constantly and uniformly decreasing carbon content, merging into the unaffected core. The characteristics of this class of carburized zones are illustrated in the micrograph, Fig. 45 (reproduced from Fig. 142, Giolitti), and the concentration-depth diagram, Fig. 46 (reproduced from Giolitti, Fig. 45). Zones of Type III are doubtless produced in the following manner: The hydrocarbons in contact with the outer layer of the iron decompose, depositing free carbon. This carbon con- tinually regenerates the CCb diffusing outward, and the resulting CO really acts as a vehicle, transporting the carbon to deeper and deeper regions, where it could not possibly penetrate by solid diffusion on account of time and temperature limitations. Very important advances have recently been made in the construction of furnaces for the practice of cementation with solid FIG. 45. Microsection of Edge of Bar Carburized with Giolitti's Mixed Cement. 2 Hours at 1000 C. 50 X. 1 0.8 0.6 0.4 0.2 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 ? 2.2 2.4 FIG. 46. Concentration Depth Diagram of Bar Carburized with CO+3.i%C 2 H 4 . 4 Hours at 1000 C. 188 CASE-CARBURIZING 189 or gaseous cements. For a description of these improved appa- ratus, the interested student will refer to the works mentioned at the head of this general explanation. In comparison with the old-fashioned furnaces which plant-managers refuse to scrap because they fail to recognize their economic wastefulness, the extraordinary work of these new types may be illustrated by saying that with a gaseous cement a penetration of from 0.7 to 1.2 mm. may be obtained in an hour with a maximum carbon concentration approaching 0.9 per cent. Such a case, by the way, is ample for 99 per cent of all machine parts subjected to partial carbonization. Using properly designed furnaces oper- ating with solid cements, from ten to twenty cementations can be effected in a day of twenty-four hours, depending upon the depth of case required, turning out from one to five tons of material per furnace, at a maximum cost of one cent per pound one-fifth that of carburization in boxes. A perplexing and most troublesome phenomenon attendant upon the use of cemented steel objects is the exfoliation or split- ting of the carburized zone from the piece during quenching, or in use. Dr. Giolitti shows that these failures occur at points where a sudden and discontinuous variation in the carbon content occurs in the carburized zone as in the case of the middle zone of cementations of Type II. The reasons why this failure occurs at such points, and how it can be avoided have been explained by E. P. Stenger, in an article published in XVI " Metallurgical and Chemical Engineering," 424. Special Apparatus. The special, apparatus needed is as fol- lows: One piece 2-in. gas pipe, 4 in. long, threaded. Two 2-in. pipe sleeves. Two 2-in. pipe plugs. Breast drill and f-in. bit. Monkey wrench. Large pipe wrench. Electrical meter. Special squads will need a supply of f-in pipe, f-in. pipe fittings, drum of CO. 190 EXPERIMENTAL GROUP IV Supplies. The supplies needed are as follows: Three round f -in. bars of 2o-point carbon steel, 6 in. long. Can of carburizer. Laboratory Equipment. The laboratory equipment needed is as follows: Pipe vise. Ice. Emery wheel with wire buffing brush. Impact machine. Scleroscope. Gas furnace at 1000 C., with temperature regulator. Metallography Room Equipment. The metallography room equipment needed is as follows: Polishing machine, complete 1 ^ . , AT . ,. 3 . \ v. Experiment No. 22. Microscopic set, complete Etching solution of picric or nitric acid. Procedure, a. Make a carburizing box for the f-in. round test pieces as follows: On either end of a piece of 2-in. gas pipe, 4 in. long, screw a sleeve, and into the sleeve a plug. One of the plugs should have a f-in. pyrometer hole in its center. b. Pack the box by placing about \ in. of the solid carburizer in the bottom, and then standing the three test bars equidistant from each other and the walls. Fill the remainder of the box with the carbonaceous material, screw in the top plug and in- sert the asbestos protected thermo-couple (Experiment No. 7) thru the pyrometer hole well toward the center of the mass. Lute the hole shut with asbestos string. c. Place this box in a hardening furnace held at 1000 C. with a temperature regulator. The temperature of the pieces in the carburizing box should be held at a constant maximum for three hours. One squad member should give his entire time to the furnace control, reading and plotting the temperature in the box and of the furnace at short intervals. Cool in the furnace. CASE-CARBURIZING 19l d. In the following laboratory period, open the box, observe the condition of the bars, and clean them with a wire buffing brush. Break three pieces from the end of one bar in the impact machine, according to procedure/, Experiment No. 17, taking the average result as the toughness. Grind one end of the frag- ments and explore the area carefully for variations in hardness. Preserve one fragment to exhibit the coarseness of the frac- ture, and paste a label around it, giving its properties and history. e. Polish one end of another fragment (procedure d and e, Experiment No. n), etch (procedure d, Experiment No. 22), and examine under the microscope (procedure / and g, Experi- ment No. n). Sketch the appearance of the microscopic field according to procedure /, Experiment No. n, or photograph. With the aid of the instructor, estimate the carbon content of the case, and sketch an approximate concentration-depth dia- gram, like Fig. 46. /. Refine the grain of the two remaining pieces by annealing at the proper temperature, according to Experiment No. 17. Test one of the annealed pieces as in procedure d, above. g. Harden the case on the remaining piece by quenching in oil from the proper temperature, according to Experiment No. 1 8 and 19. Test the piece as in procedure d, above. h. Mount representative fragments by thrusting them thru .small holes in a sheet of cardboard. Add a tabulation of the data and a copy of the sketches and temperature curves obtained during the experiment. Post the whole on the bulletin board for examination by the various squads. NOTE. Each squad should have individual instructions as to time, temperature, and agent of cementation. In this way the final results can be examined by all students, and will exhibit how the properties of case-carburized bars vary with the con- ditions. A suggested program of such variations is as follows: I. To show the effect of constancy of temperature. Use Caron's cement. 192 EXPERIMENTAL GROUP IV Squad i. Heat three hours at 1000 C., in hand-controlled oven furnace. Squad 2. Heat three hours at 1000 C., in hardening furnace equipped with automatic temperature control. Squad 3. Heat three hours at 1000 C., in an electric tube furnace. Squad 4. Heat three hours in oven furnace purposely oscillating the temperature from 900 to 1000 C. II. To show the effect of varying the carburizing agent: Compare the work of Squad 2 with Squad 5. Use crushed charcoal only thru i^-in. on -in riddle. Squad 6. Use the same material after Squad 5 has used it once. Squad 7. Use the same material after Squad 5 and 6. Connect a drum of CO to the carburizing box by a |-in. pipe, and pass a very slow stream of gas at all times. Squad 8. Pack the bars in crushed charcoal, connect a f-in. pipe to the air supply, and pass a very slow stream of air thru the box, starting when the bars have reached the desired temperature. A few cubic centimeters of gas per minute is all that is required. Squad 9. Use an inert substance like crushed silica brick for a packing and pass a slow stream of CO thru the box at all times. Squad 10. Pack as Squad 9, but use a slow stream of natural gas. III. To show the effect of temperature: Compare the work of Squad 2 with Squid ii. Working in exactly the same way as Squad 2, except at 800 C. IV. To show the effect of time: Compare the work of Squad 2 with Squad 12. Working in exactly the same way as Squad 2, except that the heating should be continued only i \ hours, and the pieces taken from the hot container and cooled in air. Queries, a. Why is it impossible to produce a bar of uni-. form carbon content edge to center, by the method of " total " cementation? b. If a piece of wrought iron were exposed to a continual stream of carbon monoxide gas at a temperature above Ac\, what would be the ultimate carbon content of the bar? Why? c. Sketch a modern furnace for carburization with solid cement. d. Discuss the effect of temperature variation during carburi- zation on the properties of the resulting bar. e. Discuss the effect of time at the carburizing heat on the properties of the resulting bar. CASE-CARBURIZING 193 /. Discuss the results of carburizing with different agents. Why does charcoal exhaust itself? Why does not Caron's cement do likewise? Explain the " regeneration " of charcoal by a stream of CO or air. Is the case resulting from a carburiza- tion in natural gas a representative of Type I, II, or III? Why? g. Describe the process of manufacture of blister steel. h. Discuss the causes and remedy for exfoliation. EXPERIMENT NO. 24 CORROSION Object. The object of this experiment is to study the cor- rosion of steel. General Explanation.* It is a matter of universal experience that dry metal does not rust. All scientific hypotheses as to the nature of corrosion require the presence of water, which we will assume to be the rain, surface, or ground water. All terres- trial water is impure, containing a greater or less proportion of other matter; that is to say, it is a " solution " of organic and inorganic matter. It is well, therefore, first to consider the prop- erties of solutions. If a quantity of soluble salt is placed in pure water, it will enter into solution at a certain rate. The molecules of the solid tend to distribute themselves among the molecules of the liquid equi-spatially. The speed of solution progressively diminishes owing to a back-pressure called " osmotic pressure " exerted by the molecules which have gone into solution, until when the sat- uration point is reached the solution pressure is exactly balanced by the osmotic pressure, and any molecules which may go into solution will be replaced by others coming out of solution. The system is one of balanced activity, or equilibrium. Arrhenius first pointed out the fact that the molecule? of inorganic salts, acids, or bases in aqueous solution are dissociated when entering into a solution. The main arguments for this theory were drawn from a study of osmotic pressure. Osmotic pressure may be measured by suitable apparatus, and for an organic solution, such as sugar in water, it has been * The best authority on corrosion is Cushman and Gardner, " The Corrosion and Preservation of Iron and Steel." 194 CORROSION 195 found to obey the gas laws. In other words, osmotic pressure varies as the absolute temperature, and as the concentration of molecules; in fact, the osmotic pressure is exactly equal to the pressure it would exert -were the substance a gas at the same volume and temperature. Inorganic substances, however, ex- hibit an osmotic pressure two, three, or four times as much as the same molecular concentration of organic substances. Again, if a certain concentration of sugar depresses the freezing-point or the vapor pressure of a solution an amount equal to one unit, the same molecular concentration of salt (NaCl) will depress it two units; while barium chloride (BaCk) will depress it three units. These various facts were explained by Arrhenius by the assumption that the sugar molecule enters the solution as such, but that salt enters as two distinct particles (Na) and (Cl), while BaCl 2 enters as three: (Ba), (Cl), and (Cl). It is unquestionably true that the dissociated parts of an inorganic salt are in some manner different from elemental matter. In the case of common salt fNaCl) the sodium is not in the atomic state, else it would violently attack water, liberat- ing hydrogen, while atomic chlorine would color the solution and give a characteristic odor. None of these things happen when salt goes into solution. This has been explained by the theory that the instant the molecule of salt dissociates, the particles become " ionized." That is to say, the sodium (which is possibly only a definite aggregate of minute electrical par- ticles of high intrinsic energy but small mass called " electrons ") is no longer an atom but an " ion " because it contains one too many electrons, and the chlorine is no longer atomic but ionic because it lacks the one electron attached to the sodium. The sodium is thus charged positively, while the chlorine is charged negatively. Due to the attraction of the unlike charges the sodium and chlorine ions are continually recombining, but also continually dissociating the system again being one of bal- anced activity or equilibrium. A directional impetus could evidently be given these wander- 196 EXPERIMENTAL GROUP IV ing ions by immersing a pair of places into a dissociated solution and maintaining a positive charge on the one and a negative charge on the other. The first would attract the negatively charged ion, and vice versa. This experiment is easily per- formed with apparatus as shown in Fig. 47. The battery sends a constant supply of positive particles of electricity to the copper plate, and an equal supply of negative particles to the iron. In solution, the molecules of sulfuric acid (H2SO4) are dissociated into three ions which may be noted. (H) + (H) + (S0 4 )~. The hydrogen ions, positively charged and free to move, will be pulled toward the iron plate and pushed away from the copper 4- Wire 6 Volt Battery Cu /^ , AAA/ Resistance 1 Fe -Wire f -*-Beaker with Ifc Ho SO 4 Solution ^D-J FIG. 47. Electrolysis of Inorganic Solutions. plate, and will continuously come into contact with the negative iron electrode. For similar reasons we expect the (SCU)"" ions to simultaneously reach the positive copper electrode. As soon as the hydrogen ion touches the anode, a certain amount of electricity is neutralized, and the hydrogen now becomes atomic, unites with a second atom, forming molecules of hydrogen gas, which gather in bubbles on the surface of the metal and escape from the solution. The (864) ion, on coming into contact with the electrode, and losing its electrical charge becomes a chemical entity of extreme activity, attacking the metal with which it is in contact forming copper sulfate, CuSO4, which being soluble, enters CORROSION 197 the solution and immediately dissociates. The presence of copper in the solution is quickly detected by the characteristic blue color in the vicinity of the cathode. The migration of the ions in this case involves the neutraliza- tion of a considerable quantity of electricity at the electrodes. This is replenished by a continual stream sent out from the bat- tery or other source, and the net effect is that of a continuous current of electricity flowing thru the whole system, including the solution. The electrical conductivity of ionized solutions has given them the name of " electrolytes." The source of supply of electrons may even be a contact Copper Wires Seat of- Contact E.M.F. Beaker with-*. Solution Fe Cu - /- FIG. 48. Electrolysis by Contact E. M. F. electromotive force. In Experiment No. 6 it was shown that the joint between two metals of unlike chemical composition or physical constitution was the seat of a " contact electro- motive force " projecting positive electrons in one direction, and negative in the other. The potency of this force in causing metallic corrosion by methods precisely similar to those studied may be demonstrated by apparatus shown in Fig. 48. The contact between the iron and copper is a seat of elec- tromotive force discharging positive electrons to the iron, and negative to the copper. Dipped into a dissociated solution (as of sulfuric acid, for instance) the iron is the anode, attracting the sulfate radical, which on contact becomes a chemical entity and attacks the iron taking one atom into solution for each two par- 198 EXPERIMENTAL GROUP IV tides of electricity neutralized. On the hand the hydrogen " plates out " on the copper, accumulating in gas bubbles. It will be noted that the current flowing in this system rapidly drops away from its initial value, and the cell is then said to be " polarized." This is principally due to the fact that the hydrogen film collecting on the copper is a poor con- ductor of electricity like all gases and its presence largely increases the resistance of the circuit. With a constant contact electro- motive force, the current passing will necessarily vary with the total resistance. It was found in Experiment 6 that a mere difference hi physi- cal constitution existing between two metals was sufficient to induce a contact potential at their joint. In this connection one can easily demonstrate a flow of current in an electrolyte, with the consequent corrosion of one of the electrodes, induced by the contact potential existing between a hardened and an annealed wire of exactly the same chemical composition. Indeed, our study of the microscopic constitution of steels has demonstrated the essential non-homogeneity even of pure iron -carbon alloys. Practically, all commercial irons have additional discontinuities of structure due to slag or scale inclusions, segregations of ele- ments, or physically strained spots or portions. Each border area of such a discontinuity will be the seat of contact electro- motive force, causing differences of potential at various portions of the metallic surface. One can easily explain the replacement of metal from solution on this basis. Suppose a rod of commercial iron to be immersed in a solution of copper sulfate. Consider the action of two infinitesimal areas of different potential (Fig. 49). The copper ion will be attracted to the negatively charged area, come into electrical contact and plate out as a metal. The S04 ion will be attracted to the positively charged area, come into electrical contact, the charges neutralize, and the SO4 becomes chemically active, immediately attacking the iron forming iron sulfate, which, being soluble, enters the solution and is dissociated. In this manner, all the copper in the solution CORROSION 199 will eventually be replaced by metallic iron. This method is in extensive use in copper mining districts to recover the copper from the mine waters, and, indeed, to recover copper leached from low-grade tailing dumps by rain water. In case the iron is dipped in very dilute salt or acid solutions (tap water) the case is somewhat different. Suppose the solution to be a weak solution of sulfuric acid, H2SO4. As discussed above, different surface potentials would cause a migration of the ions, the hydrogen plating out as a gas film, and a corre- sponding quantity of iron sulfate being formed. The action is Iron Rod -Cu' -so 4 - Beakcr with KCu SO* Solution FIG. 49. Electrolytic Corrosion. quickly polarized, however, owing to the non-conducting gas film offering a very large resistance to the electrical interchange. This is not the case if iron replaces copper, as the copper plate is an excellent electrical conductor, and is porous enough to allow the penetration of iron and sulfate ions. Iron in the purest water will be taken into solution slowly, owing to the fact that even chemically pure water is slightly dissociated into (H) + and (OH)~. The result of the interchange is ferrous hydrate, rust. Polarization in this case is very quick and the action is exceedingly slow. Polarization is possibly somewhat akin to the so-called " passive state " of metals, which is merely a temporary sup- 200 EXPERIMENTAL GROUP IV pression of their expected action (chiefly as far as solution pres- sure is concerned) caused by dipping the metal piece in a strong oxidizer, such as chromic acid, chromates, or strong fuming nitric acid. The passive state of metals is not yet explained, but it may be that the action of the oxidizer changes Fe ++ (assumed to be active) into Fe + " " (assumed to be passive). Much work on the prevention of corrosion has been done with the hope of discovering some method of making the metal permanently passive. The mechanism of the rusting of iron in ground water may now be illustrated in the following experiment. Ground water is Fe Fe Kesistance -==- C Yolt - _ Battery Beaker with \% Salt Solution FIG. 50. -Mechanism of Corrosion. a complex solution of mineral salts with more or less oxygen and carbonic acid in solution in general, it is an oxidizing substance. In the experiment, in order to hasten the action, two bars of iron are used for electrodes to represent areas of different poten- tial, the contact electro-motive force is replaced by a battery, and the ground water represented by a i per cent salt solution. The salt in solution is dissociated into a sodium cation and a chlorine anion, the former of which is attracted to the negatively charged electrode, and the chlorine to the other. On coming into electrical contact the chlorine becomes a chemical atom, extremely corrosive, and attacks the iron, according to the fol- lowing equation : Fe+Cl 3 = FeCl 3 . CORROSION 201 The iron chloride is soluble, enters' into solution and is dis- sociated. The sodium, on the other hand, attacks the water of the electrolyte as follows: the sodium hydrate is soluble, and dissociates in solution, while the hydrogen appears on the electrode in gas bubbles. Soon the OH ion meets an iron ion in its migration, and the reaction 3NaOH+3FeCl 3 = Fe(OH) 3 +3NaCl proceeds; the insoluble iron hydrate (rust) being formed in large quantities with the constant regeneration of the salt in the electrolyte. It can be appreciated that only a small electromotive force and a small amount of salt is sufficient to do irreparable damage to a piece of metal if a constant supply of water is available, and the action does not become polarized. The following list of factors affecting corrosion from Cushman and Gardner, " Corrosion and Preservation of Iron and Steel," page 123, needs no further discussion if the preceding experi- ments are understood Factors which Stimulate Corrosion, i. Impure and segre- gated metal. Unhomogeneous or burnt metal which may con- tain blowholes. 2. Cold-rolled or improperly annealed metal which may main- tain an uneven, stressed or strained condition. 3. Contact action, in which different types of iron and steel are used in one and the same structure. 4. The presence of hydrogen ions from any source whatso- ever that may be brought in contact with the surface in water or other electrolytes in the presence of oxygen. 5. The concentration of active oxygen that is present in the wetting medium. 6. The presence of electrolytes generally in the wetting medium. 202 EXPERIMENTAL GROUP IV 7. The stimulating or depolarizing effects of certain coatings applied to iron and steel with the object of protecting the metal. 8. The effect of indentations, scratches or other injuries which become centers of corrosion. 9. The effect of extraneous or stray currents escaped from high-potential circuits. Factors which Inhibit Corrosion, i. In most cases, the reverse of the conditions which stimulate corrosion. 2. Contact with certain substances in solution, such as chromic acid and its soluble salts, which produce a passive con- dition. 3. Alkaline solutions of all kinds, where the concentration of hydroxyl ions is sufficiently high. But this protection may be overcome in very strong boiling solutions. 4. Contact with more electro-positive metals. Much matter regarding corrosion is now being published, both by interested corporations and disinterested scientists. The results of short-time tests seldom form a basis for comparison on account of the extreme complexity of the reactions, and the difficulty of adequately controlling the influencing factors. This can be seen from a consideration of the following list of factors affecting the rate of corrosion, prepared and discussed by Friend : 1 . Quantity of dissolved oxygen in water. 2. Area of exposed metal. 3. Superficial area of water. 4. Depth of immersion. 5. Pressure of oxygen. 6. Rate of motion of water. 7. Partial immersion. 8. Physical condition of iron. 9. Light. 10. Temperature. 11. Presence of rust. 12. Time. 13. Biological influences. CORROSION 203 One very beautiful experiment called the " ferroxyl " test has been developed by Walker to show the actual migration of iron ions away from the solid bar. The reagent ferroxyl is a gelatine which contains two delicate indicators; phenolphthalein and potassium ferricyanide, the former of which will turn pink in the presence of the basic ion (OH)~ and the latter turns blue in the presence of iron (Fe) ++ . If then, a piece of iron is immersed in the gelatine, the pres- ence of a pink area will indicate a negative node, while the presence of a blue zone will indicate a positive node: iron is being corroded at this point. The gelatine is present merely to offer greater resistance to the migration of the ions, i.e., to " fix " the reaction. Such substances as litmus, phenolphthalein, methyl orange, and potassium ferricyanide are called " indicators " because they show the presence of slight traces of certain reagents. Phenol- phthalein, for instance, is a very weak acid, a compound of phenol (carbolic acid) and naphthalic acid (another coal-tar derivative); it is not dissociated in a neutral solution, and is, therefore, colorless. However, if a soluble base is added, a thalic salt is formed with the positive ion from the base, which thalic salt does dissociate, when the characteristic rose-colored Lhalic ion immediately colors the solution. Phenolphthalein is thus an indicator of hydroxyl ions, i.e., of bases, to a dilution of i part in 3 million. Homogeneous steel free from segregation and physical strains undoubtedly is a better rust-resistant material than an impure, poorly made article. There is reason to believe, how- ever, that the best wrought iron is a superior material; perhaps because of the mechanical protection afforded by the inert slag inclusions. Busy iron (such as rails and pipe carrying running water), usually last better than otherwise, perhaps because the mechanical forces break up and spread the nodes so that localized pitting is impossible. It is also known that spongy rust is electro-negative to iron and, therefore, accelerates cor- rosion; knowing this, the engineer should insist upon thoro 204 EXPERIMENTAL GROUP IV cleaning before painting or repainting a metallic structure. Mill scale (FesCU), on the other hand, is electro-positive to iron, and is, therefore,, an inhibitor; could iron be so treated that the rusting would form the magnetic oxide, the corrosion problem would be solved, as far as ferrous materials are concerned. The corrosion in boilers or hot water systems is ordinarily combated by hanging slabs of zinc in the boiler. The trouble with this practice is: First, the difficulty of getting a good, durable, electrical contact between the metals; second, the protection offered by the zinc is localized; and, third, it is uneconomic to burn zinc to save iron. A better way to control such cor- rosion is to eliminate the oxygen from the feed- and circulating- water (which is the biggest culprit in this case) by boiling it in an open-top container before injecting it into the system. Special Apparatus. The special apparatus required is as follows : Direct-current ammeter, lo-ohm resistance. Millivoltmeter. Four 250-cc. beakers. Two test-tubes, with perforated corks. f-in. round metal rods as follows: One of copper. Five of mild steel, Two of zinc. 1 2 -in. piece of steel wire. Hand towel, o.i-gram trip balance and weights. Supplies. The supplies required are as follows: N Sugar solution, 10 N Sulfuric acid solution, and 25 per cent. Copper sulfate solution, i per cent. Potassium bichromate, 10 per cent. CORROSION 205 N Sodium chloride solutions, and 2 per cent. Freshly prepared warm ferroxyl, liquid. Spool of magnet wire. Solvent alcohol. Distilled water. Metallic fragments. i6-gage steel plate, iX2-in. i6-gage ingot iron, iX2-in. Laboratory Equipment. The laboratory equipment required is as follows : 6-volt battery circuit. Soldering outfit. Quenching bath. Buffing wheel. Ice. Procedure, a. Arrange the apparatus as shown in Fig. 47, page 196, to test the conductivity of a i per cent sugar solution. Does any current flow, as indicated by the ammeter? Replace the ammeter by a millivoltmeter, and test again for the passage of any current. b. Rearrange the apparatus as at first, and test the conduc- tivity of a i per cent sulfuric acid solution. Verify the state- ments of the general explanation, page 196, for this case. c. Arrange the apparatus as shown in Fig. 48, page 197, to show the current generated by the primary cell " copper vs. iron," verifying the statements of the general explanation. d. Cut a i2-in. piece of steel wire in two pieces, anneal one piece, and harden the other. Clean these wires, connect them to the ammeter and dip the free ends into the sulfuric acid solu- tion. Observe conditions for several minutes. e. Polish a mild steel rod on a buffing wheel, wash in alcohol and distilled water. Fill a test-tube with distilled water, and immerse the rod by holding in place with a perforated cork. Set aside and leave it undisturbed, examining it from time to time, 206 EXPERIMENTAL GROUP IV /. Clean another steel rod as in procedure e and dip the end of it in a i per cent copper sulfate solution for five seconds. Remove and wipe off quickly. Repeat the operation, dipping deeper and deeper with subsequent immersions until a visible copper plate is formed on the end, which cannot be wiped off. g. Clean another rod as in procedure e, and immerse for an hour in a 10 per cent solution of potassium bichromate. Remove, wash and dry on a clean cloth. Then repeat procedure/ until a deposit of similar density is acquired. h. Assemble the apparatus as shown in Fig. 50, page 200, to show the electrolytic theory of corrosion of iron. i. Run a little ferroxyl into the bottom of two beakers, and place in ice water to solidify. Cool the balance of the reagent nearly to the solidification point. Clean several iron fragments, such as a needle, a bent case nail, a thin black sheet, etc., as in procedure e, place them on the ferroxyl bed, and cover with the reagent. Cool the whole until it gels, and set away to be exam- ined on the morrow. j. Clean small pieces of tinned, galvanized, and terne plates; dip them into just-fluid ferroxyl to give a thin film on the sur- face of the metal. Set away to be examined on the morrow. k. Fill two beakers with a 2 per cent solution of NaCl. In one Of them place an iron and a zinc electrode, not in electrical connection; and in the other similar electrodes, connected by a conductor. Allow the cells to stand overnight, and examine. /. Clean and weigh two pieces of i6-gage metal, iX2-in., one of mild steel, and the other of pure ingot iron. Immerse these in 25 per cent sulfuric acid for one hour, wash, dry, and re-weigh. Queries, a. Discuss any variation noted from the expected results of procedure b, c, and h. b. What does procedure a indicate as to the ionization of sugar in solution? c. Present the experimental data of procedure d and explain the reason for a steady decrease in current. CORROSION 207 d. What does procedure e disclose as to the solution pressure of iron in pure water? e. Discuss the results of procedure / and g. f. Sketch and explain the results of the ferroxyl tests. Give the chemistry of the ferricyanide indication for ferrous iron. g. Discuss fully the results of procedure k, explaining the chemical reactions in both cases, and the composition of the end- products. //. What would happen if a piece of lead were immersed in a copper sulfate solution? Why? i. Explain why a " gravity cell " is not polarized and how it maintains a nearly constant electromotive force. j. What influence should purity of the iron have upon cor- rosion? Why should cast iron be such a good rust-resisting material as to be used in water pipes? EXPERIMENTAL GROUP V FOREWORD TO THE STUDENT The two following experiments are an introduction into the metallurgy of cast iron. Evidently they could be extended considerably in case a student wished to specialize in this particular branch of metallurgy. Many problems could be investigated bearing upon the composition of molding sand, its proper preparation and manipulation, and the application and operation of the various styles of molding machines. Much of these and other problems of the iron founder are best studied in the foundry itself, however, as molding, core making and melting troubles are there intensified. EXPERIMENT NO. 25 MOLDING Object. The object of this experiment is to make sand molds for a simple casting, or test bar. General Explanation. The general subject of molding is treated at length in Appendix B, page 269, which is a coordina- tion paper written by Prof. Max B. Robinson, when Instructor in Coordination in the Cooperative Engineering Course of the University of Cincinnati. The subject needs no further elabora- tion in this place, as the student will study Appendix B, as well as pages 304 to 314 of Mills, " Materials of Construction," for a general view of the subject. In this experiment, instructions will be given for making molds for test bars, after the methods developed by W. J. Keep, and now in very extensive use thruout the United States. Special Apparatus. The special apparatus used is as follows: Set of molder's tools: Square-pointed shovel. Sprinkling can. Bench rammer. Strike. Fine riddle. Bellows. Slick and spoon. Four iron follow boards, with brass pattern and chills, gate sticks, and pouring basins, complete. Supplies. The supplies needed are as follows: Molding sand, preferably " stove-plate " sand. 209 210 EXPERIMENTAL GROUP V Laboratory Equipment. The laboratory equipment needed is as follows: Holder's bench. Four snap flasks with bottom and top boards. Four weights. Broom. Procedure, a. Ask an instructor to examine the sand, which has been left from the previous day's work in a neat pile near the molding bench. This sand will be somewhat too dry for use. Shovel the old sand into a new pile, sprinkling the old pile with water from time to time as instructed. Re-shovel (or " cut ") the sand at least three times more to reduce it to a uniformly damp condition. Prepare a few shovelsful of unused sand for " facing " in a similar manner. b. Use the iron follow board furnished by the stock clerk. Place it on the molding bench, cleats extending away from the molder. The brass pattern will fit in place and is made to cast from the under side two test bars, J in. by J in. by 12 in long. The pattern also forms the skim gate and runners. Place the chill yokes in their proper place on the follow board. Place tht " drag," or lower half of the flask, upside down over the follow board, pins thru cleats. c. Sift the facing sand thru a fine riddle to cover the whole pattern some 2 in. deep, and tuck in gently with the hands. Shovel the drag full of old sand, ramming it around the edge of the flask with the chisel edge of the rammer, butt end inclining toward the center. Round off the remainder of the sand with the hands, and butt-ram the entire surface vigorously. d. Scrape off the excess sand to a level with the edge of the flask with a " strike " or straight edge, sprinkle on a little loose, dry sand with the hands and rub the bottom board to a firm bearing. e. Grip the entire assemblage of follow board, drag, and bot- tom board, and turn it upside down (" roll over "). Remove the follow board, by lifting vertically, blow off any loose sand with MOLDING 211 a bellows. Ask an instructor to inspect the condition of the surface before proceeding. /. Sprinkle a little fine, dry, " parting " sand evenly over the entire surface, place the cope, or upper half of the flask on the drag, pin thru cleats, and place a gate stick in the proper position. g. Repeat procedures c, d, and e in the cope. Before the cope is rolled over, remove enough sand alongside the gate stick to allow the fire-clay pouring basin to be properly embedded, with- draw the gate stick, and round and smooth all edges, patting down the sand with the fingers. /z. The cope is now upside down on the bench beside the drag. Bevel the gate hole at the parting. Withdraw the pattern by lifting it vertically, without rapping. Sprues, runners and skim gates need not be cut as the pattern is constructed to form them. The chills, of course, are to be left in position, embedded in the sand. Be particularly careful to remove the last bit of loose sand in the mold. Mark the mold by making one or more small conical dents in the sand, so that the finished bars will have dis- tinguishing marks for identification. Finish any imperfections by hand or " slick," replace the cope on the drag, removing the top board. Lift the bottom board, flask, and mold to the floor where it can be poured, and place a flat cast-iron weight on the cope to hold the parting tight shut. i. Each squad-member shall make* a mold individually, calling upon an instructor to inspect and grade his work at the end of procedure h. At the end, the flasks should be " shook-out," that is, the sand returned to the original pile, and the floor cleaned up. Queries, a. Sketch and describe a molding machine. b. Sketch and describe the mechanical arrangement used in " vibrators " to produce the rapid, slight, jarring effect. c. Why is it necessary to peen-ram with a tool inclined toward the center of the flask? d. Describe a method for exact control of the moisture in molding sand. 212 EXPERIMENTAL GROUP V e. Discuss the limitations of molding. For instance, why are columns usually of built structural shapes? What portion of the foundry field has the manufacture of " stampings " appropriated? What methods are employed to cheapen and expedite " quantity production?" /. Give specifications of sand for molding stove-plate; gen- eral machine parts; brass; steel. EXPERIMENT NO. 26 COMPOSITION OF CAST IRON Object. The object of this experiment is to determine the effect of chemical composition upon the physical properties of cast iron. General Explanation. In the experiments of Group IV, the properties of steel were studied with the aid of the modern equi- librium diagram, Fig. 23, page 131. In that place it was stated that steels may be regarded as iron-carbon alloys with a low percentage of carbon. Similarly, cast iron is essentially different only in having a higher percentage of carbon in the alloy. The point of demarcation may be taken as 1.7 per cent carbon, 98.3 per cent iron, which is the maximum solubility of carbon in iron, that is to say, the highest carbon austenite. Suppose, now, that a pure alloy containing 3 per cent carbon, 97 per cent iron, be cooling from the molten state. Reference to Howe's diagram, Fig. 23, page 131, shows that at 1280 C. the first particles of solid appear. This solid is not pure iron, how- ever, for at that temperature iron and carbon form a solid solu- tion having a carbon content of 0.6 per cent. Further cooling would dissipate the latent heat of the solidifying austenite, as well as the sensible heat of the mixture, and the temperature would fall gradually, accompanied by a continuous solidifica- tion of austenite. Examine the state of the melt at 1200 C. According to the equilibrium diagram, the just-solidifying liquid must have a carbon content of 3.75 per cent carbon, while the just-con- gealing solid has a carbon content of 0.95 per cent C, both of composition far removed from those at 1280 C. 3.0 per cent and 0.6 per cent, 'respectively. The first crystal of austenite formed 213 214 EXPERIMENTAL GROUP V at 1280 C. (containing 0.6 per cent C, 99.4 per cent Fe) naturally enriched the remaining mother liquor in carbon by withdrawing a high-iron alloy from the melt. It is easy to see, therefore, that the composition of the mother melt must slide down the liquidus with decreasing temperature until the eutectic point (4.3 per cent C) is reached. The variation- in the carbon content of the progressively solidifying austenite is not usually so well appreciated. Just after the first crystal of 0.6 per cent austenite has formed, it finds itself in an unstable condition at a slightly lower tempera- ture. The 0.6 per cent austenite is no longer the saturated solution, but it must continuously absorb carbon from its sur- roundings, becoming higher and higher in carbon-content to match that of the continually solidifying austenite. In this way (if time be given during a slow cooling), the already solid austenite is absorbing carbon from the mother liquor, and the whole solid, no matter at what time its parts have been born, has a composition which slides down the solidus to the maximum solubility, 1.7 per cent C, arriving there at exactly the same time and temperature as does the mother liquor at the eutectic. At this instant, the mixture contains primary crystals approx- imating 1.7 per cent austenite, and eutectic mother liquor of 4.3 per cent carbon. The temperature will remain constantly at 1130 until the mother liquor has solidified into an euteclic mixture of 1.7 per cent austenite and cementite, cementing the primary crystals together into one solid mass. The radiating heat is entirely supplied at this temperature by the latent heat of solidification of these two components. A further cooling of the entirely solid iron will bring further changes. As the temperature falls, the saturation solubility of iron carbide in iron continually decreases. The primary massive crystals of austenite, and the lamellar eutectic austenite must both progressively eject this excess of cementite. This ejection is continuous until a temperature of 720 C. is reached, when the constitution of the metal is made up of 0.9 per cent, austenite, cementite bordering this austenite (having been precipitated COMPOSITION OF CAST IRON 215 therefrom) and lamellar cementite of the original eutectic, which might be called ' k eutectic cementite." At 720 C. the remaining 0.9 per cent austenite breaks up into pearlite; and no further change should take place, other than a coagulation of the like constituents into favored areas. Such is the constitution of the purest white cast iron, and it is illustrated in Fig. 51 by Wlist. It seems, however, that if more than about 1.25 per cent of carbon is present in the alloy, the cementite acts as a very unstable compound, especially at FIG. 51. White Cast Iron. (Wiist.) temperatures lower than 1100 C. On even moderately slow cooling, the iron carbide breaks up into ferrite and graphite, the latter accumulating in long, greasy flakes, filling in a metallic network with a lubricant, void of tenacity. These flakes give gray iron its characteristic color and fracture, and the other physical properties are strictly dependent upon the discontinuity of the metallic aggregate. Fig. 52, by Boylston, shows the constitution of this iron to be really a low-carbon steel, intersected by graphite flakes. Intermediate stages in the decomposition of cementite into graphite and ferrite are sometimes encountered, and are termed 216 EXPERIMENTAL GROUP V " mottled irons." Fig. 53, by Wiist, shows their constitution so clearly that discussion would be superfluous. FIG. 52. Gray Cast Iron. (Boylston.) FIG. 53. Mottled Iton. (Wiist.) It is easy to see that the physical properties, such as strength, shrinkage, etc.. are primarily dependent upon the amount of COMPOSITION OF CAST IRON 217 carbon present and its state of aggregation. For instance, dis- regarding the total carbon present in cast iron, it may be said that carbon as cementite (" combined carbon ") will produce the following results: Combined Carbon Softest Iron 0.15 per cent or less Iron with maximum tensile strength 0.45 per cent Iron with maximum transverse strength 0.7 per cent Iron with maximum crushing strength i .o per cent or more These matters have been thoroly explained by Howe, and the student is referred to Mills, " Materials of Construction," pages 323 to 332, for their discussion. They should be thoroly understood, as the effect of other alloying metals and metalloids is indirect, chiefly due to the influence they exert upon the state of the carbon. With the exception of sulfur, it is impossible to detect the presence of even large percentages of other alloying impurities in cast iron by microscopic examination. Silicon, for instance, is thought to form iron silicides, which dissolve in the ferrite and somehow produce marked instability in cementite. For this reason it may be said that silicon should be kept within the following limits to produce the indicated results: Per Cent of Silicon For maximum hardness 0.8 per cent or less For maximum crushing strength 0.8 per cent For maximum density and modulus of elasticity i .o per cent For maximum transverse strength i .4 per cent For maximum machinability 2.5 per cent Sulfur is thought to form an iron sulfide, which combines with iron to form a low-melting, tenuous, weak eutectic which surrounds the crystalline grains, and induces great red-shortness. It also counteracts the effect of fifteen times as much silicon, 218 EXPERIMENTAL GROUP V and consequently should be kept as low as possible (less than 0.15 per cent) in good gray foundry irons. Manganese counteracts the action of sulfur by forming manganese sulfide, and should always be present to this extent. The manganese sulfide readily liquates and may burn at the surface of a quiet melt, thus eliminating the sulfur from tjie iron. Excess of manganese above the sulfur requirement forms a double carbide with iron, and thus increases the combined car- bon; more than i per cent manganese affects the mechanical properties of the iron directly by altering the characteristics of the pearlite. Phosphorus probably forms a solid solution with ferrite, increasing fluidity and producing cold-shortness. Its effect is much masked by other impurities, however. If it is kept below 0.5 per cent, it will hardly affect the strength of the material; if thin castings are the prime consideration, upwards of i per cent may be used. Numberless analyses of irons have been given as recom- mended practice for this or that service. Some of them show such small variations between various classes of castings as to be well within the limit of error of melting, sampling and analysis. The strength and adaptability of castings depends to such a great extent upon molding, melting, pouring and annealing practice that it is dangerous, if not foolish, for the engineer to specify analyses within narrow limits. Even if the foundryman luckily furnishes the proper analysis, the castings could easily be of inferior grade due to other causes than the conflicting influence of the half-dozen constituents of the alloy. For infor- mation on this subject the student should first read T. Turner's book on '* Cast Iron " and his article in the Journal of the Iron and Steel Institute, volume I of 1886, or another paper on the " Analysis of Machinery Irons," 15 Metallurgical and Chemical Engineering, 683, and then turn to the Proceedings of the American Foundrymen's Association. Special Apparatus. The special apparatus needed is as follows : COMPOSITION OF CAST IRON 219 Set of molder's tools: Square-pointed shovel. Sprinkling can. Bench rammer. Strike. Fine riddle. Bellows. Slick and spoon. Three iron follow boards, with brass pattern, chills and gate sticks complete. o.i m.g. trip balance and weights. Optical pyrometer. Closed-end observation tube. Shrinkage gage. Cold cut. Graphite electrode for stirring rod. Supplies. The supplies needed are as follows: Three fire-clay pouring basins. Molding sand, preferably " stove-plate " sand. White cast iron, in small pieces of the following anal- ysis: C, 2.5 per cent Si, 0.2 per cent. P, 0.3 per cent. , ,1 as low as possible. Gray charcoal iron, in small pieces, of the following anal ysis: C, 3.5 per cent. Si, i.o per cent. P, o.i per cent. S, as low as possible. Mn, 0.25 per cent. 220 EPXERIMENTAL GROUP V Powdered and analysed alloys as follows: Pyrrhotite. Ferro-silicon. Ferro-manganese . Ferro-phosphorus. Laboratory Equipment. The laboratory equipment needed is as follows: Molder's bench. Three snap flasks, with bottom and top boards. Three weights. Broom. Platform scales. Tilting crucible furnace, with annealed crucible. Thermit welder crucible, clay lined. Cast-iron pig-mold. Emery wheel with wire buffer. Impact machine. Scleroscope. Anvil. Sledge. Procedure, a. Compute the amount of iron necessary to pour three molds such as were made in Experiment No. 25. Figure the constituents necessary to make iron of the composition re- quired by the instructor, weigh and charge into an annealed graphite crucible in a tilting crucible furnace. Start the gas flame, gradually increasing the temperature to a maximum, with a reducing flame. b. Make up three test-bar molds as directed in experiment No. 25, procedure a to h, inclusive. Place one of the molds directly underneath the opening in the bottom of a thermit welder crucible, and the whole so placed that the furnace, on tilting, will pour a stream of metal which will strike the side of the welder crucible, and from this be delivered to the pouring COMPOSITION OF CAST IRON 221 basin of the mold. This arrangement is shown in Fig. 54. Remove the flask. c. When the iron in the crucible appears to be thoroly molten, observe the temperature by means of an optical pyrom- eter observing a closed end tube thrust thru the exhaust hole into the crucible, as instructed in Experiment 15. Obtain inspection and O.K. by the instructor before proceeding. X^':V;M'Flame Exhaust Tilting Crucible Furnace Pyrometer 'Observation Tube Flame Exhaust Hand Wheel and Wo for tilting furnace Floor Level FIG. 54. Arrangement for Pouring Test Bars. d. Without shutting off the gas flame, tip the furnace over carefully, filling the mold with a thin stream of metal. Allow the mold to cool for ten minutes and carefully shove to one side, replacing with a fresh mold. Pour all three molds in this manner with iron of the same temperature. Pour any excess remaining in the crucible into an iron pig-mold. In pouring, be careful to keep a steady stream running continuously, but interrupt it promptly as soon as the pouring gate runs full. 222 EXPERIMENTAL GROUP V e. Shake out the castings and chills. Saw and grind off the sprues, and brush off the sand with wire buffer. Measure the shrinkage by assembling the bar and its corresponding chill on the proper follow board, and slip the wedge-shaped shrinkage gage between the end of the bar and the chill until it comes in contact with each, under no pressure other than its own weight. Read the mark at the upper surface of the bar for the measure of shrinkage. /. Break the bars in a beam-testing machine in the Strength of Materials Laboratory, according to directions. Observe the fracture for size of grain, color and segregation. g. Take three of the bars which show close results in the above test, and test in the impact machine, after procedure /, Experi- ment No. 17. h. Explore the hardness of these three bars with a sclero- scope. i. Observe and sketch the depth of chill by splitting the ends of the bar on the anvil with a cold-cut. j. Metallurgical students should polish, etch, and examine the metal under the microscope, after Experiment No. 22. k. All sprues, gates, bars and fragments should be reserved. NOTE. Each squad should have individual instructions as to the analysis of iron to be tested. In this way the final results can be examined by all students, and will exhibit how the prop- erties of cast iron vary with the composition. A suggested pro- gram of such variations is as follows: Squad i. Melt the white cast iron and test without addi- tions. Squad 2. Melt the gray cast iron and test without additions. Squad 3. Melt the gray cast iron, and stir continuously for an hour with thin iron wire. Squad 4. Add 1.5 per cent silicon as ferro-silicon to the metal from Squad i. These additions in every case should be made within ten minutes of the time of pouring. Drill a hole in the end of a graphite electrode, put in the weighed amount of COMPOSITION OF CAST IRON 223 powdered alloy, and hold in place with a wad of paper. Use this electrode as a stirring rod; the paper will char, and the alloy will be introduced below the surface of the molten metal in this manner. Squad 5. Add 1.5 per cent silicon as ferro-silicon to the metal of Squad 2. Squad 6. Add enough sulfur as pyrrhotite to counteract the silicon in the metal of Squad 4. Squad 7. Add 0.75 per cent sulfur as pyrrhotite to the gray cast iron. Squad 8. Add i per cent phosphorus as ferro-phosphorus to the white cast iron. Squad 9. Add i per cent phosphorus to the gray cast iron. Squad 10. Add i per cent manganese to the white cast iron. Squad n. Add i per cent manganese to the gray cast iron. Squad 12. Add i per cent manganese to the metal from Squad 6. Queries, a. Make up a neat tabulation of the properties of the cast iron investigated, and post it on the proper bulletin board. b. Make a tabulation of the results of all squads, interpreting and discussing the data fully. c. How could the irons be tested for fluidity? d. How can a white iron be changed into a gray iron by a heat treatment? e. Automobile cylinders are to be made of the following com- position: Si : 1.75 per cent S 0.08 per cent P 0.40 to 0.50 per cent Mn 0.60 to 0.80 per cent The foundryman has the three following irons available: 224 EXPERIMENTAL GROUP V Iron I Iron II Iron III Si 2.5 1 . 10 1 . 24 S 0.04 0.07 O-O5 P o . 80 o . 40 o . 50 Mn o . 60 o . 30 o . 50 Calculate the number of pounds of each he must use in order to make the required analysis, allowing 0.02 per cent gain in sulfur from the coke in the cupola, and 0.20 per cent loss in silicon due to oxidation during melting. APPENDIX A ELEMENTARY METALLURGICAL CALCULATIONS Besides noting in a shorthand form the reactions which are taking place, the chemical equation also gives three sets of quan- titative data: First, quantities of weights involved; second, in the case of gases, quantities of volumes involved, and third, quantities of heat involved. These relations will be briefly discussed seriatim. The Weight Relation. The weight relation .follows from the rule that in a balanced chemical equation, the mass of the dif- ferent materials reacting is proportional to their molecular weights. The symbolism is a shorthand form of writing the fact that when two hydrogen molecules, each of two atoms, combine with one oxygen molecule containing two atoms, the reaction produces two molecules of water, each containing two atoms of hydrogen and one of oxygen; furthermore, the combination is attended with the liberation of a considerable amount of heat, whose quantity is represented by the abstract figure 116,120. The weight of an atom of hydrogen is taken as unity; the weight of an oxygen atom has been determined to be 1 6 times as great. We therefore compute that four parts of hydrogen by weight burned in 32 parts of oxygen by weight, will give 36 parts of steam, by weight. The unit of mass is immaterial; it may be the weight of an hydrogen atom, or an ounce, pound, gram, ton, or any other convenient quantity. The point to remember is that whatever the weight of the material reacting, 225 226 APPENDIX A and whatever the units used to express this quantity, these masses will be proportional to the abstract numbers representing the weight of the substances appearing in the balanced equa- tion, i.e., the " reacting " weights, computed on a scale in which the weight of the hydrogen atom is unity. The reacting weights are usually noted in arable numerals below the symbol of the molecule, the above chemical equation completed to show the weight relation being 4 +32=36. Atomic Weights. The following list of approximate atomic weights will, therefore, be useful: Element. Symbol. Atomic Weight. Hydrogen H Boron B Carbon C Nitrogen N Oxygen O Fluorine F Sodium Na Magnesium Mg Aluminum Al Silicon Si Phosphorus P Sulfur S Chlorine Cl Potassium K Calcium Ca Chromium Cr Manganese Mn Iron Fe Copper Cu Zinc Zn Tin Sn Antimony Sb Barium. . Ba Mercury Hg Lead... Pb i ii 12 14 16 iQ 23 24-3 27.1 28.3 3i 32.1 35-S 39-1 40.1 52 54-9 55-8 63.6 65-4 118.7 I2O. 2 137-4 2OO.6 207.2 ELEMENTARY METALLURGICAL CALCULATIONS 227 In addition to the terms atomic weight and reacting weight, the chemist uses the term " molecular " weight. This denotes the abstract number formed by adding together the atomic weights of the atoms constituting the molecule. In this manner the molecular weight of sulfuric acid (H 2 SO 4 ), is found to be 98.1 as follows: H2 = 2 atoms at i = 2 S = i atom at 32.1 =32.1 64 = 4 atoms at 16 =64 H2SO4 = i molecule at 98.1 The number 98.1 merely means that the molecule of sulfuric acid weighs 98.1 times as much as one atom of hydrogen. In other words, the weight of the molecule of sulfuric acid is 98.1 if the weight of an atom of hydrogen is taken as unity. This abstract number representing the molecular weight is used quite often, and the abbreviation " mol " has been proposed for it. The unit of weight is usually prefixed, as follows: one gram- mol of sulfuric acid is 98.1 grams of the pure material. Similarly, one oz.-mol of water would weigh 18 oz. ; three kg.-mols of water, 54kg.; etc. Illustrative Problem. In working problems, numerical or otherwise, the student should cultivate the habit of preceding carefully step by step, rather than trying to arrive at the con- clusion in one grand leap. Step i. Read the problem carefully and understandingly. " What weight of SCb can be formed by the combustion of 100 kg. of pyrite, FeS2? " Apparently the compound splits up, and the sulfur at least, burns in a supply of oxygen to form sulfurous acid anhydride. Disregarding whatever may happen to the iron, which is evidently the intent of the query, proceed to Step 2. Symbolize the chemical reaction The compound splits up: FeS2 Fe+S2heat,~ The sulfur burns : S 2 +O 2 -> S0 2 heat. 228 APPENDIX A This is merely an abbreviated qualitative statement of what happens. This should next be adjusted so that the quantities balance, and the two reactions added together. Step 3. Balance the equations Fe+S 2 heat, Check the work by adding the number of atoms on both sides of the equality signs. Add the equations, cancelling out terms alike on both sides: Step 4. Check the equation by atoms i FeS 2 mol with i atom Fe on the left; i atom Fe on the right. 1 FeS 2 mol with 2 atoms S on the left; 2 mols SO 2 each with i atom S on right. 2 2 mols, total 4 atoms on left; 2 mols SO 2 each with 2 atoms O on right. The equation of the chemical reaction is evidently correctly written. Proceed to Step 5. Re-read the problem, and decide upon the next procedure. " What weight of SO 2 can be formed by the com- bustion of i CXD kg. of FeS 2 ?" Evidently weight is required. Weights vary as the molecular or reacting weights. The amount of heat evolved or absorbed seems not to be required in the solution of the problem, and it will, therefore, be disregarded hereafter. Step 6. Compute the reacting weights One FeS 2 has i Fe atom at 55.8 =55.8 2 S atoms at 32.1 =64.2 i FeS 2 molecule at 120.0 ELEMENTARY METALLURGICAL CALCULATIONS 229 One 62 mol has 2 O atoms at 16 =32.0 2 62 molecules at 32.0= 64.0 One S02 has i S atom at 32.1 =32.1 2 O atoms at 16 =32.0 2 SO2 molecules at 64.1 = 128.2. The complete equation is now written with the weights under- neath the respective reagents. FeS 2 +2O 2 = Fe + 2SO 2 120.0+64.0 = 55.8 + 128.2. Check by addition 184.0 = 184.0 O.K. Step 7. Interpret the results be very careful here. The problem (read it) asks: 100 weights FeS2 produce x weights SO2? The equation says: 1 20 weights FeS? produce 128.2 weights SO2. These two statements can be combined and worked by propor- tion; merely draw a line between the two statements, converting them into two fractions: ICO X 120 128.2 Whence x = 1^128.2 = 106.7 weight units of SO2 (slide rule computation). The weight is the kilogram, Hence 106.7 kg. Ans. Step 8. Examine the result as to whether it appears rea- sonable. In the reaction, one Fe weighing 55.8 is replaced by 2Oi> weighing 64. The resulting gas should weigh slightly more than the original pyrite. 106.7 kg. is slightly greater than 100 kg. The answer is therefore reasonable. This procedure may seem needlessly profuse. It is the exact 230 APPENDIX A fashion in which an expert would subconsciously solve such a problem, and it is insisted that all problems assigned in this course be attacked by the beginner in the same manner. The wording of the thought processes need not be written out at length, but all problems to receive credit must be solved in an orderly manner following the above example condensed into somewhat the following style: Condensed Solution. Problem. Find the weight of metal required for the pro- duction of gram-mol of its oxide. Assumed metal: Al. Assumed oxide: Process: Balanced equation Check 10 atoms = 10 atoms O.K. Reacting weights 4A1 = 4 atoms at 27.1 = 108.4 362 = 6 atoms at 1 6 = 96 In AhOs are A1 2 = 2 atoms at 27.1= 54.2 63 = 3 atoms at 16 = 48 2 mols A^Os at 102.2 = 204.4. Complete equation : 4A1 +362 = 2A1 2 O3 dzheat 108.4+96 =204.4 Check 204.4 =204.4 O.K. The problem asks how much metal to give gram-mol of oxide; i.e., xAl to give 102.2 gm. AkOs. The equation says 108.4 Al gives 204.4 A1 2 O3. Divide the two: X _I02.2 t 108.4 204.4' whence IO2.2 x = 108.4 = 54.2. 204.4 ELEMENTARY METALLURGICAL CALCULATIONS 231 Answer: 54.2 grams of Al will form gm.-mol A1 2 C>3. Is the answer reasonable? The problem asks how much Al2 in A1 2 O,3. 2X27.1=54.2 O.K. Queries, a. Expand the above condensed statement into the form of the illustrative problem given just preceding it. b. How much oxide is formed from 1.7 Ib. of some metal? c. How many pounds of oxygen are required to burn 0.63 ton of a metal? NOTE. Use the condensed solution, step by step, for queries b and c. The Volume Relation. The volume relation expressed in a chemical reaction involving gaseous materials follows from Avogadro's law: " Equal volumes of gas under like temperature and pressure contain the same number of molecules." Another way of stating this law is that the volume of a gas is proportional to the number of molecules present, or more simply, that all gaseous molecules, no matter what their composition . or weight, occupy the same space (temperature and pressure remaining the same). Be careful to note that the molecule of the gaseous ele- ments nearly always contains more than one atom, usually two. We therefore write rather than H 2 +O = H 2 O + 58,060, because the former equation states a reaction between two stable gases whose molecular formulae are H 2 and 2 , respectively. The latter equation states a reaction involving atomic oxygen, which is incapable of continued free existence, and, therefore, represents no stable substance. Furthermore, since the former equation involves molecular formulae of gases, we can immediately compute that two volumes of hydrogen burning in one volume of oxygen will produce two 232 APPENDIX A volumes of steam. This volume of steam produced is predicated on the condition that the heat generated by the combustion (116,120 units) has radiated into cold surroundings, and the product of the reaction has cooled down to the temperature and pressure of the original hydrogen and oxygen. As is the case with weight units, these volume units can be called any con- venient unit, such as the volume of a gaseous molecule, or a cubic foot, or a cubic meter, or the volume of a gram-mol, or any- thing else. The unit volumes are usually expressed in Roman numerals placed above the symbol; the complete chemical equation showing the volume relation being: II I II = 2H 2 O+Il6,I20. Conservation of volume is not existent, as is conservation of weight; therefore it is not possible to check this step by adding the volumes on either side of the equality sign. It will be noticed however, that the volume unit is the same as the coefficient numeral of each formula representing gaseous molecules. Queries. NOTE. These queries are to be worked out and recorded step by step after the method of the condensed solution, page 230. a. How many cubic feet of steam at 100 C., and i atmos- phere will result from the combustion of 31.2 cu.ft. ethane, C2H6, measured at the same temperature and pressure? b. How many cubic meters of oxygen are required to burn 7.35 cu.m. of ethane, if both reagents and products are measured at 80 F. and 27 in. of mercury? c. What is the total volume of all the products of combustion of 2.3 cu.in. of ethane, if all gases are measured at the same temperature and pressure? d. If air is a mixture of gaseous molecules in the ratio of 21 molecules of Q 2 to 79 molecules of N 2 , how much air, measured at standard conditions, will be required to burn 1000 cu.ft. of methane, CHU, measured at o C. and 760 mm. Hg. Heat Evolution. The third set of quantitative data given ELEMENTARY METALLURGICAL CALCULATIONS 233 by the complete chemical equation is the amount of heat evolved or absorbed by the system during the reaction. The reactions most common in ordinary experience are oxidations the burn- ing of coal or gas under a boiler is merely a progressive chemical reaction evolving considerable heat. Other reactions common in metallurgy and chemical technology absorb heat (are " endo- thermic ") and would quickly stop unless that amount of heat were constantly supplied to the reagents from outside sources. As an example of exothermic reactions, take the burning of hydrogen and oxygen in the oxy-hydrogen blowpipe, which produces a very high temperature sufficient to fuse silica and platinum. The quantity of heat evolved by this reaction has been determined by experiment, using the calorimeter (see Chapter 7, Franklin and Macnutt, "General Physics")- It has been found in this manner that four kilograms of hydrogen burning in 32 kg. of oxygen and producing thereby 36 kg. of steam, will produce enough heat to raise the temperature of 116,120 kg. of water i C., if the heat is also abstracted from the hot products of combustion by cooling them down to the original temperature of the gaseous reagents. The quantity of heat, 116,120, is therefore correct for the basis of " from and at zero "; that is to say, the weighed amount of oxygen and hydrogen were at o before the combustion, and the resulting steam has been cooled to that same temperature without condensation. The 116,120 heat units in a free oxy- hydrogen flame are mostly absorbed in heating the resulting steam to high temperature in a calorimeter the sensible heat of the hot steam is abstracted and included in the final result. The metric unit for measuring heat is called the " calory," and, in quantity, is that amount of heat which is required to raise the temperature of i gin. of water i C. Inasmuch as the calory is a rather small unit, and, furthermore, as there are many units of mass in common use, and at least two thermometric scales, it is well to designate the heat unit by terming the calory a " gram-degree-Centigrade," or abbreviated into gm.-C. It will be found very convenient to use larger units of heat, cor- 234 APPENDIX A responding to larger units of mass, such as the kg.-C.; lb.-C.; oz.-C.; oz.-F., or lb.-F., which last, by the way. is the British thermal unit (B.t.u.). In the above calorimeter experiment, the unit of weight kg. was the same thruout; therefore, the equation is correctly written as 2H 2 +O 2 = 2H 2 O+u6,i2okg-C. 4 kg. +32 kg. = 36 kg. As long as the Centigrade thermometer is used, the number of heat units (116,120), will be correct for the reacting weights of the substances involved. The kind of heat unit in that case, depends only upon the unit of mass used in the computation. Thus, 4 Ib. of hydrogen burning in 16 Ib. of oxygen will evolve 116,120 lb.-C. units of heat; 32 grams of oxygen burned in hydrogen will evolve 116,120 gm.-C. units of heat. Conversely, 36 oz. of water will absorb 116,120 oz.-C. units of heat on becom- ing dissociated. The figure 116,120 does not remain true if the Fahrenheit thermometer is used, because the Fahrenheit degree represents only | that absolute difference in temperature which the Cen- tigrade degree denotes. The lb.-F., for instance, which is that amount of heat which raises one pound of water i F., is only f the quantity of heat as that represented by the lb.-C., the latter being that amount of heat which raises i Ib. of water i C. It is clear, therefore, that while the combustion of 4 Ib. of H 2 will raise 116,120 Ib. of water i C., the same quantity of heat will raise 209,020 Ib. of water i F. Thermochemical Data. Since the amount of heat absorbed or evolved by chemical reactions cannot be predicted a priori, it is necessary to furnish a list of calorimetric results. The fol- lowing heats of formation are abstracted from a more complete list in Part I, of Richards, " Metallurgical Calculations," and are arranged in each class according to the atomic weight of the basic element. All heat units are for degrees Centigrade, and on the basis " from and at zero." ELEMENTARY METALLURGICAL CALCULATIONS 235 REACTION Aluminates: 3 Ca+ 2 Al-f 3 O 2 = (CaO) 3 (Al 2 O 3 ) + 789,050 Berates: 4Na+8B+7O 2 = 2Na 2 B 4 O 7 + 1, 496,200 Carbides: Carbonates: 2 Mg+2C+ 3 2 = 2MgCO 3 +539,8oo 4 Na+2C+3O 2 = 2Na 2 CO 3 +S47,40o 2Ca+2C+3O 2 = 2CaCO 3 +547,7oo 2Na+H 2 +2C+3O 2 = 2NaHCO 3 + 454,000 Hydrates: 2A1+ 3 O 2 + 3 H 2 = 2Al(OH) 3 +6o2,6oo N 2 + 5 H 2 +0 2 = 2 NH 4 OH+i77,6oo Ca+O 2 +H 2 =Ca(OH) 2 +2i5,6oo Hydrocarbon gases: C + 2 H 2 = CH 4 +22,25o (methane) 2 C+ 3 H 2 = C 2 H 6 +26,65o (ethane) 2C+2H 2 =C 2 H 4 -u,25o (ethylene) 6C+ 3 H 2 = C 6 H 6 - 7,950 (benzene) 2C+ H 2 = C 2 H 2 - 54, 750 (acetylene) Nitrides: Oxides: 2NH,+2 4 , 4 oo (gas) 2H 2 + O 2 2C + O 2 C + O 2 4Na+ O 2 2Mg+ O 2 2 A1 2 + 3 2 Si -f O 2 4 P +sO 2 + 2 + ^ 2H 2 O +116,120 (gas) 2CO + 58,320 (gas) CO 2 + 97,200 (gas) 2Na 2 O+2oi,8oo 2MgO+ 286,800 S 4K 2Fe 4Fe 3 Fe 4Cu+ SiO 2 +196,000 2P 2 O 6 +730,600 SO 2 + 69,400 (gas) 2X20 +196,400 2CaO +263,000 2FeO +131,400 2Fe 2 3 + 3 9i,2oo Fe 3 O 4 +270,800 2Cu 2 O+ 87,600 REACTING WEIGHTS .2+96=270.5 92X88+224 = 404 40.1 + 24= 64.1 48.6+24+96=168.6 92 +24+96=212 80. 2 + 24+96=200. 2 46 + 2 + 24+96=168 54.2+96+6=156.2 28 +IO+ 3 2= 70 40.I+ 3 2+ 2= 74.1 12+4=16 24+6 = 50 24+4=28 72+6=78 24+2 = 26 28+6=54 + 32 = + 32= + 32 4 24 12 92 -|- 32 48.6+ 32 108.4+ 96 28.3+ 32 124 +160 32.1+ 32 156.4+ 32 80.2+ 3 2 III. 6+ 3 2 223.2+ 96 167.4+ 64 254.4+ 3 2 36 56 44 = 124 = 80.6 = 204.4 = 60.3 = 284 = 64-1 = 188.4 = 112.2 = 143-6 = 319.2 = 231.4 = 286.4 236 APPENDIX A Oxides: 2Cu 2Zn 2Pb Pb REACTION O 2 = 2CuO + 75,400 O 2 =2ZnO +169,600 O 2 = 2PbO +101,600 O 2 = PbO 2 + 63,400 Silicates: 4 Na+ 2 Si+30 2 = 2 (Na 2 O) (SiO 2 ) + 65 2 , 200 2Ca+ 2 Si+3O 2 = 2(CaO)(SiO 2 )+6s8,7oo 6,Ca+ 2 Si+5O 2 =2(CaO) 3 (SiO 2 ) + i ! 2o6,io 2Zn Sulfates: H 2 +S+ 2 2 = Ca+S+2O 2 = Fe+S+2O 2 = Cu+S+ 2 O 2 = Zn+S+2O 2 = Pb+S+2O 2 = = 2 (ZnO)(SiO 2 )+ 549,600 : H 2 SO4+I92,2OO = CaSO 4 +3i7,40o = FeSO 4 + 2 14,500 = CuSO 4 +i8i,7oo =ZnSO 4 + 2 29,600 = PbSO 4 +2i5,7oo Sulfides: H 2 +S = H 2 S+ 4,800 (gas) Fe+S = FeS+ 24,000 2Cu+S = Cu 2 S+ 20,300 Cu+S = CuS+ 10,100 Zn+S = ZnS+43,ooo Pb+S = PbS+ 20,200 REACTING WEIGHTS 127.2+ 32 = 159.2 130.8+ 32 = 162.8 414.4+ 32=446.4 207.2+ 32 = 239.2 92 +56.6+96=244.6 80.2 + 56.6+ 96=232.8 240.6 + 56.6+160 = 457.2 III.6 + 56.6+ 96 = 264.2 130.8 + 56.6+ 96=283.4 2 +32.1+64= 98.1 40.1+32.1+64=136.2 55.8 + 32.1+64=151.9 63 . 6+32 . 1+64= 159 . 7 65.4+32.1+64=161.5 207.2+32.1+64 = 303.3 = 34-1 = 87.9 = 159-3 = 95-7 = 97-5 = 239-3 55-8+32. 127.2 + 32. 63.6+32. 65.4+32. 207.2+32. Thermal Calculation. As an example, take this problem: " Required the number of pounds of carbon to be burned to raise the temperature of 100 Ib. of water from 40 F. to 200 F." Step i . How many heat units are required to heat the water? Assume the heat unit most nearly fitting the conditions of the problem, i.e., lb.-F. Temperature at end 200 F. Temperature at beginning 40 Increase 160 F. Weight of water heated 100 Ib. Heat units required = 100 Ib. X 160 F. = 16,000 lb.-F. ELEMENTARY METALLURGICAL CALCULATIONS 237 Step 2. How much heat will burning carbon produce? From the complete equation, taken from the above lists, (gas) 1 2 + 3 2 = 44 Check this. This equation means that 12 Ib. of carbon burning to CO2 will give off 97,200 lb.-C. units of heat. The combustion of i Ib. of carbon will evidently produce ^^ or 8100 lb.-C. units of heat. 12 Step 3. Read the problem and appraise the results so far obtained. One pound of carbon gives 8,100 Ib.- C. heat units. How much carbon will give 16,000 lb.-F. heat units? Step 4. The units of heat are not alike, and must be made so. i lb.-C = *lb.-F. The combustion of i Ib. of carbon, therefore, gives 1x8100 = 14,580 lb.-F. heat units. Step 5. Solve by proportion. One pound burning carbon gives 14,580 lb.-F. How much carbon to give 16,000 lb.-F.? i 14,580 x 16,000' 16,000 i.io Ib. carbon required. 14,580 Answer. Step 6. Is the answer reasonable? 12 Ib. C. will boil 972 Ib. water from o C. i Ib. C. will boil approx. 80 Ib. water. The problem calls for bringing 100 Ib. water nearly to a boil. The answer should, therefore, be nearly i Ib. O.K. 238 APPENDIX A This result gives the minimum amount of carbon theoretically required, on the basis that the product of combustion (CO 2 ), is cooled to o C. Practically, considerably more would be needed to furnish the heat carried away by the carbon dioxide escaping at a somewhat higher temperature at least as great as 200 F., the temperature of the absorbent in this problem. Queries. NOTE. These are to be solved and recorded step by step. a. How many gm.-C. of heat will be evolved in changing 10 gm. Fe into each of its oxides? b. How many lb.-F. will be evolved in burning 12 Ib. CO to CO 2 ? c. The reaction 2 H 2 +0 2 = 2H 2 + ii6,i2o (gas) is on the basis that the resulting steam is cooled to o C. and remains gaseous at that temperature. What will be the heat evolution of the reaction on the basis that the resulting steam is cooled to o C. and condenses into water, if the condensation of i kg. steam at o C. into water at o C. evolves 606.5 kg.-C. units of latent heat? d. Figure the heat evolution of the same equation, product solid at o C., if the latent heat of fusion of i Ib. of ice is 144 B.t.u. Decomposition of Compounds. All the thermal equivalent? listed above correspond to the formation of the compound from its elementary constituents. The great number of ordinary reactions are between compounds, rather than elements. In order to find the net heat effect of such reactions, it is evidently necessary to compute the input of heat necessary to decompose the reagents into their constituent elements, and deduct this amount from the total heat evolved upon their reassembly. For example, to evaluate the balanced reaction C0+Fe 2 3 = C0 2 + 2FeOheat, first compute the amount of heat necessary to decompose the reagents into their components, as follows ELEMENTARY METALLURGICAL CALCULATIONS 239 Step i. Transpose from the tabulation of thermochemical data the reactions 391,200. These equations state that it requires an input of 58,320 heat units to decompose 2 mols of CO, and 391,200 heat units to decompose 2 mols of Fe 2 O3. Evidently to decompose i mol of CO will require 29,160 and i mol of Fe 2 Oa will require 195,600, or a total of 224,760 heat units to decompose the reagents into their components C, Fe, and O2. Step 2. Now, after this decomposition has been effected, the atoms recombine in a new array, stated in the following equations : = CO 2 +97,2oo, evolving a total of 228,600 heat units. (The original equation involves the production of one mol of CO 2 and two mols of FeO, so the heat figures in the above may be added directly as they stand). Step j. Subtract the input from the output. Input ............................... 224,760 Evolution ............................ 228,600 Balance evolved .................. 3,840 The complete equation should then read. The sign would have to be minus in case the input were greater than the output, in which case, the significance would be that the equation absorbs heat, or is endothermic. Queries, a. Write the complete equation for the combustion of the gaseous hydrocarbons. b. How much heat will be evolved by the combustion of one kg. of these hydrocarbons? 240 APPENDIX A c. How much heat is evolved in the reaction of water on calcium carbide, forming acetylene? d. How much heat is evolved in the " thermit " reaction: Fe 2 O 3 + 2 Al = A1 2 O 3 + 2Feheat. Gases. Unless otherwise expressly stated, gases are measured by volume at standard conditions of temperature (the freezing point of water; o C., or32F.) and pressure (one atmosphere: 760 mm. or 29.9 in. of mercury; or 14.7 Ib. per sq. in.). Should the volume be desired under any other circumstances, the laws of Gay Lussac and of Boyle are applied; viz., that the volume of a gas varies directly with the absolute temperature, and inversely with the absolute pressure. Remember particularly that the absolute temperatures are used in figuring the volume of gases, and that absolute zero is 273 C. For instance, required the volume of 189.9 cu. ft. of gas at 500 C., and 29.5 in. pressure. 773 29.9 Solution. Volume required = 189.9 X X = 545 cu. ft. Note that the above solution can be made by multiplying the original volume by two fractions, one fraction made up of the initial and final temperatures, and the other made up of the initial and final pressures. Trouble with such problems need never be experienced if the temperatures are carefully reduced to the same thermometric scale in degrees above absolute zero, and then are arranged in a fraction in such a manner as to increase the volume in case the final temperature of the gas is higher than the original. Then carefully reduce the pressures to the same notation (inches of mercury, pounds per square inch, or what not) and arrange the initial and final pressure into a " pressure fraction " so as to decrease the result in case the final pressure of the gas is greater than the original. Or vice versa, in either case. Queries, a. If the pressure of i atmosphere is equal to that of 760 mm. of mercury, or to 14.7 Ib, per sq. in., figure the ELEMENTARY METALLURGICAL CALCULATIONS 241 specific gravity of mercury from these relations. Given that water weighs 1000 oz. per cu.ft. b. What column of mercury will balance the pressure rep- resented by i in. of water? c. Figure absolute zero in degrees Fahrenheit, if absolute zero equals 273 C. d. At a certain point in a flue, 89.9 cu. ft. of gas at 500 F. and 740 mm., are passing per minute. What will be the quan- tity at the base of the stack when the gas has cooled 200 and the pressure increased by the pressure of i in. of water? e. If the above gases weigh 0.63 oz. per cu.ft. at the base of the stack, what would they weigh at standard conditions? Volume-weight Relations. In the complete equation II I II 2H 2 +O 2 = 2H 2 0+ 1 16,120, 4 +32 = 36. The volume and weight relations are, in general, independent. For instance, it cannot be said that 4 Ib. of hydrogen unite with i cu. ft. of oxygen to make 36 Ib. of water, because i cu. ft. of oxygen does not weigh 32 Ib. From Avogadro's law, page 231, that all gaseous molecules occupy the same space under like conditions of temperature and pressure no matter what their composition and weight, it should be evident that a quantity of one gas which weighs i mol should have the same volume as another which also weighs i mol. Thus, i gm.-mol of H 2 , weighing 2 gm.; i gm.-mol of 2 , weighing 32 gm.; i gm.-mol of C0 2 , weighing 44 gm. ; etc., all have the same volume, for this good and sufficient reason that they all have the same number of molecules. The volume of a gm.-mol of gas has been called a gram-molecular-volume, it contains about 6.07 Xio 23 molecules, and equals 22.4 liters at standard conditions. Likewise, a kg.-mol-vol. equals 22.4 cu. m., and an ounce-molecular-volume equals 22.4 cu. ft. This idea immediately gives a method of computing the specific gravity or the weight per unit volume of any gas if its 242 APPENDIX A molecular composition be known, or of any mixture of gases if the percentage composition by volume is determined. Illustrative Problem. " Figure the weight of carbon monox- ide (CO), in kilograms per cubic meter, at standard conditions." Step i. Kg.-mol-vol. CO occupies 22.4 cu. m. (By rule above). C=I2 kg.-mol CO weighs 28 kg. Step 2. One kg.-mol- vol. weighs i kg.-mol. 22.4 cu. m. weighs 28 kg. whence i cu. m. weighs 1.25 kg. Answer. Queries, a. Compute a table giving the formula, molecular weight, weight per cubic meter in kilograms, and weight per cubic foot in ounces of each of the following gases: Hydrogen Water vapor Nitrogen Carbon monoxide Oxygen Carbon dioxide. b. A gram-mol-vol. has been carefully determined to be 22.39 liters. What per cent of error is involved in assuming i kg.-mol- vol. to be 22.4 cu. m.? What per cent error is involved in assum- ing i oz.-mol-vol. to be 22.4 cu. ft., if i meter = 39.3 7 in. and i kg. equals 2.205 lb.* * A good way to remember the mass conversion factor is as follows: There are three tons in general use: i short ton = 2000 Ib. i long ton =224olb. X i metric ton = 2 204 lb. = iooo kg. whence i kg. =2.204lb. Note that the last two digits are merely interchanged in the case of long and metric tons. ELEMENTARY METALLURGICAL CALCULATIONS 243 c. Given the fact that air is a mixture of approximately 21 per cent oxygen and 79 per cent nitrogen by volume, what is its weight per cubic foot in pounds? In ounces? What is its weight per cubic meter in kilograms? d. From the above data, compute the percentage composi- tion of air by weight. e. A producer gas has the following composition by volume: CO 24 per cent C0 2 4 H 2 14 CH 4 3 N 2 55 How much does it weigh per cubic meter? /. How many cubic feet would 90 Ib. of this gas occupy if heated to 235 C., and placed under 2 atmospheres pressure? Problems in Combustion. The relation between volume and weight is very important as it is the starting point of a number of calculations necessary for designing boilers, furnaces, flues or stacks. It permits the computation of the amount of air nec- essary to provide for the combustion of any fuel, should the analysis of the latter be known. It also enables one to compute the amount and analysis of the flue gases produced if the amount of excess air is known, cr vice versa. Air for Combustion. Given an anthracite coal of the follow- ing composition, by weight: C 90.0 per cent H 2 2.5 2 2.5 S i.o H 2 O 1.0 Ash 3.0 required the cubic feet of air required for complete combustion per pound of coal, under forced draft, allowing 50 per cent excess air. 244 APPENDIX A Step i. The reactions involved are a. C+O 2 = CO 2 +97,2oo 12+32 =44 b. 2H 2 +O 2 = 2H 2 O+n6,i2o 4+32 =36 c. S+O 2 = SO 2 +69,26o 32.1+32 =64.1 Step 2. Find from these reactions the weight of oxygen needed for the combustion of i Ib. of coal, by the methods of the illustrative problem on page 227. The work is here condensed to a. O 2 for 0.90 Ib. C.; C : O 2 = 12 : 32 =0.90 : x b. O 2 for 0.025 Ib. H 2 ; 2H 2 :O 2 = 4 132= 0.025 :y c. O 2 for o.oi Ib. S; 8:02 = 32.1:32=0.01 : z. Solving these proportions : x = 2.400 Ib. 3> = 0.200 Ib. 2 = 0.010 Ib. Net oxygen required =2.6iolb. Step 3. Some oxygen is already contained in the coal, and a large excess is provided. Net O 2 required 2 . 610 Ib. Less O 2 in coal o . 025 Ib. Actual requirement 2 . 585 Ib. Add 50 per cent excess i . 292 Ib. Total O 2 furnished . 3 . 877 Ib. Step 4. The problem (read it carefully) calls for cubic feet. Oz.-mol-vol. 2 weighs 32 oz. and occupies 22.4 cu. ft. Total O 2 given weighs 3.877X16 oz. and occupies x cu. ft. ELEMENTARY METALLURGICAL CALCULATIONS 245 32 22.4 By proportion . 3.877X16 x 22.4X3-877X16 . Whence x = - = 43 .4 cu. f t . Volume of oxygen furnished = 43. 4 cu. ft. per Ib. of coal. Step 5. The problem (read it) calls for cubic feet of air. Air is 21 per cent 62 by volume. 43-4 Total air furnished = = 206.7 cu - ft- Answer. Step 6. Is the answer reasonable? Ewing, "The Steam Engine" page 449, makes the statement that about 12 Ib. of air is required for the complete combustion of i Ib. of coal. Usually 12 Ib. more have to enter the fire-box if it is operated by chimney draft, while the excess can be cut down by 50 per cent with forced draft. 18X16 18 Ib. air at 1.3 oz./cu.ft.= =22001. ft. approx. The answer is reasonable. Composition of Flue Gases. The percentage composition of the flue gases or their volume at any temperature or pressure may also be calculated in the following very condensed solution for the anthracite coal just above (slide rule computation). COz from carbon =3.300 Ib., equivalent volume = 26.9 cu. ft. H^O from Hz and moisture = 0.235 Ib:, equivalent volume = 4.68 SOa from sulfur =0.02 Ib., equivalent volume = o.n Oa from excess air =1.292 Ib., equivalent volume= 14.45 N 2 from total air = 163.3 Total products at standard conditions = 209.44 cu. ft. The percentage composition by volume is now easily com- puted : 246 APPENDIX A CO2 12.8 per cent H 2 S0 2 2 N 2 100.0 per cent. Check. Queries, a. Compute the volume of air at 20 C. and 700 mm. necessary to burn 100 Ib. of the producer gas mentioned in the query, page 243, allowing 25 per cent excess. b. Compute the analysis of the products of the combustion of this producer gas by volume. c. A long-flame bituminous coal slack containing 7.55 per cent of moisture suitable for pulverizing, is first dried and crushed. The ultimate analysis then shows: Sulfur 3.5 per cent Hydrogen 5.1 Carbon 74.2 Nitrogen 1.2 Oxygen 6.6 Moisture 0.5 Ash 8.9 i oo.o per cent. The coal is then injected into a cement kiln with the exact amount of air for combustion. Compute the volume of air necessary. d. Compute the analysis of the products of the combustion. e. An anthracite coal contains by analysis Carbon 89 per cent Hydrogen . 3 Oxygen i Ash......... 7 100 per cent. ELEMENTARY METALLURGICAL CALCULATIONS 247 It is burned under a boiler producing ashes which weigh 10 per cent of the weight of the coal used. (Assume the ashes to contain all the ash of the coal with enough unburned carbon to make up the balance.) The dry chimney gases analyze CO 2 15.3 per cent N 2 80.4 62 4.3 by volume. How much excess air is entering the fire-box? Total Heat of Combustion. It should be evident that the total heat evolved in the combustion of any quantity of analyzed fuel can easily be determined by figuring the heat resulting from burning its constituents carbon, hydrogen, sulfur, or any other combustible contained in unit quantity of the fuel in the manner previously outlined on page 236. One point should be carefully noted, however, in figuring the total heat of combus- tion from an ultimate analysis of a solid fuel showing oxygen. This oxygen appears to be already combined with carbon and hydrogen in the volatile hydrocarbons of the coal in such a manner that it may be assumed (as a close approximation to the actual facts) that it exists as water (P^O). For this reason, all the hydrogen shown in the analysis is not available for heat production. The " available " hydrogen is, therefore, that amount shown by the analysis less that quantity necessary to satisfy the oxygen also present. The amount of carbon available is also reduced by that unconsumed portion entering the ashes. Illustrative Problem. Figure the heat of combustion of the anthracite coal mentioned on page 243, if the dried ashes amount to 6 per cent of the original coal. Step i. Copy the analysis for reference. C 90.0 per cent H 2 2.5 2 2.5 S i.o H 2 O i.o Ash 3.0 Check it. 248 APPENDIX A Step 2. Figure the available hydrogen and carbon ir. i Ib. of the coal. 62 present = 0.02 5 Ib. From the equation 2H 2 + O 2 = 2H 2 O + 116,120 4 +3 2= 36 it seems that the oxygen will require as much hydrogen to sat- isfy it, weight for weight. Hydrogen combined = | 0.025=0.003 Ib. Hydrogen available = 0.02 5 Ib. present minus 0.003 H>- combined = 0.02 2 Ib. In a like manner, Ashes produced =0.06 Ib. Ash from analysis =0.03 Ib. Carbon in ashes =0.03 Ib. Carbon in analysis =0.90 Ib. Available carbon =0.87 Ib. Step 3. Figure the heat resulting from the combustion of each of the constituents. Be careful to use the available quan- tities. a. 0.022 Ib. H 2 - H 2 at Il6 > 120 lb.-C./lb. = 639 lb.-C. 4 6.0.87 Ib. C ->C0 2 at -lb.-C./lb.== 7,047 lb.-C. c. o.oi Ib. S -> S0 2 at 6gA lb.-C./lb.= 22 lb.-C. 32-1 _ Total heat from i Ib. coal =7,708 lb.-C. Xl gives lb.-F. = 13,900 B.t.u. Answer. Step 4. Is it reasonable? Anthracite coal is nearly pure carbon. Pure carbon gives = 8,100 lb.-C./lb. Therefore the answer is reasonable. ELEMENTARY METALLURGICAL CALCULATIONS 249 Queries, a. Compute the heat of combustion of the pro- ducer gas whose analysis appears on page 243, in B.t.u. per 100 cu. ft., and in kg.-C. per cubic meter. b. Natural gas from the West Virginia fields has the following analysis : C2H 4 0.4 per cent CH 4 93.0 H 2 2.0 CO 0.6 C0 2 0.3 N 2 3-0 H 2 S 0.2 Figure its heat of combustion in B.t.u. per 1000 cu. ft., and in kg.-C. per cubic meter. c. Gas from a by-product coke oven has the following analysis: C 2 H 4 2.4 per cent CH 4 29.2 H 2 50-5 CO 6.3 CO 2 2.2 O 2 0.3 N 2 9.1 Figure its heat of combustion in B.t.u. per 1000 cu. ft., and in kg.-C. per cubic meter. d. Compute the heat of combustion of the bituminous coal whose analysis appears on page 246. Proximate Analysis. The production of an ultimate analysis of any fuel is a slow and tedious proceeding, necessitating the use of rather special and expensive apparatus. Solid fuels, there- fore, are usually analyzed, not for their ultimate constituents, carbon, hydrogen, etc., but merely for moisture, volatile hydro- Carbons, fixed carbon, and ash. (See White, " Technical Gas and Fuel Analysis," pages 193-208.) These latter determina- 250 APPENDIX A tions are very easily made, consisting merely of weighing the sample before and after certain definite heating procedures. The total heat of combustion can be figured within limits from a proximate analysis by the use of Goutal's formula: Calorific power in gm.-C. per gram = 82 C+kV. Where C is the percentage of fixed carbon, V is the percentage of volatile hydrocarbons, and k is a factor figured by interpolation in the following table: V c+v k -5 - 145 0.12 124 0.17 113 - O.26 IO2 o-35 94 0.40 80 Illustrative Example. What is the calorific power of a coal containing Volatile hydrocarbons 10.05 P er cent Fixed carbon .86.7 Moisture 1.8 Ash 1.45 Step i. Figure a value for C+V Fixed carbon 86.7 per cent Volatile hydrocarbons 10.05 Total fuel matter. . . .96.75 Volatile hydrocarbons 10.05 Total fuel matter = 96^75 =aio 4- ELEMENTARY METALLURGICAL CALCULATIONS 251 Step 2. Figure a value for k by interpolation: 0.05 = 145 0.12 =124 A difference of +0.07 makes a difference of 21 whence +0.054 makes a difference of 16.2 in k V Then when = 0.05 and k = 145 C- ~t~ V and the corrections = +0.054 and 16.2 V Then if = 0.104 = 128.8. Step 3. Insert these values in Goutal's formula. Calorific power = 82X86.7 + 128.8X10.05 = 7109 + 1294 = 8403 calories per gram of coal. Answer. Step 4. Is it reasonable? Direct calorimetric experiments (see Hofman, " General Metallurgy," page 114) gave the value 8404 for this coal. This is not as good a check as it appears to be, inasmuch as the bomb calorimeter condenses all the water vapor from the hydrogen reaction into a liquid. This bomb calorimeter value (which is easily obtained and is universally used for coal valuation) must, therefore, be decreased by 606.5 calories for every gram of moisture formed in the bomb, in order to give results comparable with those derived by com- putation from analyses. These latter are on the basis of 2H 2 O+O 2 = 2H 2 0+ 1 16,120 (gas). Goutal's formula may be of prime use if applied to coals from a certain field, whose characteristics are known, and for which a new list of values for k may be derived from experience. This formula and the accompanying values of the constant may give quite divergent results, however, if used indiscriminately for coals from different regions. 252 APPENDIX A Queries, a. Figure the calorific value of one of each of the several types of American coals listed on page 177, Hofman, " General Metallurgy," by using the proximate analysis, and GoutaPs formula. Compute the percentage error in each case. High Temperature Reactions. All the figures given for the heat evolution of the various reactions listed on page 235, are correct only in case the reagents are originally at o C., and the products are cooled to o C. The heat abstracted by the cold calorimeter from the hot products of the reaction is therefore included in the figures representing the heat effect; in other words, the reactions proceed " from and at zero." It should, therefore, be apparent that in case the products of any reaction (such as the chimney gases from a boiler setting) leave the focus at a higher temperature than o C., the sensible heat which they carry away with them must be deducted from the theoretical maximum thermal equivalent. Conversely, any heat brought in by reagents at temperatures above o C. should be added to the heat evolved by the reaction as such, on the basis of " from and at zero." We, therefore, construct the general statement that The heat evolved by any reaction equals the heat of the reaction " from and at zero," plus the heat in the reagents, minus the heat carried away by the products. The use of this rule presupposes a knowledge of the amount of heat required to raise the temperature of various bodies, in other words, the specific heat. Specific Heat. The amount of heat required to raise unit mass of a body thru unit temperature is called the specific heat of the body. Thus, the specific heat of water is i gm.-C. per gram or more simply, one calory. Water requires more heat to raise its temperature than any of the common substances, whose specific heats, therefore, are expressed by fractional ab- stract numbers. For instance, the following short table taken from Watson, " A Text Book of Physics," page 233, gives the specific heat of some common substances at room tem- peratures : ELEMENTARY METALLURGICAL CALCULATIONS 253 Substance. At Specific Heat. Ice -10 0.502 Paraffin wax IO 0.694 Copper eo o 002 Zinc Iron 50 1C 0.093 o. 109 Platinum e,o o 032 Mercury Petroleum .... 20 4O 0.0331 o. 51 The specific heat is not a constant, but its value usually increases materially with higher temperatures; therefore, in general, it takes more heat to raise a body from 999 C. to 1000 C. than it does to raise the same substance from i C. to 2 C. This statement is symbolized by the equation where H is the heat required to raise the temperature of a body from t % to 2-fJ; a is the specific heat at o C.; and b is a numerical coefficient. For the purpose of elementary met- allurgical calculations, the usual problem will be to find the total heat required to raise a body from o C. to t C. The numerical value of most convenience is, consequently, the " mean specific heat," which represents the average value of the specific heat for any range of temperature. Evidently, the mean specific heat is a function of the temperature, as is shown in the following analysis: Specific heat at o C. = a Specific heat at t C. = a+bt Mean specific heat, o to 2a+bt a+-t. 2 This is the average amount of heat required to heat the body i C. The total amount of heat required to heat the body t C. is - to /C. = (+-< 2 Total heat, o to t C. = 254 APPENDIX A that is to say, for the total heat, the mean specific heat must be multiplied by /. The following tables of mean specific heats of gases are abstracted from more extensive lists in Vol. I, Richards, " Metal- lurgical Calculations," and the values are true either for calories per liter, kg.-C. per cu.m., or oz.-C. per cu.ft. GAS MEAN SPECIFIC HEAT Air 0.303+0.000027* Benzene, CeHe o. 76 Carbon dioxide, CO 2 . 0.37 +o. 00022* Carbon monoxide, CO o. 303+0. 00002 7* Ethylene, C 2 H4 . o. 46 +o. 0003* Hydrogen, H 2 o. 303+0. 00002 7* Hydrogen sulfide, H 2 S o. 34 +o. 00015* Nitrogen, N 2 o. 303+0. 00002 7* Methane, CH 4 0.38 +0.00022* Oxygen, O 2 ' o . 303+0 . 000027* Sulfur dioxide, SO 2 0.36 +0.0003* Water vapor, H 2 O 0.34 +0.00015* The following specific heats of solids are for calories per gram; kg.-C. per kg.; or lb.-C. per Ib. SUBSTANCE MEAN SPECIFIC HEAT Alumina, Al 2 Os o. 2081+0. 0000876* Antimony, Sb 0.0486+0.0000084* Carbon, C o. 2142+0.0001662 Copper, solid, Cu o. 094 +o. 0000178* Copper matte, 47 per cent Cu 0.211 o. 0000366* , Copper sulfide, Cu 2 S o. 1126+0.00009* Hematite, Fe 2 Os o. 1456+0. 000188* Iron, Solid, Fe o. n +0.000025* Iron sulfide, FeS o. 1357 Lead, solid, Pb 0.0292+0. 000019* Lime, CaO o. 1715+0.00007* Limestone, CaCOs o. 2086 Mercury, Hg 0.03340.00000275* Silica, SiO 2 o. 1833+0.000077* Slag, copper blast furnace o. 202 +0.0000302* Zinc, solid, Zn o. 0906+0. 000044* As an illustration of the use of this table, the total sensible heat above oC., in i cu. ft. of air at 4oC. is (0.303+0.000027X40)40 = 12.16 oz.-C. ELEMENTARY METALLURGICAL CALCULATIONS 255 The second factor in the specific heat is not of much im- portance at low temperatures, but figure, for instance, the sensi- ble heat in i Ib. of carbon at 1000 C. It is (o. 2 142 +0.000166 X 1000) 1000 = 380.2 lb.-C. Net Heat of Combustion. The practical use of such com- putations may be illustrated as. follows: Suppose the anthracite coal whose composition has been given on page 243, is burned under a boiler producing 6 per cent ashes; the air necessary for complete combustion with 50 per cent excess enters the fire- box at 40 C.; and the products of the combustion leave the boiler setting at 200 C. How much heat would then be available for steam generation? Step i. Obtain the amount of heat of the reactions from and at zero. Heat of combustion, from page 248 = 7708 lb.-C. per Ib. Step 2. Obtain the sensible heat brought in by the air fur- nished for combustion, by virtue of its being at some tempera- ture above o C. The amount of air provided, from page 245, = 206.7 cu. ft . Mean specific heat of air, from table = 0.303+ 0.00002 7/ For the range, o to 40 C. = (0.303+0.000027X40) Total heat in i cu.ft. of air, from o C. to 40 C. = (0.303+0.000027X40) X 40 oz.-C. = 12.16 oz.-C. Total heat in 206.7 cu - ft- air = 206. 7X12. 16 = 2510 oz.-C. Reduce to same unit as Step i =-fiP = 157 lb.-C. Step 3. Obtain the sensible heat taken out by the smoke leaving at 200. Condensing the detailed operations of Step 2, and obtaining the analysis of the flue gases from page 245. CO2..26.Q cu.ft.x(o.37 +0.00022 X2oo)2oo= 2,225 oz.-C. H2O. . .4.68 cu.ft. X (0.34 +0.00015 X2oo)2oo= 346 SO2... .0.11 cu.ft. X (0.36 +0.0003 X2oo)2oo= 9 O2. . ..14.45 CU.ft. X (0.303 +O.OOOO27 X2OO)2OO= 891 N2. . .163.3 cu.ft. X (0.303 +0.00002 7 X 200) 200 = 10,070 Total heat in gases = 13,541 oz.-C. 8 4 61b.C 256 APPENDIX A Step 4. Apply the general statement of high-temperature reactions. The net heat available is that evolved by the equa- tion from and at zero, plus that heat brought in by the hot air, minus that taken away by the flue gases. Net heat = 7708 -f 1 57 846 = 7019 lb.-C. Answer. Regenerative Principle. The advantage of using heated air in a furnace is easily seen, as it counterbalances in part the heat carried up the chimney, which latter can amount to a very large amount of the total heat generated in the fire-box, unless special care is taken to keep the stack temperature low. The temperature of the gases leaving a furnace must be somewhat higher than the temperature required for the operations going on within. In other words, a steel melting furnace must be at all tunes considerably hotter than the liquid metal (which melts at 1530 C. or less) in order that any heat may flow from the heating atmosphere to the melting slag and metal. Flue gases at 1600 C. or more carry an enormous amount of sensi- ble heat, and this would be entirely wasted if passed directly to the chimney. Siemens' regenerative system (see Mills, " Materials of Construction," pages 386-391) utilizes this heat by passing the hot waste gases thru a brick checkerwork, cooling the temperature to 800 C. or less. This heat thus reclaimed and stored in the hot brickwork is returned to the furnace by reversing the gas stream. Cold air or producer gas now go in the opposite direction, arriving at the furnace ports in highly superheated condition. Queries, a. The producer gas whose analysis appears on page 243 is burned in 20 per cent excess air in an open-hearth furnace. The gas and air are both preheated and enter the furnace at 1000 C. The products of combustion leave the furnace at 1650 C. What is the net heat effect of the com- bustion? b. Hot air at 800 C. from the stoves is blown into a blast furnace, meeting incandescent carbon at the tuyeres at 1580 C. The CO formed leaves the focus at 1700 C. What is the net ELEMENTARY METALLURGICAL CALCULATIONS 257 output of the reaction in heat available for melting iron and slag? Maximum Temperature of a Flame, or Calorific Intensity. The problem of computing the maximum temperature attain- able by a reaction is solved by the application of the foregoing principles. Evidently the highest degree (calorific intensity), attainable in combustion will be that point where the entire heat evolution of the equation is absorbed by the products of the reaction. In other words, the highest temperature attainable in an exactly adjusted oxy-hydrogen blowpipe, using cold gas, is when the 116,120 heat units evolved in the combustion is entirely absorbed and carried away by the resulting white-hot steam. That temperature will now be computed. Step i. The reaction is 2H 2 O+O2 =2H 2 0+ 116,120 (gas) 4+32 =36 36 gm. H 2 will absorb 116,120 gm.-C. of heat. The question resolves itself into this: If 36 gm. steam at o C. absorbs 116,120 gm.-C. of heat, what temperature will it then attain? Step 2. The specific heat of gases are given in gm.-C. per liter. How many liters of steam will 36 gm. form, at standard con- ditions? 1 gm.-mol-vol. steam = 22.4 liters = 18 gm. 2 gm.-mol-vol. steam = 44.8 liters = 36 gm. Step 3. Form an equation in / representing the total heat in this steam, from o to some unknown temperature /C. : Mean specific heat of steam, o to / = 0.34 +0.000152. Total heat of steam, o to f = (0.34 +0.0001 $t)t per liter = 44.8(o.34+o.oooi5/)/ for 36 gm. But this total heat is to be equal to the entire product of the reaction, i.e., 44.8(0.34 +0.0001 5/)/ = 116,120. 258 APPENDIX A Step 4. Solve the equation for t. Transposing, o.oo6j2t 2 + i$.2$2t = 116,120. This is a quadratic of the form ax 2 +bx = c whose solution is bVb 2 -\-4ac $ . 2d Substituting, and solving, ^ = 3175 C. Step 5. Is it reasonable? This temperature seem? very high, inasmuch as Burgess, on page 456 of " Measurement of High Temperatures," gives the temperature of the carbon arc as 3500 .150, and 2200 to 2300 C. for the temperature of the oxy-hydrogen flame (page 340). The solution has been checked over, and there appear to be no numerical errors. The trouble may be in the value for the specific heat of water. This physical data has been determined for moderate tempera- ture, as most containers become permeable to gases at temper- ature near or above 1000 C. It is quite possible, therefore, that the formula for mean specific heat should have a third term in t 2 , which would largely increase the amount of heat required to heat the gas at temperatures higher than 1000 C. Again, it may be possible that the reaction between hydrogen and oxygen is not complete at such high temperatures as exist in the oxy-hydrogen flame in other words, perhaps only about 75 per cent of the gas fed to the flame combines. This, of course, would cut down the heat evolved, and, consequently, would largely reduce the calorific intensity. Queries, a. What is the maximum temperature attainable in burning hydrogen in dry air? b. The producer gas whose analysis appears on page 243, is burned in 20 per cent excess air in an open-hearth furnace. Both air and gas enter cold. What is the maximum temperature attainable in the furnace? Would it melt steel? c. In case the producer gas and air of query b were pre- ELEMENTARY METALLURGICAL CALCULATIONS 259 heated to 1000 C. by regenerative checkerwork, what would be the maximum temperature attainable in the furnace? d. Suppose an open-hearth steel plant had available a supply of natural gas of the composition shown on page 249. This gas cannot be preheated because the hydrocarbons would decom- pose and choke up the checkerwork with carbon. It is, there- fore, burned cold, and only the air preheated to 1000 C. What would be the maximum temperature attainable in this case? Cementation Index. Portland cement has been denned by A. P. Mills ( u Materials of Construction," p. 92) as a " finely pulverized product resulting from the calcination to incipient fusion of an intimate artificial mixture of argillaceous and cal- careous materials." The relative proportion of the various constituents lies within narrow limits, for the researches of S. B. and W. B. Newberry show the essential constituents to be the tri-calcium silicate [(CaO)3(SK)2)L an d the di-calcium aluminate [(CaO) 2 (Al 2 3 )]. In the compound (CaO)sSi02, the ratio of base to acid by weight is had from the atomic weights, -as follows: BASE ACID Ca =40.1 Si =28.3 O = 16 O 2 = 32 CaO = 56.1 SiO 2 = 6o.3 (CaO) 3 =i68. 3 Base : Acid = weight (CaO)3 : weight SiO 2 = 168.3 : 6-3> whence Weight CaO = ~ (weight Si0 2 ) = 2.8Xwt. SiO 2 . 00.3 Similarly in the compound (CaO) 2 (Al 2 03), the weight ratio is Base :Acid = weight (CaO) 2 : weight Al 2 0a = 112. 2 : 102.2, whence 112 2 Weight CaO = - - (weight A1 2 3 ) = i.i X wt. A1 2 3 . 260 APPENDIX A The relative amounts of lime, silica and alumina required to make a proper mixture for cement must, therefore, bear the relation CaO = 2.8SiO 2 +i.i A1 2 3 , in order that the two essential constituents may have sufficient material for their formation. The formula merely states that the number of pounds of CaO provided in the " slurry " entering the cement kiln must be equal to 2.8 times the number of pounds of silica plus i.i times the number of pounds of alumina. All the rocks or slags available for the manufacture of cement carry a considerable amount of other oxides. Within limits, the most important of these, magnesium oxide (MgO), is commonly regarded as being able to replace the basic calcium oxide (CaO) , molecule for molecule, while the iron oxide (Fe 2 O 3 ) molecule acts as an acid, and is the equivalent of the alumina molecule (A1 2 O 3 ). Other compounds in the raw materials are purposely kept low, and for the purposes of this computation, will be disregarded. On the basis of replacement, molecule for molecule or more properly, radical for radical one MgO weighing 40.3 hydrogen atoms will replace one CaO weighing 56. i hydrogen atoms. Or in larger units, 40.3 Ib. MgO will replace 56.1 Ib. CaO; that is to say, i Ib. MgO will replace Ib., or 1.4 Ib. CaO. Stated in 40.3 other words, the base MgO is 1.4 times as effective, pound for pound, as CaO, and the magnesia content of the concrete forming materials will replace the usual base (CaO) in that proportion. The formula including this statement then becomes CaO+i-4 MgO = 2.8 SiO 2 + i.i A1 2 O 3 . Following a similar train of reasoning, it is found that if iron oxide will replace aluminum oxide, molecule for molecule, then Fe 2 O 3 : Al 2 O3 = i59.6 : 102.2 = 1.1 10.7. The .relation which must finally exist is expressed by the equation CaO+i.4 MgO = 2.8 Si0 2 -fi.i Al 2 O 3 +o.7 Fe 2 O 3 . ELEMENTARY METALLURGICAL CALCULATIONS 261 Transposing, we have the ordinary statement of Eckel's rule that 2.8 SiO 2 + i.i A1 2 O3+Q.7 Fe 2 Q 3 _ CaO + i.4MgO If a cement be analyzed, and the constituents found be sub- stituted in the above formula, the quotient, or " cementation index," would probably not equal exactly one. A cementation index less than one would mean that an excess of base is present that free or uncombined lime may possibly be present in the clinker. In order to be sure to avoid the harmful effects pop- ularly attributed to " free lime," the cement chemist ordinarily figures his mixture for a cementation index approximating i.i. Illustrative Example. As an illustration of the use of Eckel's rule, suppose there is available for the production of Portland cement of cementation index 1.08, the following materials: IRON BLAST FURNACE SLAG LIMESTONE CaO = 49 . 8 per cent SiO 2 = 3.9 per cent SiO 2 =33.2 A1 2 O 3 = 1.2 A1 2 O 3 = 12.6 Fe 2 O 3 = i . o MgO = 33 CaO =53.5 Fe 2 O 3 = i.i MgO =0.8 Assume 100 Ib. slag and X Ib. limestone as a basis for computa- tion. Substituting directly in the formula 2.8(33. 2 +0.039^0 + 1. i (i 2. 6+0.012^0 +0.7(1. i -j-o.oiX) _ 49.8+0.535^ + 1.4(3.3+0.008*) Multiplying, as indicated, clearing of fractions and collecting, we have 0.46*= 48.8, whence X = io6\b. Consequently, the ingredients should be combined in the ratio of loo slag to 1 06 limestone. Queries, a. Figure the analysis of the above cement. Com- 262 APPENDIX A pare this analysis with the average figures given on page 127 of Mills " Materials of Construction." Discuss the probable effect of discrepancies. b. According to the latest researches of the Geophysical Laboratory, the essential constituents of a cement are tri- calcium silicate and tri-calcium alumina te. Figure a formula similar to Eckel's rule on this basis. c. Compute the relative amounts of slag and limestone by this new rule, assuming a cementation index equal to one. d. Repeat a for this cement. Furnace Charges. An important duty devolving upon the smelter metallurgist is that of computing the furnace charge or " burden." This operation consists in assembling information as to the composition and relative amounts of the various ores avail- able, the composition of the fuel and flux on hand, and in combin- ing these various substances at the charging floor so that they shall issue from the furnace in the form of a metallic material con- taining practically all the values, and of a slag holding the waste substances which cannot be converted into a gas and volatilized. (See Mills, " Materials of Construction," pp. 277, 278.) In the case of a blast furnace, the composition of the slag, matte, or metal tapped from the crucible depends not only upon the com- position of the ore, but also to a large extent upon the tempera- ture and atmosphere existing in the furnace; that is to say, upon the amount and kind of fuel which is burned, the quantity and temperature of the air blown in, and the speed with which the charge passes thru the combustion zone. For these rea- sons, it is not possible to predict the furnace operation a priori, but one must make certain assumptions regarding the expected action of the furnace, borne out by past experience with its operation. Illustrative Example. The method may best be illustrated by a particular case. Suppose it is the duty of a reverberatory furnace in a copper works (see Mills, " Materials of Construc- tion," p. 551) to smelt roasted concentrates or calcine, (v. p. 13) of the following composition: ELEMENTARY METALLURGICAL CALCULATIONS 263 Cu 9.9 per cent SiO2 21.5 FeO 40.5 S 8.9 A1 2 O 3 6.1 CaO 5.0 and flue dust recovered from the smoke pipes and dust chambers containing: Cu 8.1 per cent Si0 2 25.3 FeO 27.5 S 15.8 A1 2 O 3 :... 7-7 CaO 1.3 The relative amounts of these two materials available is deter- mined by the production of the allied smelter departments, and for the purposes of this problem may be expressed as follows: For every 70 Ib. of calcine smelted, 30 Ib. of flue dust must also be charged. A small but unknown amount of limestone, which may be called X pounds, will be added to bring the slag to the proper composition. The composition of the limestone avail- able is: SiO 2 5.6 per cent FeO 0.7 A1 2 3 0.6 CaO 50.3 Past experience with the operation of these furnaces smelting calcine and flue dust will indicate the production of a molten alloy of iron and copper sulfides, called " matte," of approx- imately the following composition: Cu 30 per cent Fe 37 S.. ..26 264 APPENDIX A and that 96 per cent of the copper charged will be recovered in this matte, the remaining 4 per cent being lost in the slag or by dusting. The weight of the matte produced can be imme- diately computed. Copper from calcine = 9.9 per cent of 70 Ib. = 6.93 Ib. Copper from flue dust = 8.1 per cent of 30 Ib. = 2.43 Ib. Total copper charged 9 . 36 Ib. Total copper recovered = 96 per cent of 9.36 Ib. = 8 . 99 Ib. Weight of matte at 30 per cent Cu = = 29 . 95 Ib. The amount of the iron oxide required to furnish the iron for the matte should also be computed as follows: Iron required for matte =37 per cent of 29.95 Ik. = n .08 Ib. Iron oxide to furnish this iron = ' X 1 1 .08 = 14 . 26 Ib. The above procedure for obtaining the weight of the matte i? much superior to the oft-recommended scheme of assuming that all the sulfur charged to a reverberatory furnace will enter the matte primarily as Cu2S, while any excess sulfur will enter in combination with iron as FeS. This is an unsafe assumption, however, for sulfur is such a volatile element that even in the reverberatory, which is ordinarily run with a neutral atmos- phere as a simple melting furnace, a large and uncertain amount of the sulfur is eliminated in the smoke. A balance of the materials may then be constructed as follows: MATERIALS BALANCE Charges To 29.95 Ib. Matte To Slag To Gas 70 Ib. calcine: Cu 9.9 per cent 6 . 93 Ib. Cu SiO2 21.5 iS.oslb. SiOa FeO 40.5 n.oSlb. Fe 14. 09 Ib. FeO S 8.9 6.2 3 lb. S A1 2 O 3 6.1 4.27lb. A1 2 O 2 CaO 5.0 3. 50 Ib. CaO ELEMENTARY METALLURGICAL CALCULATIONS 265 Charges To 29.95 lb. Matte To Slag To Gas 30 lb. flue dust: Cu 8.1 per cent 2 . 43 lb. Cu SiO 2 25.3 7 . 59 lb. SiO 2 FeO 27.5 8.25lb. FeO S 15.8 i.s61b. S 6.37lb. SO 2 A1 2 O 3 7.7 2. 3 ilb. A1 2 O 3 CaO 1.3 o.39lb. CaO X lb. limestone: SiO 5.6 per cent 0.056^ lb. SiO 2 FeO 0.7 o.oo7Xlb. FeO A1 2 O 2 0.6 o . oo6X lb. A1 2 O 3 CaO 50.3 o.503Xlb. CaO We will assume for the purposes of computation that the various slag-forming constituents unite to form a slag which is essentially a ferrous silicate where acid : base = 47. 5 per cent : 52.5 per cent. In this case alumina will be regarded as an acid capable of replacing SiC^, molecule for molecule, while CaO will, in like manner, be figured to its equivalent in FeO. Tabulate the slag constituents as follows: ACID BASE SiO 2 A1 2 3 FeO CaO From calcine. . . I5-05 4.27 14.09 3-50 Flue dust 7-59 2.31 8.25 0-39 Limestone . . . o . 056^ o.oo6Z O.OO7.X' 0.503^ Total ...... 22.64+0.056^ 6.58+0.006^ 22.34+0.007^" 3.89+0.503^ Total acid = Si0 2 +-- A1 2 O 3 Total base = FeO + CaO 102.2 56.1 = 26.52 +0.060 X =27.32+0.651 X. Acid .475 26.52+0.060^ By hypothesis, = =_tl^ = 2 \ - 'Base .525 27.32+0.651^' Whence ^ = 3.4 lb. limestone. To suit the conditions of the problem, the charge would be made up in the proportion of 266 APPENDIX A Calcine 70 Ib. Flue dust 30 Ib. Limestone 3.4 Ib. The mixture is practically self-fluxing without the lime rock. Indeed, the roasting practice (page 13) is so regulated that all the iron not needed to form a matte will be oxidized from the sulfide condition so that it will actually go into the slag as a base, thus replacing the CaO as much as possible. A longer and more perfect roast could have made a matte with less iron and cor- respondingly higher copper, when the slag would absorb the excess iron. However, economic conditions will govern such factors as the time and degree of roasting, grade of matte (that is, its copper content), and the silicate degree of the slag. The student will notice that in copper smelting, the iron contained in the ore is regarded as an impurity, is eliminated in the slag, and wasted; while, in iron smelting, on the contrary, every effort is made to save the iron, recovering it as metal. In general, then, the solution of problems relating to furnace charges may best be effected by constructing a balance sheet showing the origin and expected disposition of each constituent making up the total burden. Economic considerations will fix the relative proportions of the various ores available, and the analysis of the metallic product. To obtain this result, former experience will indicate the required amount of fuel and the ratio of acid to base in the slag Algebraic solution will be relied upon to obtain the unknown amount of flux or to evaluate other vari- ables existing in the problem. Queries, a. A hematite ore of the following composition: H20 9.8 per cent SiO2 10.2 A1 2 3 ... 3-5 Fe 2 O 3 75.5 MnO 1.0 is to be smelted in a blast furnace producing a pig with ELEMENTARY METALLURGICAL CALCULATIONS 267 Fe 94.5 per cent Mn 0.5 c 3.9 Si i.i using i Ib. of coke of the following composition for every pound of pig iron produced : SiO2 5-4 per cent CaO 4.4 H 2 O i.i C 87.9 FeO 1.4 A " neutral slag " (essentially a lime silicate with the ratio acid 47.5 per centA . - = - is required for this particular pig iron, and base 52.5 per cent/ is attained by charging limestone of SiO 5.2 per cent MgO 4-8 CaO 47.4 CO 2 ...42.6 Assuming that all the Fe2Os of the ore will be reduced and fur- nish the iron for the pig iron, while the iron oxide in the coke ash will enter the slag, figure the weight of ore required to pro- duce 100 Ib. of pig iron. b. Construct a balance sheet of the furnace charge, repre- senting the amount of limestone charged by X. c. Compute the pounds of limestone needed for 100 Ib. of pig iron. Regard AkOa as an acid, and equal, molecule for molecule, to SiO2; and compute the other oxides to their CaO equivalent. d. The silicon and manganese of the pig iron may be regarded as being reduced by carbon, forming CO. How much C is used in this manner, and how much enters the pig iron as such? e. Air blown in at the tuyeres burns the balance of the C 268 APPENDIX A in the coke to CO. Assuming dry air at standard conditions, and blowing in 10 per cent excess, how much air must be charged? /. Assume that all of the iron oxide has been reduced by CO; that the CO2 of the limestone and all moisture has been driven off undecomposed; what is the percentage composition, by volume, of the gas at the top of the furnace? How many cubic feet of this gas, measured at standard conditions, would be formed per 100 Ib. of pig iron produced? g. How much heat is absorbed in reducing the elements con- tained in the pig iron? How much heat could be evolved by the total combustion of the carbon? What is the heat efficiency of the furnace on this basis? h. What percentage of the total calorific power of the coke is recovered in the top gas? APPENDIX B FOUNDRY PRACTICE * Introductory. Castings, as used in the building and manu- facture of machines and metallic equipment, are commonly made of iron, steel, aluminum, brass and bronze. Castings of iron, however, are the least expensive in proportion to strength, and their production constitutes the largest part of the foundry business. And since the methods of producing iron castings are quite similar to those for the other materials just mentioned, the differences being largely metallurgical in character, we shall confine our present study to the practice of the average iron foundry. This study is of importance to every engineer, because while he may not be concerned with the direct production of castings, he is sure to encounter their design, purchase, or use, and it is therefore necessary that he understand the basic principles of molding on which their design, cost, or use must depend. The cost of a casting is proportional to three values : a. The amount and nature of the metal and materials used. b. The time and labor required to make the mold. c. The difficulty of securing a casting which is free from defects. The amount of metal used has the greatest effect on the cost in the case of large castings. A careful design will require the * One of the coordination papers in use at the University of Cincinnati, by Max B. Robinson, M.E., Professor of Mechanical Engineering, University of Akron, formerly Instructor of Coordiration at the University of Cincinnati. 269 270 APPENDIX B minimum of metal consistent with the necessary strength, rigidity and mass, according to principles of Strength of Materials and Mechanics. For example, economical specifications will not call for steel or aluminum where cast iron would serve. Neither would they call for an expensive grade of gray iron, when white iron would satisfy the requirements. The time and labor required to make this mold can be greatly increased by unintelligent or careless designs. In fact, designs actually impossible to produce are not uncommon, but the piece could easily be constructed by paying heed to certain unalterable limitations of molding. It should be remembered that very slight changes in design can alter the entire method of molding, and the consequent cost of production. Attention will be called to points wherein the design depends upon methods of molding, and vice versa. The third variable, mentioned above, namely,, the risk of securing an unsound casting, will be discussed under " the Defects of Castings " on page 284. Many of these are caused by faulty design. Molding. In general, a casting is made by pouring molten metal into a mold, the mold being made of refractory material such as sand or loam, its interior shape determining the shape of the poured metal after solidifying. Some molds are made of metal and produce " die-castings," but they will not be consid- ered here, as their use is rather specialized. Most commonly, the cavity in the mold is shaped by " ram- ming up " molding sand inside a flask and around a wood or metal pattern, which becomes embedded in the sand, and which, when removed, leaves a cavity of the required shape. An extremely simple form of mold, for example, would be that for making a rectangular block, 4X8 in., and 2 in. thick, using a one- piece pattern. This would be molded entirely within the drag, or lower half of the flask containing the mold, since it has a flat top and straight sides and can, therefore, be easily removed. A pattern such as the above is called a " flat-back," and the order of operations in molding it would be as follows: FOUNDRY PRACTICE 271 1. Place the mold board on the bench, the cleats extending away from the molder to prevent tipping when turning over. 2. Place the pattern on the mold board, top side down. 3. Place the drag upside down over the mold board and pat- tern, with the pins extending downwards on either side of the board. 4. Sift some facing sand over the pattern until it is covered, using a fine riddle. FIG. 55.- Section thru a Mold Containing a Rectangular Pattern. A = bottom board. B = drag. C = cope. D = parting, or joint. E = pattern. F = sprue. G=gate. H = pouring basin, or skim gate. 5. Tuck the riddled sand around the edge of the pattern with the fingers. 6. Shovel the drag full of old sand. 7. Ram around the inside edge of the flask with the peen or sharp edge of the rammer, butt end inclining toward the center of the flask. 8. Butt-ram entire surface, using more sand if necessary to fill the flask. 9. Scrape off any surplus sand with a " strike." 272 APPENDIX B 10. Throw a little loose sand on the flat surface. 11. Place the bottom board on the drag, and rub it to a firm bearing. 12. " Roll over " the drag, bringing the pattern up. 13. Remove the match plate or mold board. 14. Brush the surface with a soft brush, or use a bellows. 15. Sprinkle parting sand over the sand to separate the cope and drag portions of the mold. 1 6. Blow the excess of parting sand from the surface of the pattern. (In case a two-piece pattern were used, the upper half would here be placed on the lower half.) 17. Place the cope on the drag, with pins fitting in to ears on the other half. 18. Place a " gate stick " in position, extending a slight distance into the drag. 19. Repeat operations 4 and 9 in the cope. 20. Prick vent holes where needed. 21. Remove the gate stick from the sand. 22. Lift the cope from the drag, and place it at one side, bottom up. 23. Smooth off any rough places appearing on the sand sur- faces with a slick or other hand tools. 24. Bevel the gate hole at the joint, and ream the top into a bell-shape for pouring, (In some cases a " skim gate " would be advisable. H, Fig. 55.) 25. Withdraw the pattern, using a draw nail and rapping iron. 26. Cut a sprue from the gate hole to the mold in the top surface of the drag, using a sprue cutter. 27. Finish by hand any imperfections on the surface of the mold. 28. Replace the cope on the drag. 29. Place the mold in position to be poured. 30. Unlock and remove the flask. 31. Place a flat cast-iron weight on the cope to hold the sand down and to prevent a " run-out." The skill required of the molder in performing the above oper- FOUNDRY PRACTICE 273 ations lies mainly in ramming, in hand finishing and retouching, and in handling the finished parts. The necessary skill increases in proportion to the complexity of the pattern. Correct ram- ming is largely a matter of experience, the hardness with which the sand is rammed depending on the size of the mold, the size, " temper " and composition of the sand, and the weight of the casting. Ramming too hard will cause blowholes, since it re- duces the porosity of the sand and gas cannot escape from the mold. Ramming too loose will allow the sand to sink or bulge out under the pressure of the iron, and a swelled casting is the result; or, sand may be washed from* the face of the mold, forming scales or sandholes on the finished work. The bottom of the mold and the joint must sometimes be rammed harder than the other parts because the former must stand the weight of the metal, and the latter is exposed to much handling. Beyond sufficient hardness to maintain proper shape, the risk of losing the casting increases with the hardness of the ramming. The ramming tool should never strike nearer than i in. to the pattern, else a hard spot in the sand will be formed which may cause a " scab " on the casting. Retouching is largely a matter of the deft use of hand tools. All slicking should be lightly done, else scabs will be caused. Care must also be taken not to get the sand too wet. The cope should always be finished before the drag, because should it be spoiled in any way, the drag still contains its pattern and a new cope can at once be made. In commercial work, but few castings are as simple to mold as the rectangular pattern illustrated above, irregularities in shape greatly complicating the molding operations. In order that the pattern may be withdrawn from the sand without break- ing away any of the surface of the parting, most patterns not having a flat top surface, or not having every horizontal section as great or greater in width than every horizontal section below it, must be molded partly in the cope, instead of wholly in the drag. A pattern having a horizontal section less in width than that of sections both above and below it, as a sheave pulley, 274 APPENDIX B Fig. 56, requires a three-part flask, or else a two-part flask and a false cheek. Pulleys with more than one sheave require as many cheeks as there are sheaves. Molding is also complicated by setting of " gaggers " and " soldiers " to support overhanging bodies of sand, by the fre^ quent necessity for drying the mold or skin-drying the surfaces, by the use of blackings and special facings, by special forms ol pouring gates, by the placing of shrinkheads and risers, and b)J the placing of cores. Oftentimes patterns must be placed with a certain surface on the bottom in order to receive the purest iron (impurities rise to the top), but if this is not required, the pattern is placed in whatever position will allow the mold to be made by FIG. 56. Sheave Pulley Pattern in Three-part Flask. the simplest method. This requires good judgment on the part of the molder. To illustrate some of the variations of molding from the order of operations on page 271 consider the molding of a cylinder 6 in. in diameter and 12 in. long. This cylinder can be molded in several ways, each of which illustrates a molding method in common use, since most shapes are modifications or combinations of cylinders and rectangles. Where several methods are possible such considerations as the weight of the casting, the equipment at hand, the finish required, and the number of one kind required will determine the best method to use for an> given pattern. Probably the simplest way of molding the cylinder on its side is by the " split pattern " method, in which the cylindrical pattern is split into upper and lower halves, pins extending down FOUNDRY PRACTICE 275 from the upper half engaging holes in the lower half. The half of the pattern which goes in the drag is placed flat on the mold board, and the drag rammed up as in operations i to 13, page 271. After turning over the drag, the upper half of the pattern is placed on the lower, held in position by the pins, and the mold is finished as in operations 1 7 to 3 1 . Another common method of molding this cylinder is the " bedding in " method. The drag is placed on the mold board in an upright position, rammed full, struck off even with its top edge, and the one-piece pattern is then bedded in to its parting line, Fig. 57, by scooping out a sufficient space, tucking the sand firmly around the sides and ends of the pattern, and patching FIG. 57. Pattern that is " Bedded In." the broken parting by hand. This method is crude, but it saves turning the drag, and is much used in England where heavy iron flasks are the rule. Another common method of bringing the parting line of the pattern to a level with the joint between the cope and drag, is to place wedges between the mold board and the edges of the drag, holding it up to a height equal to that which the pattern is to project into the cope. The pattern is then placed in posi- tion on the mold board, Fig. 58, and the drag rammed up. The drag is turned over, the excess sand on the top struck off and the parting made. The disadvantage of this method lies mainly in the extra work required in making a good parting. The method of " coping down " is also frequently used. In this, the drag is rammed up with the pattern in place exactly as 276 APPENDIX B in operations i to 14, page 271. To permit the pattern to be withdrawn, the sand is then cut away from around the pattern down to its exact center, its parting line, making as gradual a slope as possible, as shown in Fig. 59. The surface is slicked, FIG. 58." Wedging Up." FIG. 59." Coping Down." A, section thru drag, pattern in position. B, side of cope. parting sand dusted on, and the cope rammed as usual. This method leaves a body of sand hanging down from the cope, hence its name. Molding with a Match. The fifth method of molding the cylinder employs the use of a " match," and is commonly used FOUNDRY PRACTICE 277 for patterns with an irregular top surface, which cannot be split, and yet must lie partly in the drag and partly in the cope. A match is a mold used as a mold board into which the pattern fits up to its parting line. The operations of ramming the drag and cope are identical with those already listed. Matches are made of green sand, oil sand, plaster of Paris, or metal. A green sand match is used when only a few castings are needed. When desirable to keep the match for a period of time, an oil- sand match is made, and for permanent use, it should be made of plaster of Paris. A match is made in a shallow frame the size of the flask to be used. If but one casting is required, a tem- porary match or " upset " is made by filling the frame with loose sand and then " bedding in " the pattern the same amount that it is later to project into the cope. The sand should be tucked and rammed around the pattern merely enough to hold the latter in place. The drag is then rammed on top of the upset, and the cope on top of the drag, as in preceding cases. When a green- or oil-sand match is wanted for subsequent molds, an upset is made first and the drag rammed on the upset. The drag is rolled over and the sand knocked out of the upset, which is placed bottom side up on the drag, the pat- tern still being in the drag and projecting up. The upset is then rammed the same as a cope, enabling the sand to be packed from the side which is to form its bottom, making a smooth parting between the upset and the drag. When so rammed, the upset becomes either a green-sand or an oil-sand match depending on whether water or linseed oil has been mixed with the sand as a binder. Oil matches are hardened by baking in core ovens. After ramming the match, the drag is ready to be used for ram- ming the cope. Molding with Sweeps. Molds for circular boiler fronts, tank heads, pulley and flywheel rims, large cylinders and other purely circular shapes can be made without a pattern by the aid of a " sweep." This is often of economic importance when but one or a few castings of a kind are wanted. The " sweep " consists of a board having its edge or end 278 APPENDIX B profiled to correspond with the desired contour of the surface of the casting, and is attached to a " sweep finger " which rotates about a vertical spindle firmly set in a block in the floor. The mold is shaped by turning the board around the vertical spindle as a center, its edge or end imparting the desired contour to the surfaces of the material of which the mold is made. Castings that are irregular in shape can be partially made in this manner, and the balance of the mold made with smaller " pattern pieces." Many molds are swept in loam instead of in sand, the loam being swept against a roughly constructed brick backing which takes the place of a flask. The mold is then baked, causing the loam to adhere firmly to the brick. Molding Sand. Molding sand must possess qualities which will enable it to be tempered and to retain a definite shape in the mold. It must resist fusion, must have permeability to permit the ready escape of gases, and must be capable of being retem- pered and used a number of times. These qualities depend upon its chemical composition and physical constitution. Sands from different localities possess these qualities to different degrees. Also different classes of work require different sands. The coarseness and the constituency required for any given work depend upon the size and weight of the castings, upon the tern- 1 perature of the molten metals, upon the length of time it will take the metal to cool, upon the degree of finish desired, etc. Sand as it comes from the pit must be " burned " to destroy vegetable and animal life. The color of sand, which is reddish-yellow before use, changes to black after use. (When sand becomes worn out, that is, loses its bond or cohesion thru repeated use, it should be replaced.) Sand for steel castings must be more refractory than sand for cast iron, due to the higher temper- atures involved. Sand for brass work need be less so. Facings. Oftentimes sand cannot be used for the surface of a mold without further treatment, due to its tendency to fuse with the iron under prolonged heat, and also due to its inability to form as smooth or finished a surface as desired on the casting FOUNDRY PRACTICE 279 of the mold. A facing material then becomes necessary. It is either mixed with the sand that is used next to the pattern, or painted or dusted on the surface of the mold, or both. Different facings are are used for different purposes. Common lacings are sea-coal, which is a special grade of bituminous coal, ground, and screened to the required fineness; plumbago, which is a form of graphite; coke and charcoal " blacking " compounds; talc or soapstone; and gas-house carbon. Certain mixtures and cer- tain amounts of the above are used, depending upon the temper- atures dealt with, the kind of sand used, the duration of the intense heat, whether the mold is skin-dried, dry-sand, or green sand, and the finish required on the surface of the casting. Cores. Cavities in castings are formed by cores, usually made at the bench with a special grade of sand called " core sand." This is mixed with a " binder," and is molded to shape in wood or metal core boxes, the interiors of which are shaped so as to produce the required forms. Cores are some- times built upon special forms or drums. The cores are then baked in " core-ovens " to render them hard and to fix their shape. Core sand has different qualities from molding sand, and selection of the right sand is important. The binders commonly used are wheat flour, rye meal, powdered rosin, linseed oil, and glue, prepared in various ways, and used in various proportions. Cores must be vented to allow the ready escape of gases. Wax tapers, which melt and leave vent holes when the core is baked, are often inserted when molding the core. These are used especially in thin cores, and where it would otherwise be difficult to lead a vent around a corner, as, for example, in locomotive cylinder ports. The location of cores in the mold is usually determined by the " core prints " on the pattern, where they are usually painted red or black. When set in molds, cores must be supported on the bottom, sides and top, either by core prints or by chaplets. Chaplets come in various forms to fit against flat, convex, or. concave surfaces and for both light and heavy work. Some types come in assorted sizes, and some types are adjustable to 280 APPENDIX B suit any required height. Most chaplets are tinned to fuse readily into the molten iron. Different Classes of Molds. Green-sand Molds. These are molds made from moist and tempered sand and not dried by artificial means. Most small castings are molded in green sand. Dry-sand Molds. Where a very strong mold is required, either to resist erosion by the iron when pouring, or to maintain its shape against a high fluid-pressure, the sand is mixed with a binding material and the finished mold is baked in a mold oven. This process is similar to that used in baking cores. Examples of work cast in these molds are : steam and gas-engine cylinders, air compressor and hydraulic cylinders, printing-press cylinders and rolls, anvil blocks, engine beds and similar heavy castings. Skin-dried Molds. These molds are dry on the surface only, being a compromise between green-sand molds and dry-sand molds. They are made of green sand without a binder, and the drying is effected by gas or oil flames either placed in posi- tion or played on the different surfaces by workmen. Most medium and heavy molds which do not require a complete dry- ing are skin dried, the process materially strengthening the sand walls. Bench Molds. These are small molds whose size makes it most convenient to mold them on a bench. Floor Molds. The majority of molds are made in flasks which rest on the floor. In large floor work, the copes contain ribs and cross ribs, or " chucks," for the purpose of supporting the large areas of sand contained in them. These ribs are cut away where it is necessary to allow patterns to project into the cope. Care must be taken in placing the pattern to see that it does not come too near such ribs. In floor work, where it is often possible to mold several pieces in the same flask, nicety of judgment is required in determining which pieces can go together in this manner. Pit Molds. Often a pit dug in the floor serves the purpose of a drag, the cheek and cope being the only parts of the flask re- quired. This saves expense, especially in extra large work. If FOUNDRY PRACTICE 281 the earth is too damp, a permanently dry pit may be constructed by sinking an iron tank in the ground, or the inside of the pit may be cemented or lined with tar paper. Machine-made Molds. Where many small molds are required from the same pattern, or where there is difficulty in ramming the same or in removing the pattern by hand, molding machines are often employed. See under the heading " Molding Machines." Molding Equipment. Flasks. Flasks are made of iron, steel or wood. Iron flasks are used for very heavy castings. These are made so as to be interchangeable, with pin-holes in their flanges bored with a templet. Iron flasks should be ribbed to prevent springing from the pressure of the iron in the mold. There should be ample room for sand between the pat- tern and the side of the mold not only to avoid heating, but to absorb any gases which are evolved. Trunnions for turning with a crane should be made of steel cast into sides of the flask, rather than of iron cast in one piece with the flask. Steel flasks are light and serviceable, but on account of their lightness, heat up rapidly and are apt to warp out of shape. Wooden flasks should be of substantial, thick material, as they are subject to burning at the joints, which will later cause run-outs. Molding Machines. Molding machines are in general use on light and medium work where a number of molds are to be made from the same pattern. The common machines used are the Hand Squeezer, The Power Squeezer, the Split-pattern Machine, the Roll-over Machine and the Jarring Machine. Where the ramming time is the largest factor in the making of the mold, and where the mold is thin, one of the squeezer machines is used. But as a squeezer machine packs the sand by pressure on its surface, and the top is packed harder than the bottom it is not as well adapted to deep molds as the jarring machine. A jarring machine in which the .impact of the mold on the table is absorbed by air cushions and springs within the machine itself, instead of being transmitted to the surrounding floors and walls, is called " Shockless Jarring Machine." 282 APPENDIX B Where a mold is such that the pattern is difficult to draw by hand without injury to the mold, and the consequent finishing time is the largest factor in the making of the mold, either a Split Pattern or Rollover Machine is used, depending on the shape of the pattern. For description of each type of machine see " Foundry Practice," by Palmer. General Foundry Equipment. Besides the molding opera- tions, the elements of foundry work involve : 1. The handling of raw and finished materials. 2. The melting and pouring of the metals. 3. The cleaning of the castings after they have cooled. The equipment necessary for the foregoing includes: 1. Suitable cranes, hoists and rigging for handling the raw materials, molds, ladles and the finished castings. 2. A cupola for melting the iron, with equipment for fur- nishing a suitable air-blast, and ladles for distributing iron to all parts of the floor. 3. Rattlers or tumbling barrels, pickling vats, and tools for cleaning the castings by hand. Important points to above equipment are as follows: Cranes. The old-fashioned jib-cranes, operated by a hand winch, limited in application, and useless beyond the radius of the crane arm, have been displaced by the traveling crane of either the hand or electric type. The important requirement for a satisfactory crane, aside from its ability to safely carry the weight, is a delicate and sensitive control. The handling of green-sand molds requires gradual stopping and starting, with no shocks, and the safe carrying and pouring of molten metal demands the ability to operate at very slow speeds, on either a light or heavy load. Rigging. The rigging required depends upon the class of work being done, and generally consists of chains, slings, yokes, etc. The Cupola. For melting iron, one of two types of furnaces may be used, either the cupola, or the reverberatory or air furnace. The more common cupola is an upright cylindrical FOUNDRY PRACTICE 283 shaft furnace, open at the top and bottom, and lined with fire brick. It is provided with a charging-door at about the middle of its height. There are " tuyeres " near the bottom, thru which air is blown which consumes the fuel charged and melts the iron. The shell is formed of separate rings of boiler-plate, with angles riveted at intervals to the interior to support the fire-brick lining. The cupola shell is carried on a cast-iron bed-plate ring sup- ported by cast-iron legs. The opening in this ring is tightly closed during operation by hinged doors which are allowed to open at the end of the heat. Molten iron is removed just above the bottom thru the " tap-hole." Slag is removed thru the " slag-hole " on the opposite side, slightly higher than the tap-hole. Cleaning holes or doors on each side of the wind box permit removal of slag or dirt therefrom. Peep-holes opposite each tuyere allow the melter to judge the temperature of the focus. The space in the cupola constitutes zones, as follows: 1. Crucible zone, or hearth, extending from the bottom to the lower row of tuyeres. 2. Tuyere or combustion zone, next above, where the blast comes in contact with and burns the red-hot coke. Its upper limit depends on the blast pressure, but should be between 15 in. and 24 in. above the tuyeres. 3. Melting zone, next above, which is about 7 in. high. 4. The "stack" extending from the melting zone to the charging door. Its function is to contain the material, allowing it to absorb heat before it reaches the melting zone. Before starting the fire for each day's melt, the lining of the cupola should be inspected and repaired. The bottom doors are then closed, and tempered sand is rammed on the bottom, forming a floor inclining toward the tap-hole. After the first charge of coke is ignited, the " breast " is built around the tap- hole. This is often done by removing the coke that has fallen into the breast opening, building a wall of the cold coke back of the opening, and then building up a wall of fire-clay, thru 284 APPENDIX B which a pin is inserted. This pin, when removed, leaves the tap-hole. When the wall of coke is ignited, the breast is baked, and the furnace is ready for the charge of iron and the blast. The blast pressure should not exceed i Ib. per sq. in. Coke is most commonly used as fuel, altho anthracite is occasionally used. Ladles. Ladles vary in size from the small hand ladles car- ried by one man to the large bull ladles having a capacity of many tons. Small ladles are usually lined with fire-clay and the larger ones with fire-brick. Large ladles must be provided with a mechanism for tilting them for pouring. All ladles should have their lining heated by a gas flame or coke fire before receiving the molten iron, to prevent both the cracking of the lining and the chilling and spattering of the molten iron. Rattling or Tumbling Barrels. These should be of different sizes to suit different kinds of work, with a mechanism for rotat- ing. The cleaning is done by the impact of " stars " or " picks " with which the barrel is filled. Sometimes there is used an equipment for directing a blast of sand against the castings being tumbled. Pickling Vats. It is sometimes necessary to pickle castings in dilute sulfuric, hydrochloric or hydrofluoric acid in order to remove green-sand scale. Vats and troughs of wood or stone filled with acid are then used. Castings so treated should always be well rinsed in clean water. Cleaning Tools. Small tools operated by hand or compressed air are necessary for cleaning the scale from the outside of cast- ings and for breaking out the core sand from the interiors. Common Defects of Iron Castings. The most common faults of iron castings are blowholes, sponginess, shrink-holes, scabbiness, sand-holes, floating cores, cold shuts, cold shot, strains, shrinkage strains and warp. These may occur even though the mold is in good condition. i. Blowholes are probably the most common defect. These are holes in the castings caused by imprisoned gases. They are often invisible on surface inspection and are most apt to occur FOUNDRY PRACTICE 285 near the top of a casting. Other conditions permitting, those parts of the casting which are to be machined or which require the most strength, should, therefore, be at the bottom of the mold when it is poured. The gases which form blowholes con- sist of: a. Steam generated from moisture in the sand. b. Gases arising from decomposition of the facing material. c. Air entrapped in various cavities of the mold. The remedy for all these is to provide a means for the escape of gases by venting and by risers. Since vent holes cannot extend to every point near the surface of the molds, the gases must pass thru the sand until they reach the nearest vent hole. There- fore, the sand must have porosity, which it cannot have if too tightly rammed or too wet, or if the grains are not of the correct size and shape for the work in hand. 2. Sponginess is caused by the formation of gas bubbles in the iron at the instant of solidification. It is due to an improper mixture of iron charged to the cupola. If a casting is thick at one place and thin at another, sponginess can be prevented by providing a riser or shrink-head at the thick part. To avoid sponginess where it is not possible to place a riser, let no part of the iron be more than 3 in. from a sand surface. 3. Shrink-holes are caused by the shrinking of the metal in cooling. It occurs where a casting is extra thick. Shrink- holes are prevented by avoiding sudden changes in the thickness of a section, by the use of a shrink-head, or by chilling the thick portions. The designer should avoid heavy sections where possible, but where such are necessary, they should be so ar- ranged as to permit the use of a riser directly over them. 4. Scabbiness is caused by loosening the surface sand, or erosion of the fillers and partitions by the force of the inflowing iron. The remedy is to avoid sharp edges on fillets and tongues of sand, and to gate the mold so that the current of entering iron will be evenly spread. 5. Sand-holes are associated with scabbiness. They are caused by the same conditions, except that in their formation 286 APPENDIX B the iron has solidified at the .surf ace, thus preventing the loosened sand from floating entirely to the top. This leaves it imprisoned in the body of the casting. 6. Floating Cores are caused by a core of insufficient strength to resist the buoyant effect of the molten iron. 7. Cold Shuts are caused by the imperfect union of two or more streams of molten iron in the mold, too cold to coalesce on meeting. They occur when the iron must flow some distance thru a thin part of the mold. The remedy is to use iron as hot and fluid as possible, and to arrange gates so that the iron will quickly fill up the mold. 8. Cold Shot are small globules of iron, imperfectly united with the rest of the casting, and hard or impossible to machine. They are caused by the splashing of the iron in pouring, the spattered drops becoming chilled and dropping to the bottom of the mold. Splashing is usually caused by a gate dropping iron directly into a web, or by pouring so slowly as to allow the iron to " dribble in." The remedy is obvious. 9. Strains. This term covers a casting with warped walls due to the outward yielding of the sand walls of the mold. This is caused by loose ramming and insufficient weights. 10. Shrinkage or Internal Strains are caused by unequal rates of cooling. To avoid them, arrange the thickness of the parts so that the entire casting will solidify at about the same time, or else chill the thick portions. 11. Warping is caused by shrinkage strains that are suf- ficient to alter the shape of the finished casting. It is due either to a want of symmetry in the design of sectional parts, and the consequent unequal contraction in cooling, or to dumping out the casting from the sand while still hot and soft, allowing them to become chilled on one side by air currents, or to sag of their own weight. Reference on Foundry Work. " Elementary Foundry Prac- tice," Richards; " Foundry Practice," Tate and Stone; " Foundry Practice," Palmer; " Metallurgy of Iron and Steel," Bradley Stoughton. GLOSSARY OF TERMS IN COMMON USE ARBOR. A bar or mandrel used as the center on which is built a core. ANNEAL. To soften or render ductile by the application of heat either with or without a carbonaceous material packed around it. BAKED CORE. A dry-sand core which has been subjected to heat (usually in an oven) to render it hard and to fix its shape. BASIN. The portion of a cupola in which the molten iron collects. BED CHARGE. The first coke charged into the cupola. BENCH WORK. Molds of such size that they can be made at the bench. BINDER. A bar of wood or iron with slotted ends to receive bolts, placed across a cope to hold the cope to the drag. BLAST. The supply of air to the cupola. BOD. A ball of clay for closing the tap-hole. BOSH. Same as a swab. See page 290. BOTTOM BOARD. A board placed under a mold. BLOWHOLE. A defect in a casting caused by gases which do not escape. BREAK-OUT. A rupture of a mold permitting metal to flow out at the joint. Also called a " run-out." BREAST. The portion of the lining of a cupola immediately surrounding the tap-hole. BUCKLES. Swellings in the surface of a mold due to the generation of steam below the surface which cannot escape. CARRYING PLATES. Iron plates used to support certain parts of molds. CHAPLET. A piece of metal, shaped in various ways, placed in a mold to support a core. CHARGE. The iron and fuel placed in the cupola or air-furnace. CHEEK. That portion of a three-part mold between the core and the drag. CHILL. An iron surface of a mold, sometimes water-cooled, used to chill the molten iron rapidly and thus produce a hard surface. CHURNING. See Pumping. CLAMPING BAR. A bar used to tighten clamps on a flask. CLAMPS. Devices for fastening copes and drags together. COLD SHUT. An imperfection in the casting due to the metal entering the mold by different sprues, cooling, and failing to unite. COPE. The upper half of a mold. 287 288 APPENDIX B COPE DOWN. To build projecting bodies of sand on the surface of the cope to form surfaces on the casting which are below the level of the joint of the drag. CORE. A body of sand, either green or dry, placed in a mold to form a cavity in the casting. CORE Box. A box in which cores are formed. CORE-PLATE. A flat iron on which green cores are placed for baking. CORE PRINT. The cavity in a mold in which the ends of the cores are set. The projections on a pattern which form and locate the prints in the mold. DRAFT. The taper given in the sides of a pattern which enable it to be easily withdrawn from the molds. DRAG. The lower section of the mold. FALSE CHEEK. A body of sand in a mold, occupying the same position and performing the same function as a cheek, but contained within the cope and drag, altho separate from them. FLASK. The framework of wood or iron in which the sand is packed while being molded around the pattern. FLAT-BACK. A pattern with a flat surface at the joint of the mold. Thus a flat-back pattern lies wholly within the drag, and the joint of the cope is a plane surface. FLOW-OFF. A channel cut from a riser to permit metal to flow away from it when it has risen to a certain height. GAGGERS. Rods of wrought or cast iron with one end bent at a right angle used to support hanging bodies of sand in a mold. GATE. The hole in the cope thru which metal is poured into a mold. GREEN CORE. A core which has not been baked. GREEN SAND. Ordinary molding sand which has not been baked or given other heat treatment except by contact with molten metal in a mold. GREEN-SAND MATCH. A false cope in which patterns are placed while a drag is being made. Its object is to avoid making a difficult joint in each mold where a number of castings are to be made from the same pattern. JOINT. That portion of a mold where the cope and drag come together. LOAM. A mixture of molding sand and clay for making loam molds. MELTING ZONE. That part of the cupola above the tuyeres zone where the metal fuses. MOLD BOARD. The board on which patterns are laid when making a drag. NOWEL. A large core, usually in a loam mold. PARTING. The place on which the pattern is split. PEG GATE. A round gate leading from a pouring basin in the cope to a basin in the drag, whence sprues lead into the mold. GLOSSARY OF TERMS IN COMMON USE 289 POURING BASIN. A basin formed in the cope into which iron is poured. PUMPING. The action of feeding iron to a casting from a shrink-head by forcing it in with a rod moved up and down in the shrink-head. RISER. A gate formed over a high portion of a mold to act as an indi- cator when the mold is filled with metal and also to act as a feeder to supply iron to the casting as it shrinks. RUNNER. A deep channel formed in the top of the cope connecting with gates into which the iron is poured. SCABS. Imperfections in a casting due to the surface of the mold having broken and allowed the loose sand to become imbedded in the iron. SHRINK-HEAD. A large riser containing a sufficient body of metal to act as a feeder for metal contracting upon solidification. SHOT. Globules of metal formed in the body of a casting and harder than the remainder. SKIM GATE. A sprue so arranged as to skim any impurities from the surface of the molten iron as it flows into the mold. SKIN-DRIED MOLD. A green sand-mold whose surface has been baked for a depth of an inch or more. SNAP FLASK. A flask hinged at the corners, and separable at one corner so as to allow its removal from a complete mold. SOLDIER. A wooden stick or rod, clay-washed, used to support bodies of hanging sand or large green-sand cores. SPRUE. The channel leading from the gate to the mold. The metal which solidifies in this channel and adheres to the casting after cooling. SWEEP. A piece of wood or iron revolved about a center to form the surface of a mold. TIGHT FLASK. A flask with a rigid frame; the opposite of " snap flask." TUYERES. The openings in a cupola thru which air is blown. UPSET. A shallow frame set over a flask in which a green-sand match is formed. VENT. A small hole formed in a mold to permit escape of gas from it. WHIRL GATE. A gate or sprue arranged to introduce metal into a mold tangentially, and to thereby give a swirling motion. COMMON TOOLS USED BY THE MOLDER Shovel used for cutting up a sand heap and filling a flask. Water pail used for wetting down the sand for tempering, and for wet- ting a swab, or a bosh on floor molding. Riddle, a sieve used for sifting sand on the surface of a pattern when starting a mold. The size of riddle is designated by the number of meshes 290 APPENDIX B per linear inch. Fine castings with surface detail require finer sand, hence a finer riddle. Rammers used for pounding sand around the patterns in the flask. Small rammers are of maple, larger ones of iron. Strike, used for scrape off extra sand not wanted from the top surface of a cope or drag. A board or thin strip of bar iron. Gaggers and soldiers for holding sand pockets. See page 289. Bellows used to blow parting sand from the pattern, also to blow loose sand and dirt from surface of mold. Bosh or swab. A bundle of hemp, pointed at one end; bound with twine at the other. Used to squeeze water around the edge of the pattern before drawing. Used also the apply wet blacking to dry-sand molds before drying. Rapping and clamping bar of steel; pointed at one end to enter rapping plates in patterns; flattened and turned up at the other. Rapping iron used to strike a rapping bar entering thru a gate hole in order to jar all faces of the pattern at same time. Draw-screws or eye-bolts threaded on one end. Used to draw large wood or metal patterns from a mold. Draw-spike, steel, pointed at one end. For rapping and drawing pat- terns. Used mostly on bench work for small patterns. Spring Draw Nail. Used for drawing small patterns. Two rods joined by spring, press outward and grip the pattern when released. Wooden Gate Pin, or Sprue. A round tapered pin used to form a gate thru the cope into which iron is poured. Gate- or Sprue-Cutter. A sheet of brass, semi-circular at the edge, used to cut a channel in the drag from the gate to the mold. Vent Wires. Steel wires, upset on one end, and with a handle on the other, used to " vent " or to make perforations for the escape of gases from the mold. Clamps, used in many styles and sizes in conjunction with wedges, for holding the cope and drag tightly together. MOLDER'S SMALL TOOLS (The following are tools usually furnished and owned by the molder himself.) Trowels of different styles and sizes, used for making the joint on a mold, and for finishing and smoothing surfaces. Slickers, used for smoothing or slicking, patching, building up sand, forming corners, etc. Different styles are known as " bead " slickers, " spoon " slickers, " double-enders," etc., and are used on different kinds of curved and straight surfaces. GLOSSARY OF TERMS IN COMMON USE 291 Corner Tools, for slicking corners, both inside and outside. Pipe tools, usually of cast iron, with a handle set vertically in the center. Used for slicking interior of cylindrical surfaces. Flange tools, of steel. Used for slicking flanges on pipes or cylinders. Hub tools, used in cylindrical portions of molds such as hubs of pul- leys which are too small to permit the use of a pipe slicker. Lifters, used in lifting loose sand from deep places in the mold. The heel of the lifter is also used to slick the deep places after the loose sand has been removed. APPENDIX C GENERAL DIRECTIONS FOR WRITTEN WORK * MECHANICAL DETAILS Paper. Standard ruled note-book paper or a prescribed equivalent must be used. For typewritten work, title pages, and sketches, blank paper is required. Tabulated data and curves should be put on the special forms of paper designated for these parts of the report. Ink. All exercises should be written in dark ink. Pencil sketches are sometimes permissible in inspection-trip and laboratory reports, but, unless otherwise specified, drawings should be inked in with India ink. Page Numbering. All pages should be numbered in the upper right-hand corner. It is well to indicate also the experi- ment or series of experiments to which a particular page refers. Paragraph Indentation. Paragraphs should be indented at least i in. Margins. The left-hand margin should be kept straight, and flush with the red line. The right-hand margin may be slightly irregular, but should not be crowded. Headings. Sectional headings should be placed in the middle of the page and should be separated by at least two spaces from the divisions between which they occur. Sub- headings should appear at the left, as in the foregoing instruction sheets, and preferably one space above. Formulas and Chemical Equations. Equations and for- * By C. W. Park, A.M., Associate Professor of English, University of Cincinnati. 292 GENERAL DIRECTIONS FOR WRITTEN WORK 293 mulas should be dropped to the middle of the line below the one in which they are introduced, thus: or, in the case of a chemical equation, Ca(HCO 3 ) 2 + CaO = 2 CaC0 3 +H 2 0. Chemical Symbols. In connected discourse, the name of a chemical compound and not its formula should be used. If the formula is given, it should be placed in parentheses immediately after the name of the compound, thus: " The carbon dioxide (CCfe) escapes, and the calcium car- bonate (CaCOs) is precipitated." It should not be written, " C02 escapes, CaCOs precipitated," unless, perhaps, in notes. In equations and in tabular statements, of course, the symbols should regularly be used. Abbreviations. Only authorized abbreviations, such as B.t.u., etc., should be used. If, for example, the statement just made were written, " Only auth. abs. should be used," the confusion arising from the use of the unfamiliar and illegitimate abbreviations would more than offset the saving in space. A false notion of economy of expression often leads writers to invent for their own use a sort of cryptic shorthand which is neither English nor any other intelligible language. It is much safer to write out all words for which there is not a recognized abbreviation. A few of the more common abbreviations used in technical writing are given below. For additions to this list and for explanation of unfamiliar abbreviations which he encounters in his reading, the student sliould consult one of the larger dic- tionaries. ans ..................... answer B.t.u ................... British thermal units B.w.g ................... Birmingham wire gage C ...................... centigrade 294 APPENDIX C cc cubic centimeter cf compare (Lat. confer) cp candlepower cm centimeter cu cubic c.p chemically pure deg. (or ) degree e.g for example (Lat. exempli gratia) e.m.f electromotive force Fig Figure F Fahrenheit f.o.b free on board ft. (or feet gal gallons gm. gram gr grains H.P horsepower H.P. hr horsepower hour i.e that is i.h.p indicated horsepower in. (or ") inches inst the present month kv kilovolt km kilometers Ib pounds min minutes mm millimeters ms manuscript no number oz ounce r.p.m rotations per minute sec seconds sp.gr specific gravity sq.f t square feet t tons ult the preceding month GENERAL DIRECTIONS FOR WRITTEN WORK 295 viz namely vol volume vs against yd yard Figures. Use figures for dimensions, distances, measures, weights, degrees (of angles or temperature), dates, specific gravity and decimals. Examples are as follows: 4 by 6 ft.; 100 yd.; 12 gal; 13 oz.; 212; Jan. 5; 0.005. Exception: Figures should not be placed at the beginning of a sentence. Spell out isolated enumerations of one or two words; e.g., " four propellers," " twenty-five separate parts." In a compound adjective, one element of which is a numeral, the figures, rather than the word should be used; e.g., 28-inch mains, a 1 6-pound sledge. Sketches and Other Illustrations. Wherever possible, a sketch or a photograph should be used to illustrate the more difficult and the more important parts of a report. These illus- trations should be placed within the text, or in the case of full- page sketches and photographs, immediately after that part of the text in which they are described. Each figure should be given a title that will explain its connection with the read- ing matter, and should be designated as " Fig. 2," etc., in order that reference may be made to it in other parts of the report. Curves. In plotting curves, choose the scale of abscissae so that the largest value reaches nearly or quite across the page. Choose the scale of ordinates in a similar manner. Always select the units so that each graduation of the cross-section paper represents a convenient number of units or a convenient fractional part of a unit. The two scales need not be alike, but, in general, similar quantities should have similar scales. The independent variables are usually plotted as the abscissae and the dependent variables as the ordinates. An exception to this rule is found in the case of load-deformation curves. Defor- 296 APPENDIX C mations (deflections, elongations) are invariably plotted as abscissae and the loads as ordinates. When the graph is a calibration curve, and it is certain that there is no error of observation, the curve should be a smooth line passing thru every point. In case of other curves, always draw the straight line or the smooth curve that most nearly fits all of the points plotted. Deviations from the points in the curve usually indicate errors of observation. Such errors are corrected by drawing the curve thru the mean position of the points. Give each curve a title; e.g., " Cooling Curve of Pure Iron." The curve sheet should also contain the name of the observer, the names of his assistants, the source of the data, and a desig- nation of the apparatus or material to which the curve applies. The data and information given on the curves should be sufficient for the application or interpretation of the curves without reference to any other part of the report. Black drawing ink, French curves, and a ruling pen must be used in drawing curves. The curve should ordinarily be a full line, but in case two or more curves are drawn close together some distinction should be made in order to avoid confusion. Such distinctions may be made clear by broken lines, different colored inks, and different geometrical figures (circles, triangles, etc.), to indicate the points of the curve. Broken lines may also be used to indicate hypothetical extensions of known curves. Details for Identification of Manuals. Each laboratory manual should contain the following data: Name of student. Course. Squad number. Condition of Manuscript. No manuscripts should be sub- mitted which contain blots, insertions, or crossed-out passages. Erasures, if any are necessary, should be neat and inconspicuous. All work should be carefully revised and carefully prepared in a THE WRITING OF REPORTS AND ABSTRACTS 297 neat and legible form before it is submitted to the instructor. The sheets should be inserted in their proper sequence, and should be bound securely in the laboratory book. THE WRITING OF REPORTS AND ABSTRACTS Typical Outline for a Laboratory Report. Altho labora- tory reports differ widely in subject matter, the problem of organization is very similar in all of them. The outline given below represents a logical development of the subject as well as a convenient distiibution of the material in the report of a test. With a few changes here and there to fit special conditions, this form will be found serviceable in the writing of nearly every kind of laboratory report. (1) Object. This division should consist of a clear, full, and concise statement of the object, preferably in the form of a simple declarative sentence. Since the report is not written until after the test has been performed, the statement should be put in the past tense; e.g.: " The object of this test was to determine the steam consump- tion of a Brownell 10X12 engine." (2) Theory. This division should contain a general state- ment of that data to be obtained in such a test, together with the fundamental principles on which the test depends. For example, the second paragraph of the report indicated above might begin as follows: " In testing for steam consumption, the chief data to be obtained are (a) the horsepower of the engine, and (b) the weight of the steam passing thru the cylinder in a given time." Where formulas are to be applied in the test, they should be given in this part of the report. Since the discussion deals with general theory, this section of the report should be expressed in the present tense. (3) Apparatus. The apparatus used should be described, with emphasis on new apparatus and on special devices used in the particular test in question. At the beginning of the section 298 APPENDIX C the various pieces of apparatus should be enumerated. New and special pieces may then be described in detail. In so far as this part of the report deals with particular appa- ratus, it should be put in the past tense, e.g. : " The apparatus used in this test consisted of the follow- ing:" etc. (4) Procedure. This section contains an account of what was done in carrying out the successive parts of the test. Care should be taken to omit preliminary and non-essential opera- tions, and to follow the actual order in which the work was done. The narrative should be impersonal, and should be given in the past tense and the passive voice; e.g.: " Readings were taken at intervals of five minutes," etc. (5) Results, a. Summary of results. Conclusions drawn from the data should be stated briefly and clearly. In some cases it may be desirable to compare them with results obtained in other tests. b. Curves (seepage 295). c. Sample calculations. They may be brief, but they should be typical of the mathematical processes involved (see page 230). d. Data. The data should be presented on special ruled paper designed for the purpose. It should be arranged in tabular form and in parallel columns. (6) Sketches (seepage 295). (7) Original Data. Rough notes taken during the test should be submitted from time to time as evidence of the accuracy with which observations were made. If a log book is kept, it will answer the purpose. The Writing of Abstracts. An abstract summarizes briefly in connected discourse the gist of a discussion. Unlike the synopsis, which may be more or less fragmentary or discon- nected, the abstract is a complete composition, a miniature reproduction of a larger work. Comment, an important feature of the review, is lacking in the abstract, and the latter form con- tains little or no quoted matter. The abstractor, using his own THE WRITING OF REPORTS AND ABSTRACTS 299 language and keeping the proportions of the original, restates compactly the main points in the discussion. The making of an abstract is thus a test of his ability both to digest another's work and to write clearly and smoothly. For purposes of illustration, the foregoing paragraph may be regarded as an abstract of a short lecture on the writing of abstracts. Roughly speaking, each sentence in the summary is a condensed statement of the thought in a paragraph or a section of the lecture. The relation between the divisions of the longer discussion should therefore be reflected in the connection between the sentences in the summary paragraph. Another illustration is an abstract of the matter contained on p. 313 of " Materials of Construction," by A. P. Mills: Paragraph 351. " Pouring the Iron " " The molten iron is transferred from the cupola to the molds in heated clay-lined ladles of the top-pouring type. Small molds are filled from hand ladles; for larger castings ladles of sufficient capacity are carried on trucks or by cranes. A gentle, steady and uniform stream of quiet, slag-free metal sufficient to fill the entire cavity should be gently poured into the mold, which is ordinarily broken up soon after the solidification of the iron, to allow the casting to cool more rapidly." Place at the beginning of each abstract a complete reference to the original article. The method of noting the volume numbers by a numeral preceding the name of the publication, and the page numbers by numerals immediately following the name of the publication is universally used in legal publica- tions, is compact and clear and much to be recommended. Wherever possible page numbers should be inserted to indicate the length of the article. For instance, Aitchison, Leslie. " The Theory of the Corrosion of Steel." 15 Metal- lurgical and Chemical Engineering, 88-92. The following example is quoted from 20 (I) Science Abstracts, 15. " Apparatus for the Commercial Testing of Permanent Magnets. B. G. Betteridge. (98 Electrician, 213-215, Nov. 17, 1916.) The manufacture 300 APPENDIX C of large numbers of steel magnets, for magnetos and meters, in this country, to take the place of those formerly supplied in Germany, has necessitated the design of suitable testing apparatus for the magnets. The two most important features the manufacturer requires to know are (a) the coercive force and (b) the remnant flux. Owing to the large number of magnets to be tested, the apparatus should satisfy the following conditions : (i) Allow of individual testing of magnets. (2) The test should take the minimum time to carry out. (3) The apparatus employed should be as simple and robust as possible. (4) The indications obtained should, wherever possible, be direct-reading and readily interpreted, so that the apparatus could be used by comparatively unskilled persons. (5) The results obtained should be as accurate as possible, and should give as far as possible, the true value of the magnetic properties of the magnet. The author's method is to use an electromagnet to energize the steel under test. A disc is rotated in an air-gap in the magnetic circuit and the voltage generated in the disc measured on a milli voltmeter. By this method accurate measurements of the flux in the magnet, and the magnetizing force applied to it, can be obtained. By varying the exciting current the properties of the magnet can be investigated and its hysteresis loop obtained. The author's final design consisted of two solenoids mounted on a base con- taining the flux-measuring device proper. The two legs of the horseshoe magnet under test are inserted in the two solenoids and rest on a pair of angles in an aluminum base. These angles are separated by non-magnetic distance pieces, and in the space between runs a soft-iron disc having a copper ring attached to its outer edge. The disc is carried on a spindle made of a copper alloy. A carbon brush is run on this spindle, and another on the periphery of the disc. The positions of the two solenoids are adjustable, so that various sizes of magnets can be tested. The magnetizing force applied in the test is obtained by multiplying the magnetizing current in milliamps. by a suitable constant. The soft-iron disc is driven by a small motor; the millivolts gen- erated in the disc, when multiplied by a constant, give the flux in the magnet under test. Line drawings and illustrations are given of the apparatus, and the author states that a smart operator can test a batch of magnets for coer- cive force and remanence at a rate of about 30 to 45 seconds per magnet, and get results within an extreme error of 2 per cent." As an aid in the study of a subject, it is well to practice regularly the making of abstracts from lectures and articles. One of the best schemes for the preservation of these summaries in a readily accessible form is to take notes on thin paper, cut slightly less than the page-size of the text used, and to paste the sheets into the text book at appropriate places. Interleaving paper, gummed along one edge, may be obtained for this purpose. THE WRITING OF REPORTS AND EXTRACTS 301 Small loose-leaf sheets in a note-book may be used for abstracts not pasted in the text-book, but cards are better for cultivating the filing habit. One side of a 5 by 8-inch, or both of a 3 by 5 -inch card will ordinarily suffice for one article. Notes thus taken should be filed according to a definite indexing system. INDEX Abbreviations, 293 Abstracts, 297 Accidents, 2 Acid, definition, 27 metallurgical, 6 refractory, 27 slag, 33 Additions to melts, method of adding, 223 Air, analysis, 243 Alpha iron, 126 crystallization, 154 hardness, 154 magnetism, 154 Allotropic modifications, 1 26 theory 01 hardening, 154 Alloy steels, hardening, 153 Alloy systems, 70 Alloys, constitution of, 79 Alumina, melting-point, 25 Aluminates, thermochemistry, 235 Amorphous cement, 134 Amphoteric substances, 33 Annealing, 137, 157, 287 furnace, 19 practice, 159 Anthracite, analysis, 243, 246 Antimony, etching, 83 -lead alloys, 70 melting-point, 66 Anvils, heat treatment, 164 Apparatus, care of, 2 personal, 2 special, 3 squad, 2 Arbor, 287 Arc furnace, 104 temperature, 258 welder, 45 Arrests, 57, 70, 125 Ashes from coal, 247 Assignments, lessons, 3 queries, 4 Atmosphere, in a furnace, 10 pressure of, 240 Atomic weights, 226 Attendance, i, 4, 5 Austenite, 152, 168, 213 constitution, 168 etching, 168 formation, 152 hardness, 168 in cast iron, 213 photomicrograph, 153, 168 Augite, 35 Baked core, 287 Ball races, heat treatment, 163 Barium carbonate, as cementing agent, 183 Base, 27, 36 Base metal thermo-coupies, 51 Basic refractory, 27 slag, 33 Basin, 271, 287 Battering tools, heat treatment, 165 Battery, primary, 42. Bed charge, 287 Bedding-in molds, 275 Bench molds, 280, 287 Benzophenone, boiling-point, 66 Beta iron, 126 hardness, 154 Binders, 279, 287 Bituminous coal, analysis, 246 Black body, 112 experimental attainment, 112 Blast, 287 303 304 INDEX Blast furnace, 19 charge, 266 operation, 262, 266 Blistered negatives, 88 Blister steel, 181 Blowholes, 284, 287 Bod, 287 Boiler corrosion, 204 Boiler makers' tools, heat treatment, 164 Boiler plate, photomicrograph, 173, 174 Boiling-point apparatus, 46 table, 66 Borates, thermochemistry, 235 Bosh, 287 Bottom board, 287 Boyle, law of, 240 Brass working tools, heat treatment, 163 Breakage deposit tickets, i Break out, 287 Breast, 283, 287 Brinell machine, 95 directions for use, 99 hardness numeral table, 100 limitations, 96 -meter, 97 oil, 100 British thermal unit, 234 Brittleness, Stead^, 135 Broaches, heat treatment, 164 Buckles, 287 Bucking board cleaning, 29 Burden of furnaces, 262 Burned steel, 135 Bush hammers, heat treatment, 163 By-product coke-oven gas analysis, 249 Calcine, analysis, 263 Calibration curves, 296 of optical pyrometers, 112 of radiation pyrometers, 112 of thermocouples, 63 Calorific intensity, 257 carbonaceous-fuel furnaces, 103 electric furnaces, 104 oxy-hydrogen flame, 257 Calorimetry, 233, 251 Calory, 233 Candle lamp, 83 Captain's duties, i, 3, 4 Carbon, influence on hardening steel, 141 -iron equilibrium diagram, 126, 131 monoxide as cement, 182, 185 Carbides, thermochemistry, 235 Carbonates, thermochemistry, 235 Carborundum, dissociation tempera- ture, 25 Carburizing, 180 mechanism of, 181 Care of equipment, 2, 3, 4 Carrying plates, 287 Caron's cement, 183 Case carburizing, 180 mechanism of, 181 exfoliation, 189 Castings, cost factors, 269 defects, 284 design, 270 Cast iron: composition, 213, 217, 218 cooling curves, 213 gray, 215 mottled, 216 physical properties, 216 white, 215 Cast steel, photomicrograph, 136 Cement: blending of raw materials, 259, 261 constituents, 259, 262 refractory, 54 Cementation or case carburizing, 180 by carbonaceous materials, 184 by carbon monoxide, 182, 185 by charcoal, 183 by cyanides, 184 by hydrocarbons, 187 by molten baths, 184 furnaces, 187 index, 259 types, 185-187 variables, 192 Cementite, 127, 213 effect on strength of cast iron, 217 INDEX 305 Cementite etching, 168, 173, 177 -ferrite equilibrium diagram, 127, 131 habit, 173 hardness, 173 instability of, 215 primary, 173 Chaplets, 279, 287 Charcoal as cementing agent, 180-183 Charge of furnace, 262, 287 Cheek, 274, 287 Chemical equations, 225, 292 symbols, 293 Chill, 287 Chisels, heat treatment, 164 Chucks, 280 Churning, 287 Circular shapes, molding of, 277 Clamping bar, 287, 290 Clamps, 287, 290 Coal, heat of combustion, 247 Coke oven gas, analysis, 249 Cold-end corrections, 53 pail, 54 Cold shot, 286 Cold shuts, 286, 287 Cold- working hardness, 153 Color of hot bodies, 116 Combustion, 243 air required, 243 total heat of, 247 Concentration, 19 Conduction in furnaces, 105 in quenching baths, 147 Cones, pyrometric, 8 Seger, 8 Constitution of alloys, 79 of metals, 134 Contact electromotive force, 41 and electrolysis, 197 Convection in furnaces, 105 in quenching baths, 147 Converters, 20 Cooling curves, 56 cast iron, 213 inverse rate, 76 lead, 58 plotting, 68 Cooling curves, salt solutions, 71 steel, 126, 129 water, 58 white metal, 76 Cope, 271, 287 Coping down, 275, 288 Copper leaching, 199 melting-point, 66 reduction, 22 Core, 279, 288 box, 288 oven, 279 plate, 288 print, 279, 288 Corrosion, 12, 194 factors, 41, 201, 202 mechanism, 200 of boilers, 204 of busy iron, 203 of homogeneous metal, 203 of hot-water systems, 204 of wrought iron, 203 short time tests, 202 Counterboring forgings, 143 Cracking in heat treatment, 142 Cranes, 282 Crucibles, care of, 3, 60 Crucible steel, 162 Crystalline grains, 134 failure, 135 growth, 135 Crystallization of steel, 133-135 Crystalls, etching, 169 Cup and cone steel, heat treatment 164 Cupellation, 13, 1 6 Cupola, 282 breast, 283 operation, 283 shell, 283 Curves, standard practice, 5, 295 Cutlery, 164 Cutting hardness, 94 molding sand, 210 Cyanide as cementing agent, 184 Decomposition, heat required. 238 Defects of castings, 284 306 INDEX Deposit tickets, i Design of castings, 270 electrical furnaces, 104 Developer, 88 Development of negatives, 87, 92 of photographic paper, 88, 92 Die castings, 270 Dies, heat treatment, 163-165 Diffusion of carbon into iron, 180, 181 of CO : CO 2 into iron, 182, 185 Diopside, melting-point, 66 Dirty negatives, 88 prints, 88 Dissociation, 194 Double decomposition, thermochemis- try of, 238 Draft, 288 Drag, 271, 288 Draw screws, 290 spike, 290 Drifts, heat treatment of, 164 Dry-sand molds, 280 Eckel's rule, 261 Electric furnaces, 102 advantages, 103 design, 104 temperature attainable, 104 tube, 1 06 types, 104 Electrolysis, 196, 197 Elements, see thermo-couple elements. Emissivity, 106 hot body, 116 measurement, 123 theoretical black body, 112 variation, 112, 115 Endothermic reactions, 233, 239 Enrollment, i, 9 Enstatite-wollastonite equilibrium di- agram, 34 Equation of thermo-couple, 63 Equilibrium, 19 Equilibrium diagram, 70 ferrite-cementite, 127, 131 iron-carbon, 126, 131 Equilibrium, salt-water, 74 wollastonite-enstatite, 34 Equipment, general laboratory, IV, 3 metallography, 90, 176 Etching, 80, 177 agents for steel, 167 antimony, 83 pits, 169 polished specimens, 80 Eutectic, 34, 73, 127 appearance, 73 cementite, 213 properties, 73 Eutectoid, 127 Exfoliation, 189 Exothermic reactions, 233 Expansion at Ar 3 , 143 Facings, 278 False cheek, 288 Ferrite, 126 Ferrite-cementite equilibrium diagram, 127, 131 etching, 168, 173 habit, 173 hardness, 173 Ferroxyl test, 203 Fery radiation pyrometer, 113 focusing, 114 limitations, 114 Figures in written work, 295 Files, heat treatment, 163 Filing notes, 300 Fire clay, melting-point, 25 Fixing negatives, 92 prints, 93 Flame temperatures, 257 Flasks, 274, 281, 288 Flat-back pattern, 270, 288 Floating cores, 286 Floor molding, 280 Flow-off, 288 Flue dust, analysis, 263 gases, composition, 245 Fogged negatives, 88 Forge fire, making, 158 Formulas, 292 INDEX 307 Foundry equipment, 282 Fracture, crystalline, 135 practice, 269 examination, 133 rail, 134 steels, 138 tests, 133 Freezing-point table, 66 Fuel, heat of combustion, 247 Fullers, heat treatment, 165 Furnace, atmosphere, 10 burden, 262 cementation, 187 charges, 262 electric, 102-106 heat treatment, 142 oven, see Oven-furnace. Gaggers, 288 Gamma iron, 126 crystallization, 154 density, 154 hardness, 154 Gangue, 32, 33 Gases, volume and pressure relation, 240 volume : weight relation, 241 Gate, 288 Gay-Lussac, law of, 240 General apparatus, 2 laboratory equipment IV, 3 Glasses, composition, 35 Glossary, 287 Goggles, 2 Goutal's formula, 250 Grade of steel, 162 Grades, method of computing, 4, 5 Grain, refining, 135, 137 Granite points, heat treatment, 163 Granular pearlite, 172 Graphs, see Curves. Graphite crucibles, treatment, 60 in cast iron, 215, 217 Graver tools, heat treatment, 163 Gray cast iron, 215 Green cores, 288 Green sand, 288 Green match, 288 molds, 280 Halation, 88 Hammers, heat treatment, 163, 164 Hardening, 141 allotropic theory, 154 alloy steels, 153 carbon, 141 cracks, 142 development of metarals, 152 theories, 153, 154 Hardness, 78, 94 indentation, 95, 100 numeral, 96, 100 rebounding, 94 Heat available in boiler fuel, 255 conduction, 105 convection, 105 control, 142 evolution of reactions, 232, 233, 235, 252 in quenching bath, 147 mechanical equivalent, 103 radiation, 106 transfer in furnaces, 104 treatment, see Annealing and quench- ing. unit, 233 Heating curve of furnace, 9 furnace, 19, 142 power of liquids, 147 practice, 141 High-speed steel, 162 High-temperature reactions, 252 History of metallography, 167 Holborn and Kurlbaum pyrometer, 120 limitations, 121 operation, 121 Hot bodies, color of, 116 water systems, corrosion, 204 working of steel, 135 Hours for work, i, 4 Howe's diagram, 131 Humphrey's theory of hardening, 154 Hydrates, thermochemistry, 235 308 INDEX Hydrocarbon cementing agents, 185, 187 thermochemistry, 235 Hypereutectoid steel, 173 Hypoeutectoid steel, 172 Illumination of opaque specimens, 79 Illuminator, vertical, 79 Illustrations, 295 Impact testing, 139 Incandescent bodies, color, 116 Indentation hardness, 95, 100 Indicators, 203 Induction furnace, 104 Ink, 292 Inspection of rails, 134 Internal strains 142, 286 Inverse-rate curves, 76, 129 lonization, 195 Iron, allotropy, 126 carbide, 127 -carbon equilibrium diagram, 126, 131 cooling curve, 126 reduction, 23 Joint, 288 Joule, 102 Keep's test bars, 209, 220 Knives, heat treatment, 164 Laboratory of furnace, 7 tube furnace, 106 Ladles, 284 Latent heat, 57, 238 Lathe tools, heat treatment, 163 Lead-antimony alloys, 70 cooling curves, 58 melting-point, 66 reduction, 22 refining, 13 Le Chatelier optical pyrometer, 117 theory of hardening, 154 Lenses for metallography, 79 Lesson assignments, 3 Lifters, 291 Lighting furnace, 9 Limestone, analysis, 263 Lithium metasilicate, melting-point, 66 Loading plate holders, 91 Loam, 288 molding, 278 Machine-made molds, 281 Magnesite brick, 28 Magnetic testing, 299 Mandrels, heat treatment, 163 Manganese in cast iron, 218 Manuscripts, 296 Margins, 292 Martensite, 155, 169 etching, 169 hardness, 169 photo-micrograph, 155, 156 Mason's tools, heat treatment, 164 Mass conversion, 242 Match molding, 276, 277 Matte, 20, 263 Maximum temperature, 7 McCance's theory of hardening, 154 Mean specific heat, 253 Mechanical equivalent of heat, 103 Media for quenching, 147 Melting practice for white metal, 78, 82 zone, 288 Metallic specimens, care and storage, i7S Metallography, 79 equipment, 90, 176 historical development, 163 microscope, 79, 83 steels, 167 utility, 173 Metal protection tubes, 52 Metallurgy: laboratory work, 6 progress, 40 reagents, 36 Metals, constitution, 134 Metric mass conversion, 242 Microscope, metallographic, 79-83 ' Microscopic analysis, 93 Mill scale, formation, 13 rust inhibitor, 204 INDEX 309 Mine water, 199 Modifications, allotropic, 126 Moh's scale, 94 Mol, 227 Mold board, 288 Molding, 209, 270-290 equipment, 281 loam, 278 machines, 281 operations, 270 sand, 278 skill, 273 sweeps, 277 tools, 289 Molds, classification, 280 Molecular equivalence, 260 movements, 125 volume, 241 weight, 227 Molecule, gaseous, 231 size, 241 Monochromatic light, 87 Morse pyrometer, 120 limitations, 121 operation, 121 Mother liquor, 71 Mottled cast iron, 216 Mounting of opaque specimens, 80, 82 Naphthalene, boiling-point, 66 Natural gas analysis, 249 Negatives, faults, 88 Neutral atmosphere, 10, 18 refractory, 27 Nickel, melting-point, 66 Nitrides, thermochemistry, 235 Noble metal couples, 51 Notes, filing of, 300 Nowel, 288 Ohm's law, 41, 102 Oil baths, 148 Open-hearth furnaces, 26, 28 heating, 258 Optical pyrometers, 78 calibration, 112 Holborn and Kurlbaum, 120 Optical pyrometers, Le Chatelier, 117 Morse, 120 precautions in use, 122 principle, 116 Shore, 118 Wanner, 119 Orthochromatic plates, 87 Osmond's theory of hardening, 154 Osmotic pressure, 194 Outline of reports, 297 Oven furnace, care, 2 construction, 7 lighting, 9 operations, 7 Over-exposure, 87 Oxidation of metals, 13 Oxides, thermochemistry, 235 Oxidizing atmosphere, 10, 18 reactions, 6, 12 Oxygen, 12 Oxy-hydrogen flame, 233 temperature, 257 Page numbering, 292 Palladium, melting-point, 66 Paper, 292 photographic, 87-92 Paragraph indentation, 292 Partial cementation, 180 Parting, 288 Passive state of metals, 199 Pattern, 270, 278 pieces, 278 Pearlite: composition, 172 constitution, 171, 173 formation, 172 granular, 172 spheroidal, 172 Peen hammers, heat treatment, 163 Peg gate, 288 Peltier effect, 43 Personal apparatus, 2 Phenolphthalein, 203 Phosphorus in cast iron, 218 Photographic paper, 87 development, 88 310 INDEX Photographic printing, 92 Photographic plates, 87 development, 87, 92 exposure, 87 Photomicrography, 86-91 lighting, 87 Pickling vats, 284 Pig iron, fracture, 133 Pit molds, 280 Plane glass reflector, 79 Plates, photomicrographic, 87 handling, 91 loading, 91 orthochromatic, 87 Plate steel, photomicrograph, 173, 174 Plating metal from solution, 198 Platinum couples, 51 melting-point, 66 Polarization, 198, 199 Polishing machines, 1 73-1 75 metal specimens, 80, 82, 173 Potential drop, 102 Pot furnace, 107, 108 Pouring basin, 289 iron, 299 Pressure of atmosphere, 240 Primary battery, 42 cells, 197, 205 ferrite, 172 Prints, photographic, 88 faults, 88 manipulation, 92 Problems, method of solution, 227 Process plates, 87 Producer gas, analysis, 243 Protection tubes, 52 Proximate analysis, 249 heat calculation, 250 Pumping, 289 Punches, heat treatment, 164 Pyrometers: calibration, 63, 112 optical, in radiation, in Seger cones, 8 thermoelectric, 41 Pyroscope, Shore, 118 Quartz, allotropy, 25 melting-point, 25 protection tube, 52 Quenching, 143-148 baths, 148 fluids, 148 mechanism of, 147 media, 147 power of fluids, 147 Queries, assignment, 4 grades, 4 Rabble, 15 Radiation, 106, 115 laws, 106, in, 115 pyrometers, in calibration, 112 Fery, 113 limitations, 112, 114 selective, 114 Rails, fracture, 134 inspection, 134 molds, 273 Rapping bar, 290 Ramming bars, 290 Rare metal recovery, 13 Rattling barrels, 284 Ray filter, 87 Razors, heat treatment, 163 Reactions, 233, 239 Reamers, heat treatment, 164 Rebounding hardness, 94 Red radiations, 115 Reducing atmosphere, 10, 18 reactions, 6, 18 Refining grain of metals, 135, 137 lead, 13 Reflector, 79 Regenerative principle, 256, Refractories, 6, 25 acid, 27 basic, 27 binder, ,27 cement, 54 classification, 27 manufacture, 26 melting-point, 25 INDEX 311 Relief polishing, 80 Replacement of metal in solution, 198 Resistance furnaces, 104 Resistivity, 41, 105, 106 Reverberatory, charge, 263 operation, 264, 266 Riddle, 289 Rigging, 282 Riser, 289 Roasting, 13, 15, 266 Roll call, i Runner, 289 Safety precautions: goggles, i lead fumes, 16 lighting furnaces, 9 pouring metals, 22 Salt baths, 142, 148 solutions, 71 volatility, 68 -water system, 70 Sand holes, 285 molding, 278 Scabs, 285, 289 Scale, formation, 13 pivots, heat treatment, 163 Scleroscope, 94, 98 limitations, 97 use, 98 Scratch hardness, 94 Seger cones, 8 composition, 36 manufacture, 30 Selective radiation, 114, 115 Sensible heat, 56, 254 Sets, heat treatment, 165 Shape factor, 105, 106 Sheaves, molding, 274 Shore pyroscope, 118 scleroscope, 94-98 limitations, 97 use, 98 Shot, 289 Shrinkage, 286 Shrink head, 289 holes, 285 Siemen's regenerative principle, 256 Sighting tube, 114 Silica, allotropy, 25 brick, 28 melting-point, 25 protection tube, 52 Silicates, thermochemistry, 236 Silicate slags, 33 Silicon in cast iron, 217 Silver, melting-point, 66 Sketches, 295 Skim gate, 289 Skin-dried molds, 280, 289 Slags, 32 Sledges, heat treatment, 164 Slickers, 290 Slicking of molds, 273 Slotters, heat treatment, 163 Snap flask, 289 Sodium chloride, melting-point, 66 Soldering, 46 Soldier, 289 Solid solution, 35, 127 Solution, 194 pressure, 194 salt, 71 Solution of problems, 227 Sorbite, 155, 171 etching, 171 hardness, 171 photomicrograph, 156 Special apparatus, 3 Specific gravity of gases, 241 Specific heat, 252, 253 Specimens, illuminating, 79 mounting, 80, 82 Spheroidal pearlite, 172 Split patterns, 274 Sponginess in castings, 285 Spotty prints, 88 Spring draw nail, 290 Sprue, 289, 290 Squad organization, i, 9 Stains on negatives, 88 Standard pressure and temperature, 240 ' Standardization, see Calibration 312 INDEX Stead's brittleness, 135 Steel, blister, 181 boilerplate, 173, 1 74 burned, 135 cooling curves, 126, 129 crucible, 162 crystallization, 133 equilibrium diagram, 1 26, 131 experiments, 124 fracture, 133 hardening, 141 metallography, 167 photomicrographs, 135 Stefan's law, 106, in Strains in castings, 286 Strike, 290 Sulfates, thermochemistry, 236 Sulfide, ores, 13 roasting, 13 test, 15 thermochemistry, 236 Sulfur, allotropy, 126 boiling-point, 66 boiling-point apparatus, 46 in cast iron, 217 Supplies, 3 Surface resistance, 105, 106 Swab, 290 Swages, heat treatment, 164 Sweep, 277, 289 finger, 278 Temper, 162, 163 Temperature control, 40 minimum for carburization, 180 Tempering steel, 155, 157, 162 baths, 157 colors', 157 Test bars, casting, 209, 220 Texts, 3 Thermal analysis, 125 units, 233 Thermochemistry, 232 e.s. at high temperature, 252 data, 235 Thermo-couple calibration, 63 constitution, 49 Thermo-couple cost, 50 elements, 41, 49 melting-point, 50 sizes, 50 stability, 50 Thermo-electric equation, 63 Tight flasks, 289 Time, a factor in transformation, 152 of exposure, 87 Tin, melting-point, 66 Tool making, 162 steel, 162 Total cementation, 180 Toughening, 157 Track-layers' tools, heat treatment, 164 Transite board, working of, 109 Troostite, 171 etching, 171 hardness, 171 photomicrograph, 156 Transformation point, 125 apparatus, 128 suppression, 152 time factor, 152 Tridymite, 25 Tube furnace, 106-108 Tubes, protection, 52 Tumbling barrels, 284 Tuyeres, 283, 289 Ultimate strength derived from hard- ness, 96 Underexposure, 87 Upset, 289 Vats, pickling, 284 Vent, 289 wires, 290 Vertical illuminator, 79 Volume relation in equations, 231 Wanner pyrometer, 119 calibration, 120 limitations, 120 Warping in castings, 286 in heat treatment, 142 Water, cooling curve, 58 INDEX 313 Water, latent heat, 238 -salt equilibrium system, 70 specific gravity, 241 Wedges, heat treatment, 165 Wedging-up molding, 275 Weight relation in equations, 225 Welder, 45 Whirl gate, 289 White cast iron, 215 White metal, melting, 75, 82 Witherite, cementing agent, 183 Wollastonite : enstatite equilibrium dia- gram, 34 Work of electrical current, 102 Working steel, 135 Written work, 4, 5, 292 Wrought iron, corrosion, 203 Zinc burning, 13, 15 Zinc, melting-point, 66 UNIVERSITY OF CALIFORNIA LIBRARY BERKELEY Return to desk from which borrowed. This book is DUE on the last date stamped below. MINERAL TECHNOLOGY LIBRARY 1 3 19 LD 21-100m-9,'48(B399sl6)476 376245 UNIVERSITY OF CALIFORNIA LIBRARY