S77 
 
 METAL MIXER 
 
 The easiest, simplest and most exact method 
 
 of mixing iron by chemical analysis, with 
 
 tables and ready made mixtures. 
 
 Indispensable to Molders, Melters ^ 
 and Foundry Men. 
 
 W\ IV. ELLIS 
 
 COPYRIGHT, 1919. OAKLAND, CALIF. 
 
 - 
 
GIFT OF 
 
THE 
 
 ZMETAL MIXER 
 
 The easiest, simplest and most exact method 
 
 of mixing iron by chemical analysis, with 
 
 tables and ready made mixtures. 
 
 Indispensable to Molders, Melters 
 and Foundry Men. 
 
 W. W. ELLIS 
 
 COPYRIGHT, 1919. OAKLAND, CALIF. 
 
CONTENTS 
 
 Introduction 5 
 
 Mixture for Medium Machinery Castings 9 
 
 Soft Mixture for Pulleys, Short Method 12 
 
 Four Iron Semi-Steel Mixtures for Rolls 15 
 
 Correcting Mixtures with Ferro-Silicon or Ferro-Manganese 1 8 
 Semi-Steel Mixture for Rings, Piston Valve Liners, Gears 
 
 Etc., 21 
 
 Mixture for Marine Cylinders Liners 26 
 
 Mixing with a Certain Per Cent of Steel 29 
 
 Figuring Three or More Elements Exact 32 
 
 French Specifications for Semi-Steel Shells 36 
 
 Side Lights on Mixtures 40 
 
 Miscellaneous Mixtures 43 
 
 Analysis of Pig Irons 44 
 
 Approximate Grading Numbers 44 
 
 Approximate Analysis of Important Castings 45 
 
 The Influence Different Elements Have Upon the Iron 46 
 
 Percentage of Silicon for Different Castings 47 
 
 Judging Per Cent of Silicon in Different Kinds of Scrap 48 
 
 Decimal Fractions and Percentage 49 
 
 Cupola Practice 54 
 
 395059 
 
 3 
 
INTRODUCTION, 
 
 In presenting this book to molders and foundry foremen 
 I do so, believing there is a real demand for a book, 
 written in plain every day foundry language that anyone 
 may understand, showing a simple and easy method of mix- 
 ing iron by chemical analysis. I have endeavored to explain 
 every mixture in a manner so simple, that the man who has 
 never mixed iron, or understands anything whatever about 
 foundry work, can, with a few minutes study, make any 
 kind of a mixture, from any number of different grades of 
 iron, by this easy method, almost as well as the more exper- 
 ienced foundry man. We do not require a knowledge of 
 chemistry to be able to mix by analysis. In fact, the aver- 
 age foundry man or foreman has very little use or time 
 for it. But what we must know is the composition of the 
 iron we are mixing, and the percentage of the different 
 elements it contains. The broker generally gives an approx- 
 imate analysis. Drillings should be analyzed for the exact 
 composition. In regard to the influence and relation the 
 elements have to one another, and above all the percentage 
 of the most important of these elements, castings designed 
 for different kinds of work, should contain, I have endeavored 
 to explain, and if followed, will give the reader a good work- 
 ing knowledge of the characteristics of the different elements, 
 which is a big help in making mixtures. 
 
 Foundry iron contains several of these elements, or im- 
 purities as they are sometimes called, but there are only five 
 in which we are mostly interested in. They are silicon, 
 phosphorus, sulphur, manganese and the carbons. Of these 
 five, I think the carbons are the most important, because 
 carbon is the element that gives the iron its character. 
 Foundry iron contains carbon in two distinct forms, called 
 
graphite carbon 'and qoiAbi-ied carbon. And according to 
 the percentage. , of each ^of , these carbons, so will the iron 
 be hard or rof ; . Graphite,; or boft carbon, is always high 
 in very soft open grained iron. Combined or hard carbon 
 is always high in very hard, close grained iron. In making 
 mixtures for the cupola it is a more difficult proposition to 
 take hold of the carbons and figure their content than it is 
 silicon or some other element, so, as a rule if we wish to 
 reduce or change the carbons, we generally add some low 
 carbon steel scrap, or change the carbons by using high or 
 low silicon in the mixture, as the case may require. In 
 making mixtures for the ordinary run of machinery castings, 
 we do not trouble about the carbons because we find if 
 silicon is high, graphite carbon will be high also, and if 
 silicon is lowered, graphite carbon will be lower, and com- 
 bined carbon will be higher, and the more we lower the 
 silicon the more combined carbon we will get in our cast- 
 ings. Chemists long ago proved, if we wish to regulate 
 the carbons it can be done through the silicon, which at 
 once proves that silicon is one of the most important, if 
 not the most important element the founder has to work with. 
 Not only does it influence the carbons, but the other elements 
 also, to a certain extent. For we find if we get silicon 
 normal for the class of work we are making, the other 
 elements also will be normal, especially so, if we use the 
 ordinary run of foundry irons. As the silicon can be raised 
 or lowered as required, it is the first element that should be 
 figured in mixing iron by chemical analysis. When iron 
 contains more than 3.5 per cent silicon, it will begin to get 
 hard. Not hard and strong like a low silicon close grained 
 iron, but hard, short and brittle. So in making mixtures, 
 we never go above 3.25 per cent silicon, and even that per- 
 centage is very rarely used, except in fine stove plate, or 
 work similar to it. There are special mixtures however, for 
 acid proof castings, that call for a much higher percentage 
 
of silicon, but these are exceptional, and are not included in 
 the ordinary run of foundry products. Silicon and mangan- 
 ese are not affected by mineral and vegetable acids, like 
 graphite carbon, sulphur and phosphorus are. So in making 
 mixtures for this kind of work, the combined carbon, silicon 
 and manganese should be high, especially the silicon, which 
 of .course will make the casting very brittle. To make such 
 mixtures takes considerable experimenting, even by the 
 most experienced chemist and metallurgist. In arranging 
 the different mixtures I have tried to make them as progres- 
 sive as possible. The few mixtures from my note books, 
 I thought would give the reader an idea of what has been 
 done with steel mixing. These were made when pig iron 
 was very much cheaper than now, as some were made as far 
 back as 1904. The analysis of different pig irons will also 
 give beginners a working idea of the composition of iron. 
 The few remarks on the influence the different elements have 
 upon the iron will all help the student how to use, and mix 
 them to accomplish a certain purpose. 
 
 In selecting and using scraps, of course is more or less 
 guess work. But, by careful study and selection, and by 
 getting a determination now and then, one will soon be able 
 to judge the silicon content for all practical purposes. But 
 if special work is to be made to specification, such as shells 
 or other governmental work, all the scraps must be melted 
 and pigged, and analysis taken of each cast. Only then can 
 we say with confidence, just what is the composition of the 
 scrap. 
 
 In the mixtures showing the method of mixing three 
 or more irons together, I have used a higher per cent of 
 steel than I would advise to use without experience. Although 
 I have made mixtures containing more than 25 per cent steel, 
 still I am convinced by actual tests, that no improvement 
 or benefit can be obtained by using more. This opinion 
 seems to be general among other foundry men, who have had 
 
any experience with steel and iron mixtures. I have also 
 found if using a higher per cent the best results was obtained 
 when using all pig and steel scrap. Steel mixtures must be 
 melted hot and handled quick when in the ladle. The few 
 examples on decimals and percentage will help refresh our 
 memories, and are handy to refer to while studying this 
 method, as they deal directly with the work of the book. The 
 last chapter deals with the cupola, which are chiefly personal 
 experiences, and agrees closely with our leading foundry men. 
 And if followed as near as possible together with other in- 
 formation in the book, the young foundry man should have 
 no trouble in handling the mixing and melting end of any 
 shop, independent of the class of work being made. 
 
 W. W. ELLIS. 
 
MIXTURE FOR MEDIUM MACHINERY 
 CASTINGS. 
 
 In this mixture we will figure for silicon only, and I 
 would advise the student to work over this first mixture until 
 you understand it. You will then be surprised how simple it 
 is. You will then have the foundation for making any kind 
 of a mixture from any number of different grades of iron. 
 A mixture for medium machinery work should contain about 
 2 per cent silicon, with that percentage, castings from 54- 
 inch in section should machine quite easy. As we lose about 
 two tenths (0.2) of one per cent silicon in melting, we 
 must add that much to our mixture before it goes into the 
 cupola. On account of this loss we must figure our mixture 
 to contain 2.2 per cent silicon. To make this we will use 
 pig iron and scrap enough to make a mixture of 2000 pounds. 
 The pig contains 3.25 per cent silicon, and the scrap 1.75 
 per cent. As we desire only 2.2 per cent, you will notice 
 that I have selected one iron with a higher, and one with 
 a lower per cent of silicon. We will now put the lowest 
 silicon under A, the amount we desire under B, and the 
 highest silicon under C. So placing them in that order they 
 stand as follows: A. B. C. 
 
 1.75 2.20 3.25 
 
 RULE: By subtracting A from B we get .45 remain- 
 der; substracing B from C we get 1.05. We now add both 
 remainders together, getting 1.50. Taking the first remain- 
 der .45 and after affixing two ciphers to it, and moving 
 the decimal point two places to the right, and dividing it 
 by the sum of the two remainders 1 .50, we get 30, the 
 percentage of the C iron to be used in the mixture. Taking 
 the second remainder 1.05, and after affixing two ciphers, 
 and moving the point two places to the right, dividing it 
 also by 1.50 we get the percentage of the A iron, which is 
 70, to be used in the mixture. 
 
Example 
 
 Take A from B. 
 
 1st Remainder 
 Add both remainders 
 
 Sum of the two 
 
 A. 
 1.75 
 
 rs 
 
 B. 
 2.20 
 1.75 
 
 C. 
 
 3.25 
 2.20 
 
 .45 
 1.05 
 
 1.05 
 1 
 
 1.50 
 TABLE No. 
 
 Take B from C. 
 2nd Remainder 
 
 1st remainder .45 representing C iron. 
 Two ciphers affixed and point moved two places and 
 divided by 1.50)45.00(30% of C iron to be used. 
 45.00 
 
 TABLE No. 2 
 
 2nd Remainder 1.05 representing A iron. With two 
 ciphers affixed and point moved two places we divide by 
 1.50)1 05 .00( 70% of A iron to be used in the mixture. 
 10500 
 
 TABLE No. 3 
 
 Don't miss this one point of setting the different 
 silicons under their proper heading. 
 
 Always set the lowest silicon under A, what we desire 
 under B and the highest under C. Then take A from B, and 
 the first remainder will always represent the C iron. Then 
 take B from C and the second remainder will always represent 
 the A iron. As these remainders only represent the A and 
 C irons, they do not tell us how much per cent of each one 
 to take. To find a rate so we can figure what percentage 
 of these two irons to use, we add these two remainders 
 together, and after affixing two ciphers to each one, we 
 divide each one by the base, or sum of the two. The result 
 of this division is of course the percentage of the iron we 
 are to use that each remainder represent, which is clearly 
 shown in tables 2 and 3. When affixing the two ciphers to 
 
 10 
 
each of the remainders, be sure and move the decimal point 
 two places to the right even though it does point off ciphers 
 as in this example. 
 
 We will now check off our mixture. 
 
 According to our figures we are to use 30 per cent of 
 C iron, and 70 per cent of A iron. By multiplying the 
 3.25 per cent of silicon by the 30 per cent, we get .9750 
 per cent, and by multiplying the 1.75 per cent by the 70 
 per qent we get 1.2250 per cent. Adding these two per- 
 centages together will give us our desired percentage of 
 2.2 per cent silicon. 
 
 Example: Percentage of silicon in C iron 3.25 
 
 Percentage of C iron to be used .30 
 
 Per cent of silicon .9750 
 
 TABLE No. 4 
 
 Percentage of silicon in A iron 1.75 
 Percentage of A iron to be used .70 
 
 Per cent of silicon 1.2250 
 
 Percentage of silicon from C iron .9750 
 
 The required per cent amount of silicon 2.2000 
 
 TABLE No. 5 
 
 This method gives the percentage of silicon as well as 
 the percentage of the iron. 
 Example : 
 
 30 per cent of 2,000 Ibs. = 600 Ibs. of C. pig iron. 
 70 per cent of 2,000 Ibs. = 1400 Ibs. of A scrap iron. 
 
 Mixture of 2000 Ibs. 
 
 TABLE No. 6 
 
 In tables 2 and 3 the divisors and dividends has each 
 two decimal places which make the quotients whole numbers, 
 see decimals. 
 
 11 
 
SOFT MIXTURES FOR PULLEYS 
 
 We will make this another two iron mixture, and figure 
 for silicon only, introducing shorter method with less figuring. 
 In making mixtures for pulleys, we should try and keep the 
 silicon three or four points higher than we would for castings 
 in our first mixture. The metal will be softer and of course 
 the shrinkage will be lower. Even with this percentage, all 
 pulley hubs should be stripped and cores taken out. 
 
 Suppose we want a mixture of 1500 pounds containing 
 2.4 per cent silicon in the castings. That means with the 
 loss of silicon in melting our mixture must contain 2.6 per 
 cent before going into the cupola. To make this mixture 
 we will use some of the gates of our first mixture containing 
 2.0 per cent silicon, and No. 1 Sloss pig iron containing 3.6 
 per cent silicon. Putting the lowest silicon under A, our 
 desired under B and the highest silicon under C. We then 
 work out as in table No. 1 . 
 
 By substracting A from B we get our first remainder .6, 
 which represents the C iron to be used in the mixture. Sub- 
 stracting B from C we get our second remainder, 1 .0 which rep- 
 resents the A iron to be used in the mixture. After adding 
 these two remainders together and getting 1 .6, we take 
 the first remainder and after affixing two ciphers to it, we 
 move the decimal point two places to the right, making 
 60.0. We now divide the 60.0 by the sum of the two re- 
 mainders 1.6; which we find goes 37J/2 times. As this first 
 remainder represents the C iron, it shows we are to take 
 37.5 per cent of C iron to use in our mixture. Now, if we 
 are to use 37.5 per cent of C iron, it stands to reason we 
 must use 62.5 per cent of A iron, as 37.5 and 62.5 equals 
 100. So that being the case, further figuring are unneces- 
 sary. Example : 
 
 12 
 
A. B. C. 
 2.0 2.6 3.6 
 Take A from B 2.0 2.6 Take B from C. 
 
 1 st remainder .6 1 .0 2nd remainder 
 
 Add both remainders 1.0 
 
 Divide by 1 .6)60.00(37.5 % of C iron. 
 
 48 
 
 120 
 
 112 62.5 % of A iron. 
 
 80 
 80 
 
 TABLE No. 7. 
 
 You will notice instead of making two separate tables 
 as in our first mixture, we simply take the first remainder .6 
 put it to the right of the 1.6, add two ciphers to it, and 
 move the point two places, then divide it by the 1.6. The 
 result of this division completes all the figures required to get 
 the percentage of the irons to be used in any mixture, so 
 make yourself familiar with the first mixture, then you 
 will be able to follow table 7 with ease. For table 7 is the 
 form in which you will make all your mixtures in actual 
 practice, except, of course, when correcting mixtures, or 
 mixtures that have even number remainders, which will be 
 explained in other mixtures following. 
 
 Percentage of silicon in C iron 3.6 
 
 Percentage of C iron to be used 37.5 
 
 Per cent of silicon 1.3500 
 
 13 
 
TABLE No. 8 
 
 Percentage of silicon in A iron 2. 
 
 Percentage of A iron to be used 62.5 
 
 Per cent of silicon 1 .250 
 
 Silicon from C iron 1.350 
 
 The per cent amount of silicon required 2.600 
 
 TABLE No. 9 
 
 37.5 per cent of 1500 = 562.5 pounds of C pig iron. 
 62.5 per cent of 1500 937.5 pounds of A gate scrap. 
 
 Charge of 1500.0 pounds 
 TABLE No. 10. 
 
 Note: In multiplying by the rate for percentage, you 
 point off two for the whole numbers, and as many decimals 
 as there are in the multiplier and multipliant, which makes 
 four in table 8, and three in table 9. See decimals. In 
 actual practice we would only require tables 7 and 10. 
 Tables 8 and 9 are merely used to prove our figures. 
 
 14 
 
MIXTURE FOR A ROLL SEMI-STEEL USING FOUR 
 DIFFERENT KINDS OF IRON 
 
 In this mixture we will explain a method whereby any 
 number of different kinds of iron can be mixed together. 
 
 When making a mixture to contain several different 
 brands of iron, and not being particular how much of each, 
 the best way is to segregate them, putting all the irons 
 together that contain a lower silicon than we require in our 
 mixture into one group, and all the irons that contain a 
 higher per cent of silicon into another group. After getting 
 the mean percentage of silicon from each of these groups, 
 we have practically but two irons to figure on, and can be 
 worked out as in table 7. Then, after we have found what 
 percentage of each group we are to use, we must divide 
 each percentage into as many parts as there are irons com- 
 posing each group. 
 
 Example: Suppose we wish a mixture of 4,000 pounds 
 for a 20 inch dia-chilled roll, containing 0.6 per cent silicon, 
 adding the usual 0.2 per cent for loss of silicon, would make 
 our desired silicon 0.8 per cent before it goes into the 
 cupola. To make this mixture we will use some of the 
 following irons. 
 
 Sil. Phos. Sul. Mang. T. C. 
 
 Heavy scrap 1.50 0.40 0.08 0.60 
 
 Salisbury pig 1.29 0.30 0.045 0.40 3.85 
 
 Steel scrap 0.2 0.05 0.05 0.50 0.10 
 
 Cargo fleet-pig 0.79 1.52 0.027 0.23 3.12 
 
 TABLE No. 11. 
 
 By putting the two lowest silicons together and dividing 
 them by 2. we find their mean silicon content is 0.495 per 
 cent, which must be put down under A. Putting the two 
 highest silicons together and dividing them also by 2, we 
 
 15 
 
find their mean silicon content is 1 .395 per cent, which must 
 be put clown under C. The desired silicon of course must 
 be put down under B. This gives us practically a two iron 
 mixture, and they stand ready to be figured as in table No. 7. 
 
 A. B. C. 
 
 0.495 0.800 1.395 
 Take A from B .495 .80 Take B from C. 
 
 1 st Remainder .305 .595 2nd Remainder. 
 
 Add both .595 
 
 Divided by .900) 30.500 ( 33-8/9% of C iron 
 
 2700 66-1/9% of A iron 
 
 3500 
 2700 
 
 800 
 
 equals 8/9 
 
 900 
 
 TABLE No. 12. 
 
 Table No. 12 shows we are to take 33-8/9 per cent of 
 C iron. Then, of course we must take 66-1/9 per cent of 
 A iron. 
 
 As each of these percentages have to be divided into 
 two equal parts, we will do away with the fractions, and 
 call each one a whole number, which will save extra figuring 
 and will not affect the result any. By making them whole 
 numbers we have 34 per cent of C iron and 66 per cent 
 of A iron. As the C iron is composed of heavy scrap and 
 Salsbury pig, we must use 1 7 per cent of each, and of 
 course 33 per cent of each of steel and Cargo Fleet. In 
 checking them off at these rates, we find we have a shade 
 more silicon than we desire, brought about of course by 
 using all whole numbers instead of the fractions. 
 
 16 
 
Example : 
 
 17% of 1.50% silicon in Heavy scrap, equals 0.2550% 
 
 17% of 1.25% silicon in Salisbury scrap, equals 0.2193% 
 
 33% of 0.20% silicon in Steel scrap, equals 0.0660% 
 
 33% of 0.79% silicon in Cargo Fleet-pig, equals 0.2607% 
 
 100 Total silicon equals 0.8010% 
 
 Loss in melting 0.20 
 
 0.6010% 
 TABLE No. 13 
 
 17% of 4000 pounds equals 680 pounds Heavy scrap. 
 
 1 7% of 4000 pounds equals 680 pounds Salisbury pig. 
 
 33% of 4000 pounds equals 1320 pounds Steel scrap. 
 
 33% of 4000 pounds equals 1320 pounds Cargo Fleet-pig. 
 
 Charge of 4000 pounds 
 TABLE No. 14 
 
 To multiply one percentage by another percentage, see 
 percentage. 
 
 In shop practice when this method is understood, we 
 would require only tables 12 and 14, saving the figuring of 
 table 13, which we know would be correct. 
 
 The other elements can be figured the same way as 
 silicon. The manganese which would be low for a casting 
 of this kind, could be corrected with ferro-manganese, as 
 explained in following mixtures. 
 
 17 
 
METHOD FOR CORRECTING MIXTURES WITH FERRO- 
 SILICON OR FERRO-MANGANESE. 
 
 This method is useful when we wish to add more silicon 
 or manganese, as the case may be, to a mixture already 
 figured. 
 
 We wish to make a mixture of 2000 pounds for medium 
 floor work containing 2.2 per cent silicon. We have 1000 
 pounds of foundry scrap which we know contains 2 per cent 
 silicon and 1000 pounds of heavy scrap containing 1.5 per 
 cent silicon. As these two lots of iron are the amount we 
 require in our mixture, we will see how much silicon they 
 will bring into it. 50 per cent of 2 per cent silicon in 
 foundry scrap gives us 1.0 per cent, and 50% of 1.5% silicon 
 in heavy scrap gives us 0.75% more, making a totoal of 1.75 
 per cent silicon, leaving 0.45 per cent more to be 
 supplied by the ferro-silicon to make our mixture contain 2.2 
 per cent. This 0.45 per cent is what we want, and must go 
 down under B. The ferro-silicon containing 80 per cent 
 silicon must be put down under C, and as we are working 
 with one iron only, we will put a cipher under A. Setting 
 them down in that order, they stand ready to be figured as 
 in table 7. 
 
 18 
 
Example : 
 
 A. 
 
 Take A from B 
 
 1 st Remainder 
 Add both 
 
 Divided by 
 
 
 
 B. 
 
 .45 
 
 
 C. 
 
 80.00 
 
 .45 Take B from C. 
 
 .45 
 79.55 
 
 79.55 2nd Remainder. 
 
 80.)45.0000(.5625% of C Per. sil. 
 400 
 
 500 
 480 
 
 200 
 160 
 
 400 
 400 
 TABLE No. 15 
 
 2000 Pounds 
 00.5625 
 
 11.250000 Pounds 
 TABLE No. 16 
 
 11.25 Pounds 
 .80 
 
 20)9.0000(.45% 
 80 
 
 100 
 100 
 
 TABLE No. 1 7 
 
 19 
 
80 
 00.5625 
 
 .450000% of silicon from Ferro Silicon 
 1 .75 % of silicon from scrap iron 
 
 2.20 Total silicon. 
 
 TABLE No. 18 
 
 Table 15 shows we are to add 0.5625 per cent of ferro 
 silicon to the mixture. To find the amount of ferro-silicon 
 in pounds we are to use, we will multiply the 2000 pounds of 
 iron by .5625% making it 11.25 Ibs. Multiply 11.25 pounds 
 by the per cent of silicon it contains (which is 80,) and 
 dividing the result by 20, the number of hundred pounds in 
 the mixture will give us our required silicon 0.45 per cent. 
 Or, multiply the 80 per cent ferro-silicon by the percentage 
 we are to take, .5625, will also give us our required 0.45 
 per cent. See tables 16, 17 and 18 above. I have used 
 80 per cent ferro-silicon, it serves our purpose as well as 
 50% or any other per cent. 
 
 20 
 
THREE IRON SEMI-STEEL MIXTURES WITH APPROXI- 
 MATE FIGURING OF ALL THE ELEMENTS. 
 
 This mixture if melted hot and under proper conditions 
 would be suitable for marine piston rings, piston valve liners, 
 cut and cast gears, etc. Although we are using high steel in 
 this mixture it is advisable not to attempt high steel mixtures 
 without previous experience. In making a mixture of this 
 kind, there are always two elements we are sure of getting 
 exact. These are silicon and manganese, which will be 
 proved in this mixture. The other elements will be influenced 
 by these two, and can be made normal for the mixture, es- 
 pecially so, if we select irons suitable for the work in hand. 
 
 Keep the sulphur and phosphorus low. If the sulphur 
 is high, raise the manganese a point or two. We will make 
 a mixture of 2000 pounds to contain 
 Silicon Phos. Sulphur Mang. Total Carbon 
 1.8% 0.45% 0.07% 0.75% 
 
 From the following irons: 
 
 Buckeye 3.6% 0.55% 0.016% 0.50% 3.50% 
 
 Scrap iron 2.2 0.60 0.070 0.45 3.25 
 
 Scrap steel 0.2 0.05 0.050 0.50 0.08 
 
 TABLE No. 19. 
 
 As we are not particular what per cent of each iron 
 we use, the best way is to put the two lowest silicon's 
 together, and get their mean percentage. In this case we 
 will put scrap iron and scrap steel together. So adding 2.2 
 and 0.2 together making 2.4 per cent silicon, and dividing 
 by 2. will give us their mean per cent, 1.2. This gives 
 us now, practically, but two irons to figure on. Putting 
 them under their proper headings they stand ready to be 
 figured as in table 7. Adding 0.2 for loss of silicon in 
 melting will make our desired 2.0%. 
 
 21 
 
Take A from B 
 
 A. 
 
 1.2 
 
 1st] 
 
 nder 
 
 B. 
 
 2.0 
 1.2 
 
 .8 
 
 C. 
 
 3.6 
 
 2.0 
 
 Take B from C. 
 
 1 .6 2nd remainder 
 
 TABLE No. 20 
 
 When one remainder is just twice as much as the other, 
 it shows we are to take 33 Vz per cent of the C iron, and 
 66/ / 3 per cent of the A, and no more figuring are required. 
 If both remainders should come the same it would mean 50 
 per cent of each A and C, and if one remainder should 
 happen to come three times as much as the other, we would 
 have to take 75 per cent of one, and 25 per cent of the 
 other. In this case we are to take 33^ per cent of C and 
 66^ per cent of A iron. As the A iron is composed of scrap 
 iron and scrap steel, it means we are to use 33 Vz per cent of 
 each pig iron, scrap iron and steel. In checking off at these 
 percentages we get the following results: 
 
 Example 
 
 Silicon. 
 
 Pig iron 33J/3% of 3.6% silicon equals 1.200 % 
 
 Scrap iron 33J/3% of 2.2% silicon equals 0.733/ % 
 
 Scrap steel 33/ 3 % of 0.2% silicon equals 0.066^% 
 
 A Total silicon 2.000 % 
 
 Loss in melting equals 0.2 % 
 
 Silicon in casting 1 .8 % 
 Phosphorus. 
 
 Pig iron 331/3% of 0.55% phos., equals 0.1833/3 % 
 
 Scrap iron 33J/3% of 0.60% phos., equals 0.2000 
 
 Scrap steel 33/ 3 % of 0.05% phos., equals 0.01662/ 
 
 Phosphorus does not lose or gain in melting 
 
 22 
 
 .4000 % 
 
B 
 Pig iron 
 Scrap iron 
 Scrap steel 
 
 Sulphur. 
 33J/ 3 % of 0.016% sulp., equals 0.00533 Vz% 
 33|/3% of 0.07 % sulp., equals 0.02330/3 
 33]/ 3 %of0.05 %sulp., equals 0.0166 ^ 
 
 Gain in melting equals 
 
 .045241/3 
 .03 
 
 C Manganese. 
 
 Pig iron 33J/3% of 0.50% mang., equals 
 Scrap iron 33]/3% of 0.45 % mang., equals 
 Scrap steel 33^3% of 0.50% mang., equals 
 
 Loss in melting, equals 
 Manganese from 80% ferro-mang., equals 
 
 D Total Carbon. 
 
 Pig iron 33^3% of 3.50% carbon, equals 
 Scrap iron 33 J/3% of 3.25% carbon, equals 
 Scrap steel 33J/3% of 0.08% carbon, equals 
 
 E Total carbon 
 
 Pig iron 33|/ 3 % of 2000 Ibs., equals 
 Scrap iron 33J/ 3 % of 2000 Ibs., equals 
 Scrap steel 33J/3% of 2000 Ibs., equals 
 
 .07524/3% 
 
 0'.150 
 OA662/3 
 
 0.483/3 
 0.100 
 
 0.383 
 0.367 
 
 0.750 % 
 
 1.16662^% 
 1.0743/3 
 
 2.26762^% 
 
 6662/3 Ibs. 
 6662/3 Ibs. 
 6662/3 Ibs. 
 
 F 
 
 2000 Ibs. 
 
 TABLE No. 21. 
 
 This table shows that all the elements are nearly as we 
 want them, except the manganese, which of course can be 
 corrected to what we require with ferro-manganese. The 
 
 23 
 
sulphur is slightly higher, but the high manganese will be 
 liable to offset that much. As we desire 0.75 per cent 
 manganese in our mixture and our irons have given us only 
 0.383 per cent, it is evident we must get 0.367 per cent 
 from the ferro-manganese. 
 
 To find the number of pounds of ferro-manganese to 
 use, so as to get this 0.367 per cent manganese in the mix- 
 ture, we will set the 80 per cent under C, the desired 0.367 
 under B, and figure out as in table 15 on correcting mix- 
 tures or, using a shorter method which can always be done 
 when figuring a one iron mixture like this, we will simply affix 
 two ciphers to B, move the point two places to the right, 
 then divide it by C. Example: 
 
 TABLE No, 22. w* TABLE No. 23 
 
 A. B. C. <f* 80 
 
 0. .367 .80 .45% 
 
 80)36.700(.457/8% ^T~ 
 
 320 
 
 470 
 
 320 
 
 400 .3670% 
 
 70 
 
 -= 7 /8 
 
 80 
 
 TABLE No. 24 
 
 2000 pounds 
 457/8 
 
 1750 
 10000 
 8000 
 
 9.1750 pounds 
 3 24 
 
If you will look over Table 22, you will see we have 
 accomplished the same results as we did in table 15, with 
 considerable less figuring. The figures show we are to use 
 45%% of ferro-manganese. Multiplying 80 per cent man- 
 ganese by the .45%%, gives us our required 0.367%, as in 
 table 23, and to get the number of pounds of fero-manganese 
 we are to use to get this percentage of manganese, we mul- 
 tiply the full charge of 2000 pounds by the .45%%, which 
 shows we are to use, 9.175 pounds of ferro-manganese as in 
 table 24. By multiplying the 9.175 pounds by the 80%, 
 will give us the exact number of pounds of manganese, 
 we will get from the fero-manganese which is 7.34 pounds. 
 Now divide this 7.34 by 20, the number of 100 pounds in 
 the mixture, will again give us our required per cent, 0.367 
 of manganese, as in table 25. 
 
 Example : 
 
 9.175 
 80 
 20) 7.34000 (.367% manganese. 
 
 60 
 
 134 
 120 
 
 140 
 140 
 
 TABLE No. 25 
 
 Manganese from pig iron and scrap, equals 0.383% 
 
 Manganese from 80% Ferro-Manganese, equals 0.367% 
 
 Manganese in mixture 0.750% 
 
 In shop practice the per cent of D and F in table No. 
 21, and tables 22 and 24 is all we need figure. The other 
 tables are worked merely to prove the mixture. 
 
 25.. 
 
MIXTURE FOR LARGE CYLINDER LINERS, ETC. 
 
 Show Methods, if a Certain Per Cent of 
 
 Some of the Irons Are to be Used 
 
 We will make a mixture of 2000 pounds to contain 1.0 
 per cent silicon. We must use 500 pounds of steel scrap 
 containing 0.2 per cent silicon, and 500 pounds of liner 
 scrap containing 1.0 per cent silicon. We have besides some 
 scrap containing 1 .6 per cent silicon, and pig iron containing 
 2.2 per cent silicon. By adding the 0.2 per cent silicon to 
 make up for loss in melting, our mixture will have to con- 
 tain 1.2 per cent silicon. The first thing to do when making 
 a mixture with a given per cent of some of the irons is to 
 find how many pounds of silicon the mixture must contain. 
 Here we want 2000 pounds to contain 1 .2 per cent silicon, 
 multiplying the 2000 by the 1 .2 per cent gives us 24 pounds 
 of silicon. This is the amount we must get in this mixture. 
 The next is to find how much the given irons will contribute 
 to the 24 pounds, multiplying 500 pounds of steel by 0.2 
 per cent will give us 1 pound of silicon. The 500 pounds of 
 liner scrap containing 1 .0 per cent will contribute 5 pounds 
 more. This makes 6 pounds, leaving 18 pounds more for the 
 other 1000 pounds of iron to bring in. Now, if we had 
 some 1 .8 per cent silicon iron, 1 000 pounds of that would 
 just make our mixture complete, by giving us the 18 pounds 
 of silicon we require. And of course no more mixing would 
 be necessary, but we have only 1 .6 per cent scrap, and the 
 2.2 per cent pig iron. So we must find how much of each of 
 these we must use to complete the mixture. As we want 
 1000 pounds more iron, and 18 pounds of it must be silicon, 
 that means it must contain 1.8 per cent silicon, which of 
 course is our desired silicon, and must be put under B, set- 
 ting them down under their proper headings, they stand ready 
 to be figured as in table 1. Example: 
 
 26 
 
A. B. C. 
 
 -1.6 1.8 2.2 
 Take A from B 1.6 1.8 Take B from C. 
 
 1st Remainder .2 .4 2nd Remainder. 
 
 TABLE No. 26. 
 
 As we have shown in previous mixtures, when one 
 remainder is as much again as the other, it means we are 
 to take 33J/3% of the iron the smallest remainder represents, 
 (which is C), and 66^/3% of the iron that the largest rep- 
 resents which is A. As we only want 1000 pounds more 
 iron to complete the mixture this means we are 
 to take 3331/3 pounds of pig iron, 666^/3 pounds of 
 scrap iron. By taking the 500 pounds of steel, 500 pounds 
 of liner scrap, 333J/3 pounds of pig iron, 666^/3 pounds 
 of scrap iron, and multiply each one by the per cent of 
 silicon it contains, will give us the 24 pounds of silicon we 
 require in the mixture. Example: 
 
 Steel 500 pounds x 0.2% equals 1 Ib. silicon 
 
 Liner scrap 500 pounds x 1 .0% equals 5 Ib. silicon 
 Pig iron 333J/3 pounds x 2.2% equals 7J/3 Ib. silicon 
 
 Scrap 666^/3 pounds x 1 .6% equals 10^/3 Ib. silicon 
 
 20)24.0(1. 2% sil. 
 20 
 
 40 
 40 
 
 TABLE No. 27. 
 
 By dividing the 24 pounds by the number of 100 pounds 
 in the mixture, (which is 20), gives us our required 1.2 per 
 cent silicon. Here is another way to check it off, but as 
 we were only figuring for half the mixture in table 26, we 
 must take only half of each percentage thus obtained, mak- 
 
 27 
 
ing it 162/3% of pig, or "C" iron, and 33J/3% of the scrap, 
 
 or "A" iron. Example: 
 
 Steel 25% of 0.2% silicon equals 0.050% silicon 
 
 Liner scrap 25% of 1.0% silicon equals 
 
 Pig iron 162/3% of 2.2% silicon equals 
 
 Scrap 33J/3 % of 1.6% silicon equals 
 
 0.250% silicon 
 0.366^ silicon 
 0.533% silicon 
 
 1.200% silicon 
 
 TABLE No. 28. 
 
 By using this same percentage, the other elements if 
 known, can be figured as in table No. 21. And the Man- 
 ganese corrected as in tables 22, 23 and 24. 
 
 Any number of different grades of iron can be mixed 
 this way. Always leaving two irons, one with lower and 
 one with a higher silicon content than we desire to corect 
 the mixture with. 
 
 28 
 
METHOD OF FIGURING WHEN A CERTAIN PER CENT 
 OF STEEL MUST BE USED. 
 
 This mixture would be suitable for heavy gas and 
 hydraulic cylinders and other castings that require strength 
 and close grained enough to stand water pressure. We wish 
 a 25 per cent steel mixture of 2000 Ibs., to contain 1 .6 per 
 cent silicon, and 0.75 per cent manganese. We will use 12 
 per cent manganese scrap steel to corect the manganese 
 with, we will use the following irons: 
 
 Silicon Phos. Sulp. Mang. 
 
 Pig iron 3.00% 0.5% 0.016% 0.5% 
 
 Scrap iron 1.8 0.5 0.08 0.4 
 
 Scrap steel 0.0 0.02 0.05 0.5 
 
 TABLE No. 29 
 
 By adding 0.2 per cent silicon, and 0.1 per cent man- 
 ganese for loss while melting, will make our required silicon 
 1.8 per cent, and the manganese 0.85 per cent. When mak- 
 ing a mixture of this kind, I figure to get all the silicon from 
 the iron, because of the small amount of silicon in the steel. 
 On account of having to get all the silicon from the 1500 
 pounds of iron, that of course changes our desired silicon 
 for the present, because the 1500 pounds of iron will have 
 to carry enough silicon to give 1 .8 per cent to the full 
 charge of 2000 pounds of both iron and steel. To get this 
 new required per cent of silicon, we will divide the percent- 
 age of silicon required in the whole charge, by the per cent 
 of the iron used, which is 75 per cent. Please take note 
 of this rule. Example: 
 
 TABLE No. 30 TABLE No. 31 
 
 .75)1.800(2.4% silicon. A. B. C. 
 
 150 1.8 2.4 3.0 
 
 1.8 2.4 
 
 300 
 
 300 .6 .6 
 
 29 
 
Table No. 30 shows our new required silicon must be 
 2.4 per cent, which means we are take enough of the pig 
 iron and scrap to give us that amount and, according to 
 table 31 we must use 50 per cent of each. As 'we only 
 require 1500, that means we are to take 750 pounds of 
 each A and C, together with the 500 pounds of steel and 
 no more figuring is required, except of course for the man- 
 ganese. Example : 
 
 As we have 25 per cent steel, the other 75 per cent 
 must be divided between A and C. 
 
 Steel 25% of 0.0% silicon equals 0.0 % silicon 
 
 Pig 37.5% of 3.0% silicon equals 1.125% silicon 
 
 Scrap 37.5% of 1.8% silicon equals 0.675% silicon 
 
 1.800% silicon 
 TABLE No. 32. 
 
 Table 32 shows our figures are correct and is a much 
 shorter method than table 27. We will now figure the man- 
 ganese. Taking the same percentage as in table 32 we will 
 see how much manganese the irons figured have already 
 brought into the mixture. Example: 
 Steel 25% of 0.5% equals 0.125 % manganese 
 
 Pig 37.5% of 0.5% equals 0.1875% manganese 
 
 Scrap 37.5% of 0.4% equals 0.150 % manganese 
 
 0.4625% 
 TABLE No. 33. 
 
 Table 33 shows our mixture already has 0.4625 per cent 
 manganese. As we desire 0.85 per cent we must get the 
 other 0.3875 per cent from the 12 per cent manganese steel 
 scrap. This 0.3875 per cent of course is what we desire, 
 and must be put down under B. The 12 per cent man- 
 ganese under C, and figured as in table 22. Example: 
 
 By using the figure 3 in table 34, we save a lot of 
 figures and it does not affect the result. 
 
 30 
 
A. B. C. 
 
 0. .3875 12 
 
 12)38.7500(3.23% 
 36 
 
 27 
 24 
 
 35 
 36 
 
 TABLE No. 34. 
 
 Table 34 shows we are to take 3.23 per cent of C. 
 To find how much steel scrap we are to use, we will multiply 
 the full charge of 2000 pounds by 3.23 per cent, which 
 gives us 64.6 pounds. Multiplying the 64.6 pounds by the 
 per cent of manganese it contains, which is 12, gives us 
 the exact amount of manganese this 64.6 pounds adds, 
 which is 7.752 pounds. This again divided by 20, the 
 number of 100 pounds in the mixture, will give us our 
 required per cent of manganese. Multiplying the 12 by 
 3.23 per cent, will also give us our required per cent of 
 manganese. Example: 
 
 2000 20)7.7520 
 03.23 
 
 64.6000 Ibs. 
 
 .3876% man. 
 
 64.6 12 
 
 .12 03.23 
 
 7.7752 Ibs. .3876% 
 
 Manganese from mixture 0.4625 
 
 Manganese from 12% steel 0.3876 
 
 Total manganese equals 0.8501% 
 31 
 
TABLE No. 35 
 
 Using the figure 3 in table 34, altered the result bu 
 slightly, and saved a lot of figures. To use this 64.6 Ibs 
 of 12 per cent manganese steel, we would take out tha 
 amount from the 500 pounds of the common steel scrap. 
 
 METHOD OF FIGURING THREE OR MORE ELEMENTS 
 EXACT IN THE SAME MIXTURE 
 
 We have shown in previous mixtures how to get the 
 exact percentage of both silicon and manganese in the same 
 mixture. The silicon is figured correct, and the manganese 
 is corrected by the use of ferro-manganese. But now, sup- 
 pose we have to get another element exact, say phosphorus 
 to specification? 
 
 The best way to do it is to make two mixtures, both 
 to contain the same per cent of silicon we desire in the final 
 mixture, but one mixture to contain a lower and the other 
 to contain a higher per cent of phosphorus than we desire in 
 our final mixture. We then take the phosphorus contained 
 in each of these two mixtures to figure the exact percentage 
 of phosphorus we desire in our final mixture. So you see, 
 when we have two mixtures, each containing the same per 
 cent of silicon, no matter how much we use of each one 
 to get our desired per cent of phosphorus in the final mix- 
 ture, the silicon will not be changed. Example: 
 
 We wish to make a mixture of 2000 pounds to contain 
 2.3 per cent silicon, 0.65 per cent phosphorus, and 0.75 per 
 cent manganese. As we lose from 0.10 to 0.15 per cent 
 in melting, we will make our mixture to contain 0.9 per 
 cent manganese. 
 
 Silicon Phosphorus Sulphur Manganese 
 
 2.00% 0.4% 0.04% 0.60% 
 
 3.00% 0.7% 0.02% 0.75% 
 
 2.25% 0.6% 0.03% 0.65% 
 
 2.75% 0.9% 0.01% 1.00% 
 
 32 
 
TABLE No. 37. 
 
 When mixing for Phosphorus no allowance is made for 
 gain or loss in melting. Far silicon we will add the 0.2 per 
 cent, making our desired silicon for the mixture 2.5 per cent. 
 
 We will make our first mixture from the first two irons, 
 and our second mixture from the next two. You will notice 
 we have tried to select two irons that will give us a lower 
 and two that will give us a higher per cent of phosphorus 
 than we want in our final mixture. 
 
 Setting the first two irons under their proper heading 
 they stand ready to be figured. 
 
 First Mixture Second Mixture 
 
 Silicon Silicon 
 
 A. B. C. A. B. C. 
 
 2.0 2.5 3.0 2.25 2.50 2.75 
 
 2.0 2.5 2.25 2.50 
 
 .5 .5 .25 .25 
 
 TABLE No. 38. 
 
 In both of these mixtures we get even remainders, which 
 shows we are to take 50 per cent of all the irons. 
 
 In our first mixture we get the lowest phosphorus 0.55 
 per cent, as 50 per cent of 0.4 per cent equals 0.20 per 
 cent and 50 per cent of 0.7 per cent equals 0.35 per cent, 
 adding these two together we get 0.55 per cent phosphorus, 
 which is lower then we desire in the final mixture, but we 
 get our desired silicon 2.5 per cent. 
 
 In the second mixture taking 50 per cent of 0.6 per cent 
 phosphorus equals 0.3 per cent and 50 per cent of 0.9 per 
 cent equals 0.45 per cent. Adding these together we get 0.75 
 per cent phosphorus which is higher than we require in the 
 final mixture. But, we also have the same silicon (2.5 per 
 cent) in both mixtures. Now we have two mixtures, both 
 containing the same per cent silicon, but one has a higher 
 and the other has a lower per cent of phosphorus than we 
 
 33 
 
want in the final mixture. So we will take these two per 
 cents of phosphorus with our desired per cent and set them 
 under their proper heading and figure as in table No. 1. 
 Example : 
 
 A. B. C. Phosphorus 
 
 .55 .65 .75 
 .55 .65 
 
 .10 
 
 .10 
 
 TABLE No. 39 
 
 As both remainders are the same again, it shows we 
 are to take 50 per cent of each mixture, which will give 
 us our required per cent of phosphorus and the required per 
 cent of silicon in the final mixture. 
 
 As we are to take 50 per cent of each of the first mix- 
 tures, and each mixture contains two irons, it is apparent 
 that we are to take 25 per cent of each iron. 
 25% of 2.00% silicon equals 
 25% of 3.00% silicon equals 
 25% of 2.25% silicon equals 
 25% of 2.75% silicon equals 
 
 Example : 
 0.50 % 
 0.75 % 
 0.5625% 
 0.6875% 
 
 2.5000% 
 
 Silicon loss in melting 0.2 % 
 
 Silicon in castings 2.3 % 
 
 25% of 0.4% phosphorus equals 0.100% 
 
 25% of 0.7% phosphorus equals 0.175% 
 
 25% of 0.6% phosphorus equals 0.150% 
 
 25% of 0.9% phosphorus equals 0.225% 
 
 Total phosphorus 0.650% 
 
 25% of 0.6 % Manganese equals 0.150 % 
 
 25% of 0.75% Manganese equals 0.1875% 
 
 25% of 0.65% Manganese equals 0.1625% 
 
 25% of 1.00% Manganese equals 0.25 % 
 
 0.7500% 
 34 
 
TABLE No. 40 
 
 Our mixture gives us 0.75 per cent Manganese, leaving 
 0.15 per cent for the fero-manganese to bring in. As this 
 0.15 per cent is what we require, we will set it under B. 
 The 80 per cent ferro-manganese under C and figure as in 
 tables 22, 23, 24 and 25. 
 
 A. B. C. 
 0.15 80 
 
 Affix two ciphers to B, move the decimal point two 
 places to the right and divide by .80. Example: 
 
 80)I5.0000(.1875% of C. Ferro-Manganese. 
 80 
 
 700 
 640 
 
 600 
 560 
 
 400 
 400 
 
 2000 Ibs. 80 
 
 .001875 .001875 
 
 3.750000 Ibs. Ferro-Mang. .1 50000% 
 
 TABLE No. 41 
 
 Our figures show we are to take 00.1875 per cent of 
 C ferro-manganese, by multiplying the 80 per cent ferro- 
 manganese, by the 00.1875 per cent will give us our required 
 manganese. And by multiplying the 2000 Ib. charge by the 
 percentage of ferro-manganese 00.1875 we are to use, will 
 give us the amount of fero-manganese in pounds we 
 are to use. 
 
 Manganese from mixture equals 0.75% 
 
 Manganese from ferro-man. equals 0.15% 
 
 O90~~ 
 
 Loss in melting 0.15 
 
 Total manganese equals 0.75% 
 
 35 
 
FRENCH SPECIFICATIONS FOR SHELLS OF 122 TO 155 
 MILLIMETERS CALIBER TO BE CAST IN SAND 
 
 By Edgar Allen Custer in "The Foundry" 
 
 Silicon Phos. Sulphur Mang. C. Carb. G.Carb. 
 1.2% 0.15% 0.08% 0.70% 0.70% 2.40% 
 
 The above analysis are for dry sand molds, if cast in 
 green sand, the silicon should be about 1 .35 per cent. The 
 total carbon and silicon must not exceed 4.7 per cent. If 
 this limit is exceeded, the iron will lack toughness, at least 
 20 per cent of the total carbon must be combined to produce 
 proper fragmentation. The percentage of dust increases as 
 the combined carbon decreases. The charge should be as 
 follows: Pig iron 40 per cent, scrap 40 per cent and 
 steel 20 per cent. The term scrap is used to denote scrap 
 melted, pigged and charged according to analysis. All the 
 foundries in France engaged in this work have been mobi- 
 lized on a common basis, and are using precisely the same 
 methods of selection, analysis and general foundry procedure. 
 This has not been done without enormous losses and vex- 
 atious delays. There have been many cases where the 
 loss of a total heat has been reported, and the loss of 40 
 per cent was not uncommon in the first stages. Team work, 
 scientific methods and keeping everlastingly at it, have 
 brought results. Today, September 1917, the output has 
 reached staggering proportions, over 1 ,000,000 rounds per 
 day are being made. 
 
 This must certainly be interesting to every metal mixer, 
 and should have a tendency to induce him to try his hand 
 at making mixtures for shells, so as to be prepared, to some 
 extent, for any emergency. 
 
 We will make a mixture as near as possible to the 
 French specifications, from some Iron Mountain pig iron 
 which I recently had analyzed, and some scrap we will 
 presume contains the following analysis after it has been 
 melted and pigged. 
 
 36 
 
Sil. Phos. Sul. Mang. C. C. G. C. 
 
 Iron Mountain 1.4% 0.14 0.011 1.22 0.60 2.70 
 
 Selected scrap 2.0 0.40 0.080 0.55 0.40 3.00 
 
 Steel scrap 0.2 0.01 0.040 0.50 0.10 
 
 TABLE No. 42 
 
 In making this mixture we will use the silicon in the 
 steel, although as a rule I leave it out when making a 
 mixture to contain a certain percentage of steel. According 
 to the specifications our mixture must contain 1 .2 per cent 
 silicon, adding 0.2 per cent for loss in melting, will make 
 our desired silicon 1 .4 per cent. As we are to use 20 per 
 cent steel, we will get 0.04 per cent from it, leaving 1 .36 
 per cent for the pig iron and scrap to bring into the mixture. 
 Now, as we only want 80 per cent more iron, and this 80 
 per cent will have to carry enough silicon to give us 1 .36 
 per cent for the whole mixture of 2000 pounds, that means 
 we are to find a new temporary per cent of silicon to work 
 with. So by using the same rule as in table 30 that is by 
 dividing the actual per cent of silicon desired by the per- 
 centage of iron used in the mixture, which in this case is 
 80 per cent, we get the new per cent of silicon, 1 .7 to 
 work with, see the point. We must get 1 .7 per cent 
 silicon in 80 per cent of the mixture to give us 1 .36 per 
 cent more silicon to the whole mixture. Example: 
 
 TABLE No. 43. TABLE No. 44 
 
 .80)1.360(1.7% silicon. A. B. C. 
 
 80 1.4 1.7 2.0 
 
 1.4 1.7 
 
 560 
 
 560 3 .3 
 
 The result of table 44 shows we are to use the same 
 percentage of each pig iron and scrap, which in this case 
 is 40 per cent, with 20 per cent of steel. So figuring all 
 the elements on that basis, we will see how near our mix- 
 ture is to the specifications. Example: 
 
 37 
 
Iron Mountain 
 Selected scrap 
 Steel scrap 
 
 40% of 1.4% silicon equals 
 40% of 2.0% silicon equals 
 20% of 0.2% silicon equals 
 
 Loss in melting 
 Silicon in mixture 
 Phosphorus. 
 
 0.560% 
 0.800% 
 0.040% 
 
 1.400% 
 0.2 % 
 
 1.2 % 
 
 Iron Mountain 40% of 0.14% phosphorus equals 0.056% 
 Selected scrap 40% of 0.40% phosphorus equals 0.160% 
 Steel scrap 20% of 0.01% phosphorus equals 0.002% 
 
 Phosphorus in mixture 0.218% 
 
 Sulphur. 
 
 Iron Mountain 40% of 0.011% sulphur equals 0.0044% 
 Selected scrap 40% of 0.080% sulphur equals 0.0320% 
 Steel scrap 20% of 0.040% sulphur equals 0.0080% 
 
 Gain in melting 
 
 Sulphur in mixture 
 Manganese. 
 
 Iron Mountain 40% of 1.22% manganese equals 0.488% 
 Selected scrap 40% of 0.55% manganese equals 0.220% 
 Steel scrap 20% of 0.50% manganese equals 0.100% 
 
 Loss in Melting 
 Manganese in mixture 
 38 
 
Combined Carbon 
 
 Iron Mountain 40% of 0.6% C. C. equals 0.240% 
 
 Selected scrap 40% of 0.4% C. C. equals 0.160% 
 
 Steel scrap 20% of 0.1% C. C. equals 0.020% 
 
 0.420%, 
 Graphite Carbon 
 
 Iron Mountain 40% of 2.7% G. C. equals 1 .08% 
 
 Selected scrap 40% of 3.0% G. C. equals 1 .20% 
 
 Steel scrap 20% of 0.0% G. C. equals 0.00% 
 
 2.28% 
 TABLE No. 45. 
 
 These tables show that all the elements are very nearly 
 what the specifications call for. Even though phosphorus is 
 a little higher here, there is not the least doubt it would be 
 lower in actual practice, even from this mixture. If is was 
 not, we would use two grades of pig iron, and melt and 
 pig two different grades of scrap, and get our phosphorus 
 exact, by the same method as in table 38 and 39. The 
 carbons we cannot tell very much about till after analysis 
 has been made from the mixture, because it is a semi-steel 
 mixture. But both carbons will be well within specifications, 
 which says the combined should be at least 20 per cent 
 of the total carbon. As this mixture shows over 15 per 
 cent, it is bound to be higher in the casting on account 
 of the low silicon, and high steel in the mixture. Successful 
 mixtures of this kind are not accomplished with one trial. 
 And like the French foundrymen, only sticking everlastingly 
 at it, would we accomplish the desired results. 
 
 39 
 
SIDE LIGHTS ON MIXTURES 
 
 In making some mixtures you will find when dividing 
 your first remainder, that to get the exact result, you would 
 be compelled to carry it out to several decimal places. Now, 
 if it will not finish with one decimal place, just raise the 
 last decimal or figure in the quotient, one more. Although 
 the divisor will not go that many times, still it will save 
 a lot of figures, and will not affect the result any. But, 
 be sure and do this with the first remainder only, then you 
 will always have the full percentage of the element you are 
 figuring for; Example: 
 
 Suppose we wish a mixture of 1500 pounds to contain 
 2.2 per cent silicon. We will make it from 1 .8 per cent 
 silicon scrap, and 50 per cent ferro-silicon. 
 
 TABLE No. 46 
 
 A. B. C. 
 
 1.8 2.2 50.0 
 
 1.8 2.2 
 
 .4 47.8 
 47.8 
 
 48.2)40.000( .83% of C iron 
 3856 99. 17% of A iron 
 
 1440 
 1446 
 
 As there are two more decimal places in the dividend 
 than in the divisor, we point off two decimal places in the 
 quotient making it .85 1 jhtindredths per cent of C iron to be 
 used, and 99.17 per cent of A iron to be used. 
 A iron 1.8% silicon C iron 50% silicon 
 
 Take 99. 1 7% of A iron Take .0083% of C iron 
 
 silicon 4 , 50% si]icon 
 
 .4150 
 
 2.20006% silicon 
 
 ^40 
 
You will notice, by using the figure 3 in table 46 did 
 
 not make any material difference to the result. But saved 
 
 carrying the quotient to several decimal places. 
 
 We will try another from 3.25 per cent pig iron, instead 
 of ferro-silicon. 
 
 A. B. C. 
 
 1.8 2.2 3.25 
 
 1.8 2.2 
 
 .4 1.05 
 
 1.05 
 
 1.45)40.00(28% of C iron. 
 290 72% of A iron. 
 
 1100 
 1160 
 
 TABLE No. 47. 
 
 A iron 1.8% silicon C iron 3.25% silicon 
 
 Take 72% of A iron Take 28% of C iron 
 
 1.296 .9100 
 
 1.296 
 
 2.2060% silicon 
 
 Even by having too much by 60 in table 47 only added 
 6 thousandths of one per cent to the silicon, but saved quite 
 a lot of figures. 
 
 By taking advantage of this idea when you are dividing 
 your first remainder, if the figures are inclined to run to 
 several decimal places, you will always get the full percent- 
 age of silicon, or any other element you may be figuring 
 for. Table 46 says we are to take. .83 hundredths of one 
 per cent of the C iron and 99.17 per cent of A iron. 
 
 41 
 
Example: 
 Charge 1500 pounds 
 
 00.83% of C iron 
 
 1 500 pounds 
 99.17% of A iron 
 
 4500 
 12000 
 
 1487.5500 Ibs. of A iron 
 12.4500 
 
 12.4500 Ibs. of C iron 
 
 1500.0000 pounds 
 
 Table 47 says we are to take 28 per cent of C iron and 
 72 per cent of A iron. Example: 
 
 Charge 
 
 1500 pounds 
 .28% of C iron 
 
 12000 
 3000 
 
 420.00 Ibs. C of iron 
 
 1500 pounds 
 .72% of A iron 
 
 3000 
 10500 
 
 1080.00 Ibs. A iron 
 420.00 
 
 1500.00 pounds. 
 
 42 
 
MISCELLANEOUS MIXTURES 
 
 These mixtures are taken from my note books and was 
 cast several years ago from the following irons. The cast- 
 ings answered their purpose and finished up clean. You 
 will notice the steel mixtures are made from irons low in 
 sulphur and phosphorus with manganese from 0.75 per 
 cent up. 
 
 Piston Valve Liners 
 
 70% Carron No. 1 ; Tranverse 2800 per sq. inch. 
 30% steel scrap ; Silicon 1 .6 per cent in casting. 
 
 Hammer Block 
 
 60% Foundry scrap; Tranverse 3600. 
 40% Steel scrap; Silicon 1.16 per cent. 
 
 Stamp Heads 
 
 68% Shop scrap; Tranverse 3900. 
 32% Steel scrap; Silicon 1.09 per cent. 
 
 V Gear 8' 6" Dia. 9" Face, Hub Split. 
 74% cyl. Niagara; Silicon 1.2 per cent. 
 26% Steel scrap. 
 
 Large Marine Cylinder. Net Weight 34,020 Ibs. 
 30% Gun iron; Silicon 1.7 per cent in casting. 
 30% Cyl. Niagara. 
 30% Shop scrap. 
 10% Carron No. 1. 
 
 McCully Crusher No. 7 
 
 85% Texada No. 2; Test piece Chilled 2% deep. 
 15% Steel scrap; Silicon 0.85 per cent. 
 
 90-inch Snap and Bull Rings 
 
 67% Carron No. 1 ; Silicon 1 .6 per cent in casting. 
 33% Steel scrap. 
 
 43 
 
Two Marine Cyl-Liners, 6927 and 7134 Ibs. Net 
 
 56.25% Gun iron; Silicon estimated 1.2%. 
 
 25.00% Carrqn; 1st liner cast 1.07%. 
 
 18.75% Steel scrap; 2nd liner cast 0.97% silicon. * 
 
 16.000 pounds. 
 
 Good For Strong Castings and Semi-Steel 
 
 Brand Sil. Phos. Sul. Mang. C.C. F.C. G.C. 
 
 Carron 2.8 0.50 0.035 1.45 3.64 
 
 Texada No. 2 1 .25 0.30 0.025 0.90 
 
 Cyl. Niagara 1.80 0.50 0.044 0.75 
 
 Car Wheels 0.70 0.40 0.16 0.50 0.9 2.90 
 
 Gun iron 1.25 0.31 0.070 0.60 
 
 Iron Mountain 1.400.14 0.011 1.22 0.6 2.70 
 
 Irondale 2.300.16 0.035 1.10 
 
 Niagara No. 2 2.20 0.40 0.04 0.60 3.58 
 
 Muirkirk 2.21 0.28 0.031 2.22 0.55 3.01 
 
 Good for Soft Iron Work 
 
 Sloss No. 1 3.60 0.65 0.03 0.45 
 
 Crown No. 1 3.25 0.71 0.022 0.50 
 
 Clifton No. 1 3.500.50 0.015 1.40 0.3 3.30 
 
 Mississippi 3.34 0.294 0.022 0.90 
 
 Grading numbers will correspond closely to the follow- 
 ing percentages of silicon and sulphur. 
 
 No. 1 Pig No. 2. No. 3. No. 4. 
 
 Silicon 2.75 to 3.50% 2.25 to 2.75 2.00 to 2.25 1 .75 to 2.00 
 Sulphur 0.02 to 0.04% 0.01 to 0.03 0.01 to 0.03 0.01 to 0.03 
 
 No. 5. No. 6. No. 7. No. 8. 
 
 Silicon 1 .50 to 1 .75% 1 .25 to 1 .50 1 .00 to 1 .25 0.75 to 1 .00 
 Sulphur 0.02 to 0.04% 0.02 to 0.04 0.03 to 0.04 0.03 to 0.05 
 
 The following analysis of a few of the most important 
 castings, will give the young student some idea to work on 
 while making mixtures. 
 
 44 
 
If from five to ten per cent of steel is mixed in these 
 mixtures, it will strengthen and improve the castings for 
 this class of work. 
 
 Silicon Phos. 
 
 Hydraulic Cylinders .5 .40 
 
 Amonia Cylinders .6 .60 
 
 Air Cylinders .3 .45 
 
 Steam Cylinders, Heavy .6 .40 
 
 Steam Cylinders, Small .9 .55 
 
 Gas Engine Cylinders .8 .50 
 
 Locomotive Cylinders .6 .55 
 
 Automobile Cylinders 2.2 .50 
 
 Propeller Wheels .4 .30 
 
 Bed Plates, Heavy .9 .55 
 
 Dynamo Frames 2.5 .80 
 
 Approximate rule for weighing pig iron in piles : 
 If piled in the usual way 7^4 cubic feet will weigh one 
 ton. If very closely piled 7 cubic feet will weigh one ton. 
 
 Sulp. 
 
 Mang. 
 
 .08 
 
 .8 
 
 .09 
 
 .7 
 
 .09 
 
 .8 
 
 .09 
 
 .8 
 
 .08 
 
 .6 
 
 .07 
 
 .7 
 
 .08 
 
 .6 
 
 .08 
 
 .7 
 
 .09 
 
 .8 
 
 .08 
 
 .6 
 
 .07 
 
 .5 
 
 45 
 
THE INFLUENCE DIFFERENT ELEMENTS HAVE 
 UPON THE IRON. 
 
 Silicon. 
 
 Silicon will soften the iron up to 3.50 per cent. When 
 iron contains more it begins to get hard, short and brittle. 
 Silicon increases fluidity, decreases shrinkage, open the grain 
 of the iron and helps to turn combined carbon into graphite 
 carbon, which helps to reduce the strength of the iron. In 
 melting we lose about 0.2 per cent of silicon, which amount 
 must be taken into account when figuring for silicon. 
 
 Phosphorus. 
 
 Phosphorus helps to make iron fluid and weak, so for 
 all kinds of castings except the very thinnest, it should not 
 be over 0.7 per cent. But for light, thin castings where 
 strength is of no importance, it can run as high as 1.0 or 
 1.25 per cent. In fact iron for stove plate and that line 
 of work require that much. Phosphorous lowers the melting 
 point of iron, and decreases the shrinkage. In melting it 
 neither loses or gains very much. So no provision for loss 
 or gain is required when making mixtures. 
 
 Sulphur 
 
 Sulphur if too high will make the iron hard. Increase 
 the shrinkage and promote chill, and cause the iron to con- 
 geal quickly. If very high it will cause blow holes, shrinkage 
 cracks and dirty iron. In all machinery castings it should 
 be kept below 0.9 per cent if possible. In melting it gains 
 about 0.03 to 0.035 per cent, chiefly from the fuel. This gain 
 must be taken into account when making mixtures. 
 
 Manganese. 
 
 Manganese is one of the best elements we have in iron. 
 It is a regular scavenger. There is no element that will 
 cleanse the iron, reduce the blow holes, reduce the sulphur, 
 
 46 
 
increase the strength and improve the grain like manganese. 
 When silicon is normal for the work being made, manganese 
 from 0.5 to 0.8 per cent will be alright. In melting we lose 
 from 0.10 to 0.15 per cent. 
 
 Graphite Carbon. 
 
 Graphite carbon is a softener. It opens the grain of 
 the iron, makes it soft, weaker and reduces shrinkage and 
 chill. 
 
 Combined Carbon 
 
 Combined carbon is a hardener. It closes the grain of 
 the iron, increases the strength, shrinkage and chill. In melt- 
 ing there is no gain or loss, only that one form will change 
 to the other according to the rate of cooling, and influence 
 of the other elements, especially silicon and manganese. In 
 the common, soft foundry pig irons, combined carbon will run 
 about 0.30 per cent to 0.50 per cent. Graphite carbon will 
 run about 3.0 per cent to 3.5 per cent. But as the iron is 
 made harder by mixing, the carbons will change, Graphite 
 carbon getting lower in per cent and combined carbon in- 
 creasing in percentage, according, of course, to the per cent 
 of silicon put into the mixture. Graphite carbon will be 
 high, when silicon is high, and combined carbon will increase 
 as silicon is lowered. 
 
 Approximate per cent of Silicon for Different Castings 
 
 I have found castings containing the following percent- 
 ages of silicon, were satisfactory, both in machining and use. 
 
 For light castings from % to one-inch in section. From 
 2.25 to 1 .9 per cent silicon. Castings from 1 inch 
 to 2 inches in section. From 1 .9 to 1 .6 per 
 cent silicon. Castings from 2 inches to 3 inches in 
 section, from 1 .6 to 1 .3 per cent silicon. These figures are 
 given to give the reader an idea how to regulate the silicon 
 for castings of different section, and if the other elements 
 are kept normal by selecting irons suitable for the class of 
 work being made, will be entirely satisfactory for all kinds 
 
 47 
 
of general machinery castings when figuring for silicon only. 
 For pulleys the silicon should range from 2.3 per cent to 2.6 
 per cent, and the sulphur should be kept below 0.06 per 
 cent if possible. For large Marine cylinders with brackets 
 and flanges liable to crack through unequal sections, the 
 silicon should run from 1 .6 per cent to 1 .8 per cent. The 
 castings of course, always have liners of much harder metal. 
 Small and medium sized cylinders with no liners should run 
 from 2.0 to 1.6 per cent silicon, with 10 to 15 per cent steel. 
 For large gear wheels, blank or otherwise should contain, 
 after 20 to 25 per cent steel has been added, about 1 .6 
 per cent silicon. Car wheels, from 10-inch mining wheels, 
 up to regular passenger car wheels, from 1.5 to 0.70 per 
 cent silicon. From 5 to 10 per cent steel scrap always helps 
 the chill and strength of the wheels. All car wheels should 
 be annealed as soon as possible after casting, by putting into 
 a pit altogether. 
 
 When selecting pig iron for small and medium castings, 
 try and get iron containing less than 0.03 per cent sulphur, 
 phosphorus about 0.7 per cent, Manganese about 0.6 per 
 cent or 0.8 per cent, with graphite carbon about 3.25 per cent 
 and combined carbon 0.25 per cent or under. In the castings 
 the sulphur will average about 0.08 per cent. The other ele- 
 ments will not vary very much. In the heavier castings the 
 sulphur should not exceed 0.095%, phosphorus should be 
 kept down from 0.4 to 0.5%, manganese about 0.8 to 0.9 
 per cent. Graphite carbon will run from 2.50 per cent to 
 2.75 and combined about 0.75 per cent. 
 
 Judging the percentage of silicon in Different Kinds of Scrap. 
 
 As a general rule light machinery scrap will contain 
 about 1.9 per cent to 2.25 per cent silicon. But sometimes 
 we run across heavy scrap that runs that high in silicon. In 
 that case, as a rule, the fracture will show a dark rough sur- 
 face, full of shining particles of graphite, whereas the low 
 silicon heavy scrap, will show a lightish, slightly rough frac- 
 ture. 
 
 48 
 
Heavy scrap from 1J/2 inches to 3 inches in section, will 
 run from 1 .8 per cent to 1 .25 per cent silicon. 
 
 Standard car wheels silicon 0.7 per cent, phosphorus 
 net over 0.4 per cent, manganese 0.4 to 0.5 per cent, sulphur 
 not over 0.17 per cent, graphite carbon from 2.5 to 2.9 per 
 cent, combined carbon not over 0.90 per cent. 
 
 Steel plate scrap contains silicon about 0.2 per cent, 
 phosphorus from 0.01 to 0.05 per cent, sulphur form 0.03 
 to 0.05 per cent, and manganese 0.5 per cent, with total 
 carbon about 0.10 per cent. 
 
 Stove plate scrap runs about 2.75 per cent silicon, phos- 
 phorus about 1 .0 per cent. Although stove plate scrap is 
 high in silicon, it is a quantity that cannot be depended 
 upon, on account of its thin section, both the iron and the 
 elements in it are burnt somewhat, especially so if melted 
 under high blast. Use it with judgment. If light and 
 heavy scrap are brought together, it will pay to sort, and 
 give each its proper rating, which with a little experience 
 can soon be learned. 
 
 DECIMAL FRACTIONS 
 
 In adding a few examples on decimal fractions and 
 percentage, I thought would be an advantage to those who 
 have allowed themselves to get rusty on decimals to have 
 under the same cover, as a ready reference while working 
 over the mixtures. 
 
 Addition of Decimals 
 
 The only respect in which addition of decimals differ 
 from simple addition is, in placing the decimal point directly 
 over one another. Example: 
 
 26.346 
 .263 
 
 26.609 
 
 49 
 
Substraction of Decimals 
 
 Substract as in whole numbers, but keep the decimal 
 points directly under each other, as in addition. Example: 
 80.312 
 79.200 
 
 1.112 
 
 Multiplication of Decimals 
 
 Multiply as in whole numbers, and point off in the 
 product as many decimal places as there are decimal places 
 in the two factors, and if the product has 1 not so many, 
 supply the defect by writing ciphers on the left hand. 
 Example : 
 
 1st 2nd 
 
 .33 32.3 
 
 .2 2.3 
 
 .066 969 
 
 646 
 
 74.29 
 
 Note: In the first example there are three decimal 
 places, so must make three decimal places in the product 
 by adding one cipher to the left hand of it. 
 
 Division of Decimals 
 
 Divide as in simple numbers and point off as many 
 decimal places in the quotient, as the number of decimal 
 places in the dividend exceeds the number in the divisor. 
 If necessary prefix ciphers to the quotient; or affix ciphers 
 to the dividend. When both dividend and divisor contam 
 the same number of decimal places, the quotient is a whole 
 number, without or with a remainder as the case may be. 
 
 50 
 
Example: 
 
 No. 1 Divide 60 by 1.5. 
 
 No. 2. Divide 34.75 by 2.5. 
 
 Divisor Dividend Quotient 
 
 1.5 ) 60.0 ( 40 
 
 600 
 
 2.5)34.75(13.9 
 25 
 
 97 
 75 
 
 225 
 225 
 
 In the first example the divisor has one decimal place, 
 but the dividend has none, so one must be affixed to it. 
 As the dividend must have as many, if not more, decimal 
 places as the divisor, with the added decimal place in the 
 dividend makes the quotient a whole number. In the 
 other example the dividend has one decimal place more than 
 the divisor, so we point off one in the quotient. 
 Example No. 3. Divide 30.5 by .9. 
 Example No. 4. Divide 70 by 11.2. 
 
 .9)30.5 (33-8/9 1 1 .2) 70.000(6.25 
 27 672 
 
 35 , 280 
 
 27 224 
 
 8 560 
 
 560 
 
 In the 3rd example we find we could not bring it to 
 an end, so to save carrying it on to several decimal places, 
 we have finished with a vulgar fraction, and as the dividend 
 has the same number of decimal places as the devisor, the 
 quotient is a whole number, with the fraction 8/9. 
 
 In the 4th example we had to add three more ciphers- 
 
 51 
 
to the dividend, giving it two more decimal places than the 
 divisor, so we point off two decimal places in the quotient. 
 
 Percentage 
 
 Percentage is the process of calculating by the hun- 
 dreths. Thus 5 per cent of a quantity is 5 of every hun- 
 dred, or 5 hundredths of the quantity. When multiplying 
 for a percentage of a certain number, the multiplier is ex- 
 pressed decimally. That is, if we are to take 5%, 25% and 
 12j/2% of a number, we would set them down to multiply 
 like this: .05.25 and .125. The following table will show 
 our meaning: 
 
 PER CENT DECIMAL PER CENT DECIMAL PER CENT DECIMAL 
 
 1% 
 
 2% 
 
 3.1% 
 
 10% 
 
 50% 
 
 .01 
 .02 
 .031 
 .10 
 .50 
 
 75% 
 100% 
 150% 
 500% 
 
 .75 
 1.00 
 1.50 
 5.00 
 .0025 
 
 1!/2% 
 8/3% 
 12/2% 
 
 .005 
 
 .0075 
 
 .015 
 
 .08J/3 
 
 .125 
 
 In the first place the base is the number on which the 
 percentage is computed. 
 
 Example: Suppose we wish to take 6J/4 per cent of 
 12.7 per cent. The 12.7 is the base, and the 6J/4 is the rate, 
 so multiplying the base by the rate decimally expressed we 
 get the percentage of .79375. 
 Example: 
 
 12.7 
 .0625 
 
 635 
 254 
 762 
 
 .79375% 
 
 As explained in the multiplication of decimals, we must 
 point off as many places in the product as there are in the 
 multiplier, and the multiplicand which is five. It will be 
 noticed that although we called 12.7 a per cent, it became 
 
 52 
 
a base as soon as we wished to take a percentage from 
 it. Examples. 
 
 2. Take 50% of 2.75%. 
 
 3. Take 30% of 3.25%. 
 
 4. Take 70% of 1.75%. 
 
 2.75 3.25 1.75 Base. 
 
 50 .30 .70 Rate 
 
 1.3750% .9750% 1.2250% Percentage 
 
 When two numbers are given and we wish to know 
 
 the rate of each one, we add the two together, and divide 
 
 each number, after affixing two ciphers and moving the 
 
 points two places to the right, by the sum of the two. 
 
 Example : 
 
 What per cent of 1.50 is .45 and 1.05? 
 
 .45 
 1.05 
 
 1.50)45.00(30% 
 
 4500 
 1.50)105.00(70% 
 
 10500 
 
 When multiplying for percentage with decimals, we 
 must always point off two extra in the result for the whole 
 numbers. Example: Suppose we wish to take .5625 per 
 cent of 80 per cent 80 
 
 .005625 
 
 You notice we have added two .450000 
 ciphers, which represent the two whole numbers, and 
 of course moves the decimal point two places to the left, 
 so in pointing off the result, we count six decimal places. Of 
 course in actual practice, we imagine the two whole numbers 
 are there, and point off the result accordingly. 
 
 Hoping these few suggestions will carry the point, we 
 will not go any deeper on this subject. 
 
 53 
 
CUPOLA PRACTICE 
 
 Although it is not the purpose of this book to treat on 
 cupola practice, I feel I could not conclude it without a 
 word or two. We may make our mixtures as they should be 
 made, still there is a possibility of them going wrong by im- 
 proper handling and charging of the cupola. 
 
 If every melter would take the trouble to find the 
 proper height the coke bed should be for his particular 
 cupola, then make all his charges of iron from first to last 
 as near the same weight as possible, he will get a more uni- 
 form grade and even flow of iron, with less coke consumption, 
 than the man who crowds his coke bed to the limit with an 
 extra heavy first charge of iron. The proper practice calls 
 for the same weight of charge on the bed as every succeed- 
 ing charge, and the weight of that charge is figured by the 
 weight of coke it takes to fill four inches high in the cupola. 
 Then use a ten to one ratio, that is if it takes 150 pounds of 
 coke to fill four inches high in the cupola, the iron charges 
 should be about 1500 pounds and so on. 
 
 Experimenting foundry men have proved the melting 
 zone averages from four to five inches in depth. And they 
 have also found that amount of fairly good coke will melt 
 10 times its weight in iron and when that amount of iron 
 is melted, the bed is then ready for another four inch layer 
 of coke. Now, when I speak of a four inch layer of coke, 
 I do not mean that we must put four inches all over the 
 inside area of the cupola. That rule is only used as a stan- 
 dard on which to figure our iron charges. It has been 
 proved that the best results have been derived by putting 
 all the coke in the center, and all the iron as close to the 
 lining as possible, excepting of course, when making different 
 mixtures which must be separated by coke. By this method 
 of charging, the coke can be reduced and still have hot iron 
 if the bed and first charge have been started right. When 
 
 54 
 
the bed is the right height it is only the top four inches 
 that does the real melting, so if the bed is higher than it 
 should be the extra coke will be burnt and wasted until 
 it lets the iron down to the real melting zone, which will 
 vary from 15 to 28 inches above the tuyers according to 
 high or low blast, so the main point is to find the proper 
 height of the bed for every cupola, and the best way to find 
 it is by the time it takes the iron to drop lively after the 
 blast is on. If it takes more than three minutes at the most 
 the bed is too high, and the extra time will be taken up 
 burning coke that is not required. Now then, it is generally 
 upon high coke beds that extra heavy first charges of iron 
 are put, because we are under the impression that so much 
 coke on the bed ought to melt a much heavier charge than 
 the rest of the charges. But it is a wrong impression. An- 
 other reason for heavy first charges are, that most foundries 
 have some special mixtures to make, different from their 
 regular run of work, and if they happen to be heavier than 
 their regular charges, the bed is considered the best place, 
 so as to get them down, and out of the way of the regular 
 mixtures. But, as we must put an heavier split of coke between 
 two different mixtures, the bed is built up somewhat, the 
 amount it has lost by having an extra heavy first charge 
 to melt, and I believe, that one reason of having to put 
 an heavier split of coke between two different mixtures, have 
 saved many a coke bed from getting dangerously low with- 
 out the melter being aware of it. Now here's the point: We 
 know, if we wish to retard the melting between two different 
 mixtures, we must put an heavy split of coke between them. 
 By so doing we keep the iron high above the melting zone, 
 until a part of the coke is burnt away, when the top part 
 or last four inches of the coke wil drop to a point where it 
 can melt the iron above it. 
 
 It is just the same with the high coke bed. It is only 
 the last, or top four inches that does the real melting, and 
 like the heavy split of coke, even that four inches will not 
 melt iron until it drops to the real melting zone, and then 
 it will only melt so miich< so if burdene'cj .with an extra heavy 
 
 '.55 ' 
 
first charge of iron, the bed proper is bound to suffer, and 
 can only be built up again at the expense of irregular iron, 
 and extra coke, which would not be, if all the charges had 
 been made as near as possible what they should be, accord- 
 ing to the size of the cupola. It is not very good cupola 
 practice to let the iron soak too long in the cupola before 
 starting the blast, as I believe the iron absorbs more or less 
 sulphur from the fuel during that time. It is a fact, that 
 "converter steel" men, if they have to make castings to 
 strict specifications, will not use first charges for that class of 
 work if they can help it. The reason, that sulphur always 
 runs higher in first charges than in the following charges. 
 Fairly good practice calls for the fire to be started about one 
 hour before charging. As charging will take from three 
 quarters to one hour, the blast should then be put on as 
 soon as possible. If the bed is the right height, the iron 
 will begin to drop within a minute or so, and will be droping 
 quite fast within three minutes, as can be seen through 
 tuyer glasses. 
 
 Although I have mentioned that the melting zone will 
 average about 20 inches above the tuyers, I do not mean 
 that to be the height of the bed. We may have to make 
 it 30 inches, or even more, because the bed will settle from 
 8 to 12 inches as soon as the first charge of iron is dumped 
 on it that point will have to be settled by the time it takes 
 the iron to begin to drop after the blast is put on, the melting 
 zone being located entirely by the force of the blast. If 
 the blast is high and strong, so will the melting zone be 
 located high, so make the coke bed rather high at first then 
 reduce till you find proper height by above instructions. 
 
 56 
 
UNIVERSITY OF CALIFORNIA LIBRARY 
 BERKELEY 
 
 THIS BOOK IS DUE ON THE LAST DATE 
 STAMPED BELOW 
 
 Books not returned on time are subject to a fine of 
 50c per volume after the third day overdue, increasing 
 to $1.00 per volume after the sixth day. Books not in 
 demand may be renewed if application is made before 
 expiration of .loan period. 
 
 21 
 
 50w-7,'16 
 
\7 A /~\ I 7 O^ 
 
 YA 01 /o2 
 
 305059 
 
 UMVERSITV OF CAUFORNIA LIBRARY