DEPARTMENT OF COMMERCE TECHNOLOGIC PAPERS OP THE BUREAU OF STANDARDS S. W. STRATTON. DIRECTOR No. 2O2 RESULTS OF A SURVEY OF ELEVATOR INTERLOCKS AND AN ANALYSIS OF ELEVATOR ACCIDENT STATISTICS BY C. E. OAKES, Electrical Engineer J. A. DICKINSON, Mechanical Engineer Bureau of Standards OCTOBER 17, 1921 PRICE, 5 CENTS Sold only by the Superintendent of Documents, Government Printing Office Washington, D. C. WASHINGTON GOVERNMENT PRINTING OFFICE 1921 DEPARTMENT OF COMMERCE TECHNOLOGIC PAPERS OF THE BUREAU OF STANDARDS S. W. STRATTON. DIRECTOR No. 2O2 RESULTS OF A SURVEY OF ELEVATOR INTERLOCKS AND AN ANALYSIS OF ELEVATOR ACCIDENT STATISTICS BY C. E. OAKES, Electrical Engineer J. A. DICKINSON, Mechanical Engineer Bureau of Standards OCTOBER 17, 1921 PRICE. 5 CENTS Sold only by the Superintendent of Documents, Government Printing Office Washington. D. C. WASHINGTON GOVERNMENT PRINTING OFFICE 1921 RESULTS OF A SURVEY OF ELEVATOR INTERLOCKS AND AN ANALYSIS OF ELEVATOR ACCIDENT STATISTICS By C. E. Oakes and J. A. Dickinson ABSTRACT This report gives the results of a field survey of several thousand elevator landings equipped with various types of mechanical and electromechanical interlocks and contact devices. The survey was conducted in connection with the preparation of an elevator safety code, in which work the Bureau of Standards engineers have co- operated with engineers of the American Society of Mechanical Engineers. The elevators are classified: A, Elevators in buildings having heavy service and where maintenance service is provided; B, elevators located in buildings where the service is heavy but without maintenance; and C, elevators on which the service is light and for which no maintenance service is provided. The statistics show that 73.8 per cent of all fatal accidents might be prevented by well-designed interlocks. CONTENTS Page I. Introduction 3 II. Report on a field survey of elevator interlocks and contacts 4 1. Hoistway-door protective devices 5 2. Emergency release g 3. Field survey 9 4. Division of buildings into classes according to conditions of service and maintenance 9 5. Discussion of conditions under which devices must operate 10 6. Tabulation of results of survey n 7. Discussion of survey 15 III. Summary of advantages and disadvantages of various types of devices. . . 18 IV. Elevator-accident statistics 21 r. Group I. General: Reports of accidents received through clipping bureaus 21 2. Group II. Reports from State industrial commissions 22 3. Group III. Municipal statistics 25 V. Suggested requirements to be included in an elevator code 28 1. Definitions 28 2. Specifications 29 I. INTRODUCTION During the European War the Bureau of Standards cooperated with the Federal safety engineers in preparing elevator rules, intended primarily for the use of Government navy yards, arse- 3 4 Technologic Papers of the Bureau of Standards nals, and explosive plants. The code was based on existing regu- lations as prepared by various States and municipalities, together with suggestions from elevator manufacturers and safety engineers. This work was completed in the spring of 1918. Following the completion of this code the Bureau was requested to revise the Federal code to render it suitable for general application. The American Society of Mechanical Engineers published an elevator code in 1917, which was prepared by the subcommittee on the protection of industrial workers. A complete revision of this code was undertaken a few months after its publication. In order to avoid duplication of effort, the Bureau in 1918, at the invitation of the American Society of Mechanical Engineers, began to cooperate in the further development of this code. The Bureau placed the data it had collected at the disposal of the American Society of Mechanical Engineers' committee and later, in 1919, abandoned all idea of separate development and publication of an elevator safety code, but continued to cooperate with the A. S. M. E. committee with the idea of developing a single national code. Early in the work it became evident that the point of greatest controversy was the question of proper hoistway-door protection. The Bureau of Standards decided to make a field survey of the various types of interlocks and contacts in use and did so in 1918 and 1919. The results were submitted to the committee for its use in drafting the portion of the code dealing with such devices. This paper gives the final results of the survey and a collection of statistics from other sources which have been classified and arranged to show the relation of accidents to hoistway-door pro- tection. U. REPORT ON A FIELD SURVEY OF ELEVATOR INTER- LOCKS AND CONTACTS In the preparation of a safety code of any kind the need for reliable data upon which to base the various requirements is self- evident. There are certain hazards which are obviously apparent. It may be, however, that hazards of a far graver nature exist which are not apparent to a person making a casual survey of the in- dustry. Ofttimes these less apparent hazards prove to be more serious than those readily apparent. A careful study of acci- dent statistics taken from as large a number of sources as possi- Elevator Interlock Report 5 ble will indicate in a general way the particular parts or types of machines responsible for the more serious accidents. A summary of the accidents given in the appended tables will show that by far the larger portion of elevator accidents occur at the hoistway door. The question, therefore, of the best method of reducing this hazard is one to which considerable study has been given. 1. HOISTWAY-DOOR PROTECTIVE DEVICES All forms of hoistway-door protective devices may be broadly classified in three divisions, according to the method of locking the car and the hoistway door. THE MECHANICAL TYPE. This type of door safety depends on mechanical action to lock the car-control mechanism while the door is open and to hold the door in the locked position when the car is not at the landing. The functions of this device are gener- ally as follows : (a) Release of hoistway door, (b) locking control mechanism, (c) release of control mechanism, and (d) locking of hoistway door. In most types the locking of the hoistway door is accomplished by holding locked, either by means of a very substantial latch or a properly shaped slot in a rotating member, a bar fastened rigidly to the hoistway door. Generally, the door-locking mecha- nism is substantial in construction and will stand a large amount of wear without loss of function. Mechanisms which do not depend entirely on spring action to hold the door locked are preferable. The methods of interlocking the car and door are varied. A variety of methods are necessary, owing to the large number of different types of control mechanism used on different makes and types of elevators. Two popular styles in use are (a) a slotted flap which upon the opening of the landing door drops from a vertical to a horizontal position, the slot engaging and holding inoperative the car-control lever in the neutral position; (6) a device in which an arc of a circle properly slotted is fastened to the control lever. A bolt or rod engages in this slot, thus pre- venting the motion of the operating lever when the lever is in the neutral position and the car door open. As a rule these devices are satisfactory. Occasionally, however, installations are found in which there is too much lost motion, and in certain hydraulic installations it is possible to start the car very slowly, due to the slight motion allowed on the control lever when the 6 Technologic Papers of the Bureau of Standards car is in the "locked" position. This objection does not ordi- narily hold true in case of electrically controlled elevators, as the operating lever must move through an appreciable arc before the operating circuit is completed, although in controllers with a number of contact steps the first contact may be so close to the ' ' stop ' ' position that cases have been noted in which the car may be moved slowly with the interlocking mechanism in the ' ' locked ' ' position. As installed at the present time, a great many of these devices have accessible metal parts in the car, the removal or blocking of which may render the device inoperative. Such parts as are located within the car should be inclosed. One of the features generally incorporated in this type of lock is some form of limiting device which will prevent the opening of the hoistway door except when the car is within 3 or 4 inches of the landing at which the stop is to be made. The inclusion of a landing range in an interlock device is not a necessary adjunct to the interlocking function. It is a mechanical detail which is inherent in certain types of devices where inter- locking is secured by mechanical means. The absence of a landing range detracts in no way from the proper functioning of a well-designed interlock. The use of a device of this kind is intended to lessen the number of accidents due to persons trip- ping on entering or leaving the car. Such accidents are, however, relatively unimportant, the minor nature of these accidents being emphasized by the fact that statistics concerning them are not generally recorded separately. At the request of the Bureau of Standards a list of tripping accidents was prepared by one of the largest casualty insurance companies, the list including all tripping accidents occurring on elevators insured by them during the year 1919. An analysis of this list shows that more tripping accidents occurred when the car platform was within 3 inches of the landing level than occurred when the car platform was more than 3 inches away. In other words, a 3 -inch stopping range would not have prevented the majority of these accidents. The consideration of a more or less parallel case may be of interest: A person frequently trips over a projection in a sidewalk where the difference of elevation is slight, but almost never trips over a curbstone where the differ- ence of elevation is considerable. It appears that threshold illumination is an important consideration in preventing tripping accidents and should be given careful attention, if such accidents are to be avoided. Elevator Interlock Report 7 Leveling devices, both automatic and manually controlled, are now on the market and can be used on either passenger or freight elevators. These devices bring the car flush with the landing, and in the case of the automatic device maintain it there during loading and unloading and are doubtless the most effective devices avail- able for decreasing accidents from tripping. ELECTROMECHANICAL. The second broad division into which door safety devices may be placed is the electromechanical type, in which the door is mechanically held closed but in which the car is interlocked by means of some form of electrical control. This may be accomplished either by interrupting the master or control circuit or by running an independent circuit which will interrupt the operating current on the machine control board or, in the case of hydraulic installations, by using a current to actuate a solenoid which in turn mechanically locks the car-control lever or holds closed a valve in the supply line, thus preventing motion of the car. This type of device (electromechanical) may or may not limit the stopping of a car to within a few inches of the landing plat- form before the door can be opened. Those devices in which the electric contact is place^.jm^i:jthe_car_must necessarily be provided with a stopping range limit, and their operation pre- vents the opening of jthe hoistway door unless the car is within a predetermined distance of the landing flooFTevel, while those depending' on switches placed in the hoistway~can generally be opened without respect to the relative position of the car and hoistway landing platform. The door-locking mechanisms used in devices of this kind are varied. Those with the undercar contact generally use the same mechanism as is employed in the purely mechanical type, while those with a hoistway contact generally make use of a bar-lock or some form of toggle-joint device. Of the two the latter is prefer- able, as it does not depend on springs to hold the door locked. ELECTRIC CONTACT DEVICES. The third class of door safety devices comprises a wide variety of electric hoistway-door con- tacts. These devices are generally connected in series known as shaft-series system and operate on the car-control circuit, although they may be wired so that each door is a separate unit, this wiring being known as a door-unit system. Those in which the switch is actuated by the door itself (there are a large variety of this type) give considerable trouble, due to lack of alignment of the door and switch and to vibration and the impact of the door. 8 Technologic Papers of the Bureau of Standards The difference between a hoistway-door contact and an electro- mechanical interlock should be clearly kept in mind. A contact device holds the car inoperative when the door opposite which the car is standing is open. An electromechanical interlock is a device which holds the car inoperative when the door opposite which the car is standing is open or unlocked and in addition holds locked the hoistway door when the car is not at the landing. This door-locking function is accomplished by a mechanism inde- pendent of, and separate from, the ordinary door latch. In other words, a contact is a device which can only prevent the motion of the car when a door is open. It can not and does not hold locked the door when the car is not at the landing. The ordinary hoistway-door latch must be used as the door-locking device. The electromechanical device has a bar or other form of lock incorporated in it and so arranged and interconnected that the door must be closed and locked before the car can be started. In order to obtain information regarding the effectiveness of the ordinary hoistway-door latch, a survey was made of several hundred hoistway doors. The results of this survey indicate that 30 per cent of the latches were out of order, the principal causes of failure being worn latch parts, worn door hangers, loosening of latch from door, and lack of lubrication. Many latches were so poorly designed that if the door were closed quickly the rebound of the door would occur before the latch engaged. An analysis of elevator-accident statistics shows that about one-third of the accidents occur by persons falling down the shaft. This is in itself an indication that the ordinary latch is unreliable and shows why a contact device can not of itself offer complete hoistway-door protection. In the other two types of device (mechanical and electromechanical) the door is held shut mechanically whether or not the car interlocking feature is working. 2. EMERGENCY RELEASE Every elevator car equipped with interlocks should be provided with an emergency switch or release device located in the car. If there is a fire, panic, or other emergency, or if the car becomes stalled in the shaft due to failure of the door safety circuit, the operation of the emergency device will permit the elevator to con- tinue in operation. Unless this device is properly designed and installed its use may be subject to much abuse. Elevator Interlock Report 9 An undesirable feature of an emergency switch so designed that the switch can be turned "on" or "off" is that, in event of failure of any contact, the throwing of an emergency switch leaves the entire system with no protection other than that offered by the ordinary door latch, which is notoriously unreliable. The emergency release generally consists of a two-pole push or snap switch sometimes placed in a case under a glass cover, often without even this protection, and when once thrown is quite likely to be left in the "emergency" position for weeks or months. Except where required by law, the spring-type switch, which must be held in the emergency position by the operator, is seldom seen, although it is obviously the most desirable type. The emergency-release device for mechanical locks which remains in the emergency-operating position without being held there by the operator is equally undesirable. 3. FIELD SURVEY In order to obtain reliable information regarding door protective devices, a large number of passenger elevators were inspected by representatives of the Bureau of Standards and thousands of doorways and landings examined. These inspections covered a considerable range of territory, and territories in which the legal requirements for the protection of entrance doors to elevators differ widely. The following statement gives the various cities and States in which data were secured and the legal requirements in each place : Requirements Philadelphia Mechanical or electromechanical interlocks of the undercar contact type. Pennsylvania (outside of cities of first or second class) Mechanical or electromechanical interlocks and car-door contacts on cars driven by electricity. New Jersey Mechanical or electromechanical. New York City No interlock requirements. 4. DIVISION OF BUILDINGS INTO CLASSES ACCORDING TO CONDITIONS OF SERVICE AND MAINTENANCE For the convenience in analyzing the performance under actual operating conditions of any given interlock device or group of such devices it is desirable to divide the elevators which they protect into three general classes, the classification being natu- rally somewhat elastic. The two factors which govern this classi- 61693 21 2 io Technologic Papers of tlie Bureau of Standards fication are the amount of duty rendered and the maintenance service they receive. The grouping adopted is as follows: CLASS A. Those in which the volume of travel is heavy and for which there is a special elevator maintenance and inspection service. Into this class will fall: (a) Large office buildings, (6) first-class hotels, (c) high-grade apartment houses, (d) large rail- road terminals, (e) municipal and State buildings in cities, and (/) large department stores. This class is generally limited to the larger cities. CLASS B. Buildings in which the travel is heavy, but in which no regular maintenance or inspection service is provided. In this are included the following types of buildings: (a) Small office buildings in large cities, (6) large office buildings in small cities, (c) loft and manufacturing buildings, (d) small hotels in cities, (e) medium-size apartment houses, (/) small department stores in cities, and (g) other retail stores in cities. CLASS C. Buildings that have light service and receive but little care and attention. In this class will fall: (a) Small office buildings in towns or small cities, (b) storage or warehouse build- ings, (c) small hotels, (d) manufacturing plants where the upper floors are used largely for storage, and (e) very small shops and retail stores (such as furniture stores) where most of the selling is done on the ground floor. 5. DISCUSSION OF CONDITIONS UNDER WHICH DEVICES MUST OPERATE In Class A buildings the vital consideration is speed of service. In most installations of this class there are several load peaks, and at such times every car is run at capacity. A loss of a half minute a trip per elevator would mean a serious curtailment of service. Any device that is installed must not reduce the operating speed to any marked degree. Further, it must be provided with some form of emergency release, so that the failure of the device or of any part of it will not tie up the car during the rush period. The survey showed that there is generally very little operating trouble due to failure of door-safety equipment in Class A build- ings, because ample maintenance service is provided. As a rule, the elevator maintenance crew take pride in the condition of the equipment under its care and keep parts properly adjusted and in repair. In Class B buildings the vital consideration is simplicity and ruggedness, as the service in this class of buildings is severe and Elevator Interlock Report 1 1 the maintenance and attention which elevator parts and devices will receive is slight. Lack of alignment on the part of the elevator or of the hoistway door may cause excessive wear of the interlock or contact, causing rapid deterioration, which results ultimately in failure to operate. As a usual thing the only attention which a Class B elevator receives is an occasional "going over" by some mechanic or handy man for the purpose of lubrication. Even this perfunctory oiling is done at odd times and at very irregular intervals. It is very seldom that the elevators in a Class B building are in charge of a starter, and it is only in very exceptional cases that the operators are given any definite instructions as to the operation of the car; in other words, operating rules are usually unknown. This class (Class B) will contain by far the largest part of the elevators in almost any city or large town. For this reason the behavior of the various door-protective devices in Class B buildings is a matter of vital interest. For this class of service the most simple and rugged type of mechanical interlock or electromechan- ical device is desirable. Almost invariably electric door contacts seem to give trouble, under the combination of heavy duty and little or no attention. For elevators in Class C buildings the requirements which are to be met are not so severe. Any device that will stand up under occasional use, but with almost entire lack of attention, will probably give satisfaction. Most mechanical devices give fairly good service, although in some cases which were noted they have worn rapidly owing to lack of lubrication. Electromechanical and contact devices give fair service provided they are used often enough to prevent the contacts from oxidizing and corroding. This class is probably smaller than either Classes A or B, and is confined very generally to smaller cities and towns. Frequently in Class C installations may be found various types of makeshift devices, some of which fail to accomplish the purpose for which they were intended. 6. TABULATION OF RESULTS OF SURVEY The data regarding interlocks and electric contacts given on the following pages were obtained by examining substantially all of the elevators in many of the smaller cities and towns of Penn- sylvania and New Jersey and all of the elevators in representative blocks in the larger cities. In uptown New York the blocks were 1 2 Technologic Papers of the Bureau of Standards between Fifth Avenue and Broadway and Thirty-sixth and Forty- fifth Streets. In downtown New York both sides of Broadway were covered for a number of blocks. This was done in an effort to observe an average condition, and it is believed that the data given here are typical of the various towns or sections in which inspections were made. Summary of Conditions of Interlocking Devices in Pennsylvania (Exclusive of First- Class Cities) GENERAL ANALYSIS Elevators inspected 79 Landings inspected 513 Cars not provided with car-gate contacts (required by Pennsylvania law) ... 20 Per cent not up to requirements of Pennsylvania law 25. 2 Landings provided with interlocks 513 Landings provided with interlocks found to be inoperative 47 Per cent of landings with interlocks inoperative 9. 2 ANALYSIS BY TYPE OF DEVICE Landings provided with mechanical interlocks 244 Landings with mechanical interlocks inoperative 26 Per cent of landings with mechanical interlocks inoperative 10. 6 Landings provided with electromechanical interlocks 269 Landings provided with electromechanical interlocks found to be inoperative . 2 r Per cent electromechanical interlocks found to be inoperative 7. 8 ANALYSIS BY CLASS OF BUILDINGS IN WHICH ELEVATORS ARE LOCATED CLASS A Landings in this class 149 Landings with interlocks inoperative 10 Per cent of landings with interlocks inoperative 6. 7 CLASS B Landings in this class 313 Landings with interlocks inoperative 36 Per cent of landings with interlocks inoperative zz. 5 CLASS C Landings in this class 51 Landings with interlocks inoperative i Per cent of landings with interlocks inoperative 2. o Summary of Conditions of Interlocking Devices in New Jersey GENERAL ANALYSIS Elevators inspected 189 Landings inspected 1466 Landings with no hoistway-door protection 227 Per cent of landings with no hoistway-door protection 15- 5 Landings not provided with mechanical or electromechanical interlocks (as required by New Jersey law) 559 Elevator Interlock Report 13 Per cent of landings not provided with mechanical or electromechanical interlocks (as required by New Jersey law) 44. 8 Landings provided with interlocks or contacts 1239 Landings with interlocks or contacts inoperative 73 Per cent of landings with interlocks or contacts inoperative 5. 2 ANALYSIS BY TYPE OF DEVICE Landings provided with mechanical interlocks 243 Landings provided with mechanical interlocks found to be inoperative 29 Per cent of mechanical interlocks found to be inoperative 11.9 Landings provided with electric contacts * 329 Landings provided with electric contacts found to be inoperative * 44 Per cent of electric contacts found to be inoperative * 13. 3 Landings provided with electromechanical interlocks 2 667 Landings provided with electromechanical interlocks found to be inopera- tive 2 o Per cent of electromechanical interlocks found to be inoperative 2 o ANALYSIS BY CLASS OF BUILDING IN WHICH ELEVATORS ARE LOCATED CLASS A Landings in this class 845 Landings with interlocks or contacts inoperative 19 Per cent of landings with interlocks or contacts inoperative a. 2 CLASS B Landings in this class 350 Landings with interlocks or contacts inoperative 53 Per cent of landings with interlocks or contacts inoperative 15. i CLASS c Landings in this class 44 Landings with interlocks or contacts inoperative i Per cent of landings with interlocks or contacts inoperative 2. 3 Summary of Conditions of Interlocking Devices in Philadelphia [NOTE. Municipal regulations require the use of mechanical or undercar contact electromechanical inter- locks.] GENERAL ANALYSIS Elevators inspected 49 Landings inspected 503 Landings provided with interlocks 503 Landings with interlocks inoperative 66 Per cent of landings with interlocks inoperative 13.0 ANALYSIS BY CLASS OF BUILDING IN WHICH ELEVATORS ARE LOCATED CLASS A Landings in this class 404 Landings with interlocks inoperative 9 Per cent of landings with interlocks inoperative 2. 2 1 While contact devices without door-interlocking mechanisms are not permitted under the New Jersey law, a number of such devices were found. Some of these devices were installed prior to the enactment of the law and will, no doubt, eventually be changed to comply with the legal requirements. 1 Seventy-five per cent of the electromechanical devices were found in one "Class A" group, where the maintenance service was excellent 14 Technologic Papers of the Bureau of Standards CLASS B Landings in this class _ 99 Landings with interlocks inoperative 3 57 Landings with interlocks inoperative (final) 28 Per cent of landings with interlocks inoperative 56. o Per cent of landings with interlocks inoperative (final) a 8. o Summary of Conditions of Interlocking Devices in Uptown New York City 4 GENERAL ANALYSIS Elevators inspected 294 Landings inspected 3, 043 Landings provided with interlocks or contacts 229 Per cent of landings provided with interlocks or contacts 7. 5 Landings with interlocks or contacts inoperative 50 Per cent of landings with interlocks or contacts inoperative ai. 8 ANALYSIS BY TYPE OF DEVICE Landings provided with electric contacts 190 Landings with inoperative electric contacts 37 Per cent of landings with inoperative electric contacts 19. a Landings provided with electromechanical interlocks 39 Landings provided with electromechanical interlocks inoperative 13 Per cent of landings with electromechanical interlocks inoperative 33-3 DATA ON CAR-GATE CONTACTS Cars provided with car-gate contacts aa Car-gate contacts in operating condition 10 Per cent of car-gate contacts in operating condition 45-5 Car-gate contacts tied up or plugged 7 Per cent of car-gate contacts tied up or plugged 31. 8 Car-gate contacts broken or defective 5 Per cent of car-gate contacts broken or defective 23. 7 Summary of Conditions of Interlocking Devices in Downtown New York City * GENERAL ANALYSIS Elevators inspected 241 Landings inspected 5, 304 Landings provided with interlocks or contacts i, 926 Per cent of landings provided with interlocks or contacts 36. 4 Landings with interlocks or contacts inoperative a Per cent of landings with interlocks or contacts inoperative a 10 ANALYSIS BY TYPE OF DEVICE Landings provided with mechanical interlocks 893 Landings with mechanical interlocks inoperative i Per cent of landings with mechanical interlocks inoperative 0.10 Landings provided with electromechanical interlocks i, 014 Landings with electromechanical interlocks inoperative i * As 29 of the 57 defective landings were in a single group of buildings which had evidently been over- looked by the inspector, this estate was omitted from the final statistics as not being representative. 4 Only one mechanical interlock was found in the particular uptown section inspected, so that statistics on this type are not given. Buildings equipped with interlocks or contacts in this tabulation may be r- garded with very few exceptions as Class B. 6 All of these downtown buildings provided with interlocks or contacts in this tabulation ar CWi A Elevator Interlock Report 15 Per cent of landings with electromechanical interlocks inoperative o. 10 Landings provided with electric contacts 20 Landings provided with electric contacts inoperative o Summary of an Inspection to Determine the Effectiveness of Emergency Releases, Uptown New York City 6 ELECTROMECHANICAL DEVICES Elevators inspected 43 Landings inspected 504 Landings working satisfactorily 225 Per cent of landings working satisfactorily 44-6 Landings where elevators were running with emergency switch thrown; no electrical or mechanical trouble 148 Per cent of landings unprotected, due to emergency switch being thrown. . . 29. 4 Landings where elevators were running with emergency switch thrown or system disconnected, due to electrical or mechanical breakdown 131 Per cent of landings unprotected, due to electrical or mechanical breakdown. 26. o This survey was made to determine to what extent the emer- gency release is used as a means of defeating the purpose of door- protective devices. The emergency releases were all of the two-pole, push-button type. Most of them were without a cover glass or even a means of holding a cover glass. The results of this survey show that an emergency release of this type will be misused. The only type which should be permitted is that in which the operator is re- quired to hold the emergency release in the emergency operating position. Practically all of the devices inspected on this survey were poorly designed and cheaply made, which may in part account for the large number of defective landings noted. 7. DISCUSSION OF SURVEY RELATION OF LAW ENFORCEMENT TO ELEVATOR SAFETY. There is a marked difference between the condition of elevators in a State or city where the law or code relating to elevators is strictly enforced and a State or city where the law does not have the backing of an active and thorough administrative organiza- tion. In many parts of one State the law calling for interlocks did not seem to be generally known, and even in the larger cities no particular efforts had been made to enforce it. In many cases it has been through the notification given the building owners by insurance inspectors that the devices have been installed. By way of contrast, however, in another State there was not a single elevator examined that was not equipped with the required inter- 6 Nearly all of these are Class B buildings. 1 6 Technologic Papers of the Bureau of Standards lock, although in a few electric-elevator installations the addi- tional car-door contact which was required by law was lacking. Perhaps the most badly needed addition to present State rules and city ordinances is a set of simple, brief, and concise operating rules to be posted in every elevator under the jurisdiction of the particular State or city. Certain cities have already provided for the licensing of car operators. The passing of an examination is necessary before the applicant is granted a license. In this way it is possible to discourage the operation of elevators by the ignorant or uninstructed. Unfortunately, during the war it was deemed necessary to waive this ruling in certain cases, and the strict enforcement of this rule has not yet been insisted upon. PERFORMANCE OF INTERLOCKS IN VARIOUS CLASSES OF BUILD- INGS. In regard to the performance of interlocking devices, there seems to be but little difference between the mechanical and electromechanical interlocks. Many installations, both mechan- ical and electromechanical, were met with in Philadelphia that had been giving satisfactory service for periods of 15 years and upward, with a cost for repairs and renewals not exceeding $i per landing for the entire period, while numerous installations met with in the smaller cities of Pennsylvania were found to be in good condition after 10 years or more of service. As might be expected as the result of the combination of hard service with little or no maintenance, the Class B installations in every case have made the poorest showing, but even here the percentage of installations in working order indicates that the protection afforded by such devices is decidedly worth while. For Class A installations the average percentage of devices in working order was 98. When the fact that an interlock may be called on to function several hundred times a day is taken into consideration, this percentage must be considered remarkable. Class C installations were generally in fairly good condition, due mainly to light service requirements. INTERLOCK INSTALLATIONS IN LARGE OFFICE BUILDINGS. The results of the inspection in the down-town section of New York City brings forth two points of particular interest. First, the relatively large number of interlock installations. While interlocks are not required by the city ordinances, over one- third of the landings inspected had been so safeguarded. Second, the remarkably low number (two) of defective landings found despite the severe duty imposed on the elevators in the Elevator Interlock Report 17 large office buildings down town. Excellent maintenance service and the carefully designed and well-built interlocks which pre- dominate in this section are, no doubt, largely responsible for this fine showing. SUGGESTIONS FOR IMPROVEMENT IN DESIGN. From a general survey of results obtained it would seem that there is room for considerable improvement in the types of interlocking devices. Of the hundreds of elevators inspected none were equipped with a mechanical interlock in which the four vital functions of unlock- ing hoistway door, locking car mechanism, unlocking car mecha- nism, and locking the door are performed by means of positive mechanical motion. In other words, at least one, and generally more, of these functions are accomplished by means of springs or the force of gravity, or a combination of the two. An inter- lock in which all four of these functions are performed without the aid of a spring or the action of gravity should be not par- ticularly difficult of design. Rules to be drafted should, how- ever, take cognizance of present well-designed apparatus and should be so worded as to not hinder improvement in design or new designs. Door contacts, while usually of ample current-carrying capac- ity, are seldom rugged enough to stand up under the impact and vibration of the hoistway door, where they are actuated by the impact of the door itself. Another characteristic feature about them is that in nearly every case the making and breaking of the current is dependent directly upon and is accomplished at the same speed as the opening and closing of the door. Many such devices, if the door fails to latch, are left in but slight con- tact and may produce an arc which will rapidly destroy the con- tact surfaces. The use of a sequence coil or a quick make-and- break switch will do much to eliminate such arcing and pitting. There are several well-designed devices with switches that are not subject to door impact and which give long and satisfactory service when made a part of a substantial interlock mechanism. Well-designed mechanical locks are also available. A substantial door-locking mechanism is necessary whether the method of interlocking the car is electrical or mechanical. The door- locking function should be so interconnected with the car inter- lock that the car can not be started until the door is locked in the closed position. 1 8 Technologic Papers of the Bureau of Standards There is considerable need for some sound engineering work in connection with the proper design of interlocks. Many of the devices on the market were designed without due regard for the stresses set up by door impact, lack of alignment, and abuse of equipment at the hands of careless operators. Such devices are generally sold at a low price, the manufacturer depending upon this low price to obtain business. In many cases such devices fail after a short period of service and are, in the end, more expensive than a well-designed and carefully installed out- fit, the initial cost of which would have been greater. III. SUMMARY OF ADVANTAGES AND DISADVANTAGES OF VARIOUS TYPES OF DEVICES The following tabular arrangements set forth the advantages and disadvantages of the various types of elevator door safety devices and the characteristic methods of failure of the various types of devices for the various classes of service: Elevator Interlock Report JE w s I 11 I II I ! 15 03 to 42 3 US t-1 s 5 $* ti a =S |l I IS s I? . III? 1 ? &2.S +* "^ CB is it It ^rjS aj-e j. rfs " 3 u* CJ 53 ** 3 ll liisfii I I i 1 * 11 20 Technologic Papers of the Bureau of Standards CJXJ43 O *> |s*S| SI? If *J'I lillf Klj* :fff! !lifi ^3-K S Sf p | r".2-s^ SB.Sxjg.^ lii J3 ^3 o I 1 JB ***B W fl =il!ll aets Elevator Interlock Report 21 Public accidents. IV. ELEVATOR-ACCIDENT STATISTICS There are attached a .number of tables of elevator statistics obtained from various sources. The public accidents have been separated from the industrial accidents where this was possible, and the classification has been made to show the number of acci- dents by causes. Accidents have been classified as public, semipublic, or indus- trial, according to the nature of the occupancy of the building, as follows : Hotel Apartment house Department store (patrons). Office building [Hospital Semipublic accidents {_ IGarage Industrial accidents. . . (Mercantile establishment (employees only). The accidents have been further grouped into shaft-door and nonshaft-door accidents and the percentages computed on this basis for each table. For convenience in analysis the statistics have been divided into groups according to sources, as noted below and shown in the tables following: Group I. General: Reports of accidents received through clip- ping bureaus. These cover the entire United States. Group II. State industrial reports of State industrial commis- sions. These are taken from records of industrial accidents and do not cover accidents to the general public. Group III. Municipal statistics from records of coroners of cities or records of elevator inspection departments. 1. GROUP I. GENERAL TABLE 1. Elevator Accidents Reported Through Clipping Bureaus by the Public Press of the United States, January, 1913, to July, 1918 Fa tal Non fatal Cause Public Semi- public Indus- trial Not classi- fied Public Semi- public Indus- trial Not classi- fied. Fell into shaft from landing floor or due to motion of the car 150 15 198 121 10 181 Crashed between car and sill of landing floor or frame of 150 10 168 107 5 138 96 11 150 330 12 486 Cause not given 11 12 7 11 24 11 Total 407 36 528 7 569 27 779 11 22 Technologic Papers of the Bureau of Standards TABLE 2. Summary from Table 1 : Entire United States from Newspaper Clippings, January, 1913, to July, 1918 Classification and cause Fatal Nonfatal Public: Per cent Per cent Nonshaft-door accidents 26. 2 60 Shaft-door accidents- Fell into shaft 36.9 21.3 Crushed between car and sill or frame of door 36. 9 18. 7 Total 100 100 Semipublic: Nonshaft-door accidents 30. 5 44. 5 Shaft -door accidents- Fell Into shaft 41. 7 37 Crushed between car and sill or frame of door 27. 8 18. 5 Total 100 100 Industrial: Nonshaft-door accidents 49.6 59 Shaft -doer accidents- Fell into shaft 37. 4 23. 2 Crushed between car and sill or frame of door 13 17.8 Total. . . 100 100 2. GROUP H. REPORTS FROM STATE INDUSTRIAL COMMISSIONS OR DEPARTMENTS OF LABOR TABLE 3. Elevator Compensable Industrial Accidents, New York State, June 1, 1914, to July 1, 1915 [From data submitted by the New York Industrial Commission. N. O. C. Not otherwise classified] NONFATAL ELEVATOR ACCIDENTS Cause Num- ber Per cent Cause Num- her Per cent Caught between floor and car. Falls into shaft from floor Caught between shaft and car . Fall of car Struck by car (N. O. C.) Gates (N. O. C.) Unknown Objects falling into shaft Cables breaking Caught by machinery Caught by cable Falls into shaft from car. . . 113 63 26 41 36 39 33 61 32 25 25 13 19.6 10.9 4.5 7.1 6.3 6.7 5.7 10.0 5.6 4.3 4.3 2.3 Caught between car and gate Struck by counterweight Objects falling down shaft from car. Struck by car in pit Machinery breaking Load catching between car and shaft Cables unwinding Struck while on top of car Total. 577 2.3 1.9 2.3 1.5 1.7 1.0 .7 .7 100.0 Elevator Interlock Report TABLE 3. Elevator Compensable Industrial Accidents, New York State, June 1 1914, to July I, 1915 Continued FATAL ELEVATOR ACCIDENTS Cause. Num- ber. Per cent. Cause. Num- ber. Per cent. Falls into shaft from floor 13 27.7 Unknown .. .. .... 1 2 1 Caught between shaft and car. . . . 6 12.8 Machinery breaking 1 2 1 Caught between floor and car Struck by car (N. O. C.) 5 4 10.6 8.6 Load catching between car and shaft 1 2.1 Fall of car 3 6 4 Csble unwinding 1 2.1 Falls into shaft fronj car 3 6.4 Struck while on top of car 1 2 1 Total 47 100.0 Caught between car and gate 2 4.2 TABLE 4. Elevator Accidents in Industries and Mercantile Establishments in the State of Pennsylvania, 1918 [From statistics furnished by the Department of Labor and Industry of Pennsylvania] Fatal Nonfatal Total Number Percent Serious: Number Minor: Number Total Number Percent Number Per cent Cable breaking 4 1 3 17 3 2 1 1 5 7 1 5 8 2 6 34 6 4 2 2 10 14 2 10 27 9 43 134 30 10 10 10 35 34 15 31 29 7 49 171 36 7 17 35 26 27 16 50 56 16 92 305 66 17 27 45 61 61 31 81 6.5 1.7 10.7 35.7 7.8 2.0 3.1 5.3 7.1 7.1 3.6 9.4 60 17 95 322 69 19 28 46 66 68 32 86 6.6 1.9 10.5 35.5 7.6 2.0 3.1 5.1 7.3 7.5 3.5 9.4 Cable unwinding Cable, caught by Car, caught by Car, struck by Counterweight, struck by Defective equipment Entering or leaving car Fall of car Fall of person Falling objects ... AO others TotaL SO 100 388 470 858 100.0 908 100.0 24 Technologic Papers of the Bureau of Standards TABLE 5. Wisconsin Industrial Accidents on Elevators [From published reports of the Wisconsin Industrial Commission, 1915-1917] Cause Fatal Nonfatal Total Number Per cent Number Per cent Number Per cent Falling from floor down shaft 5 2 35.6 14.3 24 55 14 14 14 16 4 4 10 22 18 6 4 11.7 26.8 6.9 6.9 6.9 7.8 1.9 1.9 4.9 10.7 8.8 2.9 1.9 29 57 14 15 15 17 4 5 10 22 20 7 4 13.2 26.0 6.4 6.9 6.9 7.7 1.8 2.3 4.6 10.1 9.1 3.2 1.8 Platform floor catching person Caught between cage and side Struck by objects falling dowri shaft 1 1 1 7.1 7.1 7.1 Car struck person Car fall big Car sudden start or stop Struck by counterweight 1 7.1 Hoisting cable striking person Hoisting machinery catching person Platform and gate catching person 2 1 14.4 7.1 Falls from car Load or other object catching person Total 14 100.0 205 100.0 219 100.0 TABLE 6. Elevator Accidents in Industries in Massachusetts (Freight and Passenger, Insured, not Insured, and Common Law Rights) [From published reports of Massachusetts Industrial Accident Board Total Cause July 1,1913, to June 30, 1914 July 1, 1914, to June 30, 1915 July 1, 1915, to June 30, 1916 July 1, 1916, to June 30, 1917 Fatal Nonfatal Fatal Non- fatal Fatal Non- fatal Fatal Non- fatal Fatal Non- fatal Num- ber Per cent Num- ber Per cent Caught in ma- chin err 62 40 30 1 47 1 1.5 179 6.92 Caught be- tween car and shaft 7 50 6 186 7 204 6 289 26 38.8 728 28.1 Caught under- neath or on top of car. . . . 36 16 4 25 3 18 7 10.5 95 3.67 Falling car 1 38 49 1 44 1 47 3 4.48 179 6.87 Falling down shaft (person) 1 59 3 59 9 56 9 61 22 32.9 235 9.08 Struck by fall- ing object 53 1 55 24 29 1 1.5 161 6.22 Caught by fire hatch or trap . 9 6 1 9 25 .96 Miscellaneous 284 3 265 2 213 2 228 7 10.5 990 38.2 Total 9 591 13 676 23 597 22 728 67 100.0 2,592 100.0 NOTB. The classification used here is not as suitable for grouping the hoistway-door accidents into two general classes as the other tables presented. Elevator Interlock Report TABLE 7. Analysis of State Elevator Statistics INDUSTRIAL ELEVATOR ACCIDENTS, NEW YORK STATE, JUNE 1, 1914, TO JULY 1, 1915 Cause Fatal Nontatal If onshaft-door accidents Per cent 51 1 Percent 64 9 Shaft-door accidents: Fell into shaft 34.1 13 2 Caught between floor and* car or gate and car 14 8 21 9 Total 100 100 INDUSTRIAL AND MERCANTILE ELEVATOR ACCIDENTS, PENNSYLVANIA, 1918 If onshflft-door accidents 50 51 9 Shaft-door accidents: Fell into shaft 14 7 1 Caught by car and entering or leaving car 36 41 Total. 100 100 INDUSTRIAL ELEVATOR ACCIDENTS, WISCONSIN, 1915-1917 Nonshaft-door accidents 35 7 52 7 Shaft-door accidents: Fell into shaft 35.6 11 7 Platform floor catching person platform and gate catching person 28 7 35 6 Total 100 100 INDUSTRIAL ELEVATOR ACCIDENTS, MASSACHUSETTS, JULY 1, 1913, TO JUNE 30, 1917, INCLUSIVE If onshaft-door accidents 28.3 62 82 Shaft -door accidents: Fell into shaft 32.9 9 Q8 Caught between car and shaft 38 8 28 1 Total 100 100 3. GROUP m. MUNICIPAL STATISTICS TABLE 8. Philadelphia Elevator Accidents 19 14 IS 15 19 16 19 17 19 18 iype Injured Killed Injured Killed Injured Killed Injured Killed Injured 'Killed Passenger :<> Electric 1 1 g 1 Hydraulic 3 1 3 1 5 3 6 3 4 Freight: Electric 7 2 14 2 8 2 5 5 5 2 Hydraulic 6 5 4 1 9 5 5 2 2 3 Belt 17 4 22 4 24 2 24 7 12 12 Hand 2 3 3 2 1 Sling 1 1 1 Total 33 16 46 9 46 13 49 19 22 23 Interlocks required by law on passenger elevators. 26 Technologic Papers of the Bureau of Standards TABLE 9. Fatal Accidents on the Elevators of Borough of Manhattan, New York City, 1907 to 1918, Inclusive [From coroner's reports] Cause Number Per cent Fell down shaft 292 40.5 Crushed between moving car and sill or frame of shaft door 335 46. 5 Miscellaneous (broken cables, falling cars, workmen in the shaft struck by moving car.etc.) ... 94 13 Total 721 100 TABLE 10. Fatal Elevator Accidents, Cook County, HI., 1904 to March, 1916, Inclusive [From coroner's reports] Cause Number Per cent Fell down elevator shaft 169 42. 2 Crushed between elevator and floor, wall, door, etc 152 37. 9 Killed by falling elevator 24 6.0 Struck by elevator 23 5. 7 Crushed by counterweights 12 3. Crushed in elevator 8 2. Struck by falling objects other than elevator 1. 7 Crushed by machinery of elevator 4 1.0 Unknown how accident occurred 2 .5 Total... 401 100.0 TABLE 11. Summary of Elevator Accidents of Two of the Largest American Cities [Reported from coroner's records] Fatal Fatal Borough of Manhattan, New York City, 1907 to 1918, inclusive, public and industrial: Nonshaft-door accidents Shaft -door accidents- Fell down shaft Crushed between car and sill or frame of shaft door Cook County, 111., 1904 to March, 1916, inclusive: Per cent 13 40.5 46.5 Nonshaft-door accidents Shaft -door accidents Fell down shaft Crushed between elevator and floor, wall, door, etc Total.. 100 Total.. Per cent 19.9 42.2 37.9 100 An analysis of these statistics shows that the weighted average percentage of accidents presented in the preceding tables, due to causes originating from the lack of an interlocking device with which is combined a substantial lock, is as follows : Elevator Interlock Report 27 Fatal accidents to public, per cent of total accidents given in preceding tables. 73. 8 Fatal accidents to industrial employees, per cent of total 54 Nonfatal accidents to public, per cent of total 40 Nonfatal accidents to industrial employees, per cent of total 53. 2 The statistics show that the largest number of accidents is caused by falling down the shaft, the weighted average percentage of the total fatalities to the public from this cause being 36.9, and to industrial employees being 36. In the case of nonfatalities these percentages are, public 21.3, industrial employees 14.7. It should be noted that the percentage of all fatalities from falls is about the same for both classifications, public and industrial. This is a strong indictment of the ordinary door latch, and the results of the survey bear out the conclusion reached by a study of accident statistics. A substantial lock should, therefore, be made a part of every interlock and the ordinary door latch dis- carded. A somewhat minor point which should not be neglected in the design of interlocks, since it bears directly on the hazard of falling down the shaft, is the possibility of opening the door from the landing side when the car is passing a landing. Devices which permit this should not be installed. The next most prolific cause of accidents is the crushing hazard ; that is, the possibility of being crushed between car and sill of landing floor or frame of door. The average percentage of the total fatalities to the public from this cause is 36.9 and to indus- trial employees 18. In the case of nonfatalities the percentages are, public 18.7, industrial employees 38.5. These values, on comparison with those for causes originating within the shaft, show the necessity of interlocking the movement of the car with the opening and locking of the door. Not only should the interlocking function be associated with the closing of the door but also the locking of the door, inasmuch as the unlocked door exposes the user to the hazard of falling into the shaft. Another cause of accident resulting from lack of interlock devices is stepping into the shaft under the mistaken idea that the car is at the landing. A study of 978 elevator fatalities showed that i .3 per cent of the accidents were due to this cause, and of 1386 nonfatalities the percentage was 0.9. This is a real hazard and is comparable with many of the accident percentages due to causes originating within the shaft. It is recommended that provision be made against it. 28 Technologic Papers of the Bureau of Standards Perhaps the most striking figures are those of the city of Phila- delphia, which show the remarkable results obtained by the compulsory use of interlocks in a large city. Of the four people killed on passenger elevators in 1918 only one was a passenger, the other three being elevator operators or mechanics. The fact that there was not a passenger nor an operator killed in an electric elevator during the entire year of 1918 is a remarkable tribute to the effectiveness of the interlocks required by the Philadelphia law and to the administration of the code in force in that city. The contrast in the experience with freight elevators, for which interlocks are not required, is very marked. The following requirements cover, in general, the desirable interlock features to be incorporated into an elevator code. It is urgently recommended, however, that the latest issue of the A. S. M. E. Elevator Code be consulted as to the exact wording; so that greater uniformity may be secured and that advantage may be taken of criticisms of earlier drafts of that code. V. SUGGESTED REQUIREMENTS TO BE INCLUDED IN AN ELEVATOR CODE The following definitions and specifications are submitted as tentative suggestions suitable for inclusion in an elevator code and are intended to give adequate protection and to mitigate the hazards previously set forth. From the evidence thus far accumu- lated the use of a hoistway-door interlock seems advisable, and this protection should be required on passenger elevators at least. The material presented in this report was gathered for use in formulating an elevator safety code and was placed at the dis- posal of the joint committee which was formulating a national code for the American Society of Mechanical Engineers and with which the Bureau of Standards has been cooperating. The definitions and specifications which follow and those tentatively adopted by the committee agree in all essentials. Any evidence tending to support or modify these conclusions will be gladly received by the Bureau of Standards. 1. DEFINITIONS HOISTWAY-DOOR INTERLOCK. A hoistway-door interlock is a device the purpose of which is i. To prevent the normal operation of the car, except by the use of a leveling device, unless (a) (door unit system) the hoist- Elevator Interlock Report 29 way door opposite which the car is standing is locked in the closed position; or (6) (hoistway unit system) every hoistway door is locked in the closed position. A hoistway door or gate shall be considered locked in the closed position when within 4 inches of the full closure. If in this position, and any other up to full closure, the door or gate can not be opened from the landing side more than 4 inches. Interlocks which permit the starting of the car before the door is fully closed shall be so equipped that except when the door is locked in the position of full closure the door or gate can be opened from the landing side to the position approximately 4 inches from full closure. 2. To prevent the opening of a hoistway door (except by use of a key) from the landing side when the car is passing a landing; except when the car-control mechanism is in the "stop" position. HOISTWAY-DOOR ELECTRIC CONTACT. A hoistway-door electric contact is an electrical device the purpose of which is to prevent the normal operation of the car except by the use of a leveling device unless (a) (door unit system) the hoistway door opposite which the car is standing is in the closed position, or (6) (hoistway unit system) every hoistway door is in the closed position. EMERGENCY RELEASE. An emergency release is a device the purpose of which is to make hoistway-door electric contacts or hoistway-door interlocks inoperative. 2. SPECIFICATIONS HOISTWAY-DOOR INTERLOCK SPECIFICATIONS. (a) The pre- vention of the operation of the car by a hoistway-door interlock shall not be dependent on the action of springs in tension nor solely upon the completion or maintenance of one electrical circuit. (6) The agency used to perform any interlocking function shall be such that even without lubrication of the mechanism the intended functioning of the device will be completely performed. (c) The locking of the hoistway door and the interlocking of the car control shall be accomplished by an interconnection between the parts of the device. This interconnection may be mechanical, electrical, hydraulic, or pneumatic. (d) It shall be necessary to accomplish the locking of the hoist- way door opposite which the car is standing before the car can be moved by normal operation. This paragraph applies to both the door unit and the hoistway unit system. (e) If without damage to, removal of, or interference with any part of the elevator or hoistway equipment the door opposite 30 Technologic Papers of the Bureau of Standards which the car is standing becomes unlocked, it shall be impossible to start the car by normal operation. HoisTWAY-DooR ELECTRIC CONTACT SPECIFICATIONS. (a) The prevention of the operation of the car by an electric contact shall not be dependent on the action of springs in tension nor solely upon the completion or maintenance of one electric circuit. The failure of the device shall manifest itself by preventing the starting of the elevator from the landing. (6) The device shall be such that without lubrication of the mechanism the intended functioning will be completely performed. (c) All live parts shall be inclosed. EMERGENCY RELEASE. (a) The emergency release shall be in the elevator car, plainly visible to the occupants and reasonably, but not easily, accessible to the operator. (6) To operate under emergency conditions, it shall be necessary for the operator to hold the emergency release in the emergency position. The emergency release shall be so constructed and installed that it can not be readily tampered with or "plugged" in the emergency position. (c) Rods, connections, and wiring used in the operation of the emergency release that are accessible from the car shall be inclosed to prevent their being tampered with readily. WASHINGTON, May 14, 1921. 14 DAY USE RETURN TO DESK FROM WHICH BORROWED LOAN DEPT. This book is due on the last date stamped below, or on the date to which renewed. Renewed books are subject to immediate recall. LD DEC 14 1970 3- t* 1158 * 2 7 1975 5 2 24lan'61 IF 3T5 REC'D LD DEC 18^88 JAN 3 1 1961 27 REC'D LD AUG 2 1963 JAN 2 2001 ft 1971 6 3 ffr REC'D ID FEB 17 71 -5PM 10 LD 21-50m-8,'57 (.C8481slO)476 v - 31 U.C. BERKELEY LIBRARIES CDOMlMOOSk ' ,. . - * UNIVERSITY OF CALIFORNIA LIBRARY ' s fti|S s0 ^V^*s .-/ trA\C *-'* "