give little or no attention." The fact that a diamond truck has the top bar in compression and the middle bar in tension, and the result of this is shear on the bolt, is well known. It would not be at all complimentary, I think, to railroad mechanical engineers if they had disregarded this fact, because a freight truck is the most numerous structure of any; it is the structure that is used in larger numbers on the railroad than anything else, and therefore it should be a proper subject for correct design and economical construction.

All the latest practice has plainly shown that there is no advantage. in turning up the ends of the middle bar; that practice was used fifteen years ago, and was discarded as unnecessary. It has been found that in the rough work of a car shop such a fine machine fitting as would be required to make the shear of bolts coincide with the butting of the bar against the other end is not obtained; the result has been that you find freight trucks today with straight bars just as they are in the test, and I have no doubt that as the result of long experience and investigation as well, it is, all things taken into consideration, the best design.

MR. W. A. SMITH (Cloud Steel Truck Co.): The truck frame referred to in the paper of Prof. Kerr is the M. C. B. recommended practice for cars of 80,000 pounds capacity. It will be seen by the diagram that the elastic limit was reached at 45,000 pounds, and it is very evident that the arch bars were put out of serviceable condition long before the point was reached where the bolts were sheared. A certain master mechanic originally informed me that he had had a template made and had been testing all cars of 80,000 pounds capacity that came into his yards, and found that in a very large proportion of them the arch bars had taken sufficient permanent set to tilt the boxes toward each other at the top. It appears from this, that the arch bars reach their elastic limit altogether too easily and that there will be much resulting trouble from hot boxes, etc.

MR. HUBBELL : I am not able to say anything on this subject which will be perhaps of any special benefit, but I heard a paper read under the "Topical Discussions," at Saratoga, eighteen months ago, that interested me very much, about the arch bar truck, and since that time I have been watching the arch bar truck as perhaps never before; and I have been surprised at the number of instances I have seen, under cars of various stenciled capacities, where the ends of the arch bars over the oil boxes were "cocked up," or, in other

words, where the arch bars had passed the elastic limit. I think perhaps, most of our members will be more or less interested in that same subject if, they make a specialty of it and watch it.

MR. F. H. CLARK (C. B. & Q.) : I agree with Mr. Stark, that the failure recorded by Prof. Kerr is not a very common one. I have seen very few cases of sheared journal box bolts. I saw some time ago a record of failures of truck side frames, and out of the entire number there was not a failure of the journal box bolts. The failures covered by the record were distributed in the ratio of about 77 per cent in the lower bend of the bottom arch bar, 2 per cent in the top bend of the top arch bar, 7 per cent in the bottom bend of the top arch bar, that is, out near the journal box bolt, and 14 per cent in the column bolts.

Now, I do not know to what extent those ratios will hold good, but I am of the opinion they are about right. In looking into this question some time ago, I investigated quite a number of trucks in common use, some of which were considered very good trucks and some of which had bad reputations; I made stress diagrams for each truck, calculated the stresses in the arch bars and column bolts, and in calculating the stresses in the lower arch bar it was at first thought best to consider the section at the column bolt, as about one-third of the material was cut away at that point by the bolt. Upon looking a little further into the matter, it was found that we had very few failures of the arch bars at that point, the most of them being an inch and a half or two inches away, at the bend. As the stresses in the diagonals were greatest, we finally decided to base our stresses on them, and on this basis we found that in most trucks which apparently gave very little trouble, the stresses in the lower arch bars under a dead load were 2,800 to 2,900 pounds per square inch in the lower arch bars and 2,200 to 2,300 pounds per square inch in the top arch bars. These figures seemed ridiculously low; almost anyone would feel sure that the stresses might be considerably greater without hurting the trucks very much, and I came to the conclusion, after studying the thing over, that the life of the lower arch bar depends more upon the column bolt and the column bolt fastenings than upon any other thing, and that in order to make use of light material in the lower arch bar, it would be necessary to provide, in the first place, column bolts of sufficient size to safely carry the load, they having to carry practically all the load, and, in the second place, arrange a fastening of some kind at the lower end that would prevent the nuts

from losing off or slacking off; that the breakage of arch bars is not generally due to legitimate stresses or stresses ordinarily considered by the designer, but were due to nuts loose or lost off, which, of course, would greatly increase the stresses at the bends at the corner of the end castings or columns, on account of the bending and vibrating action that the arch bars there have to withstand.

THE PRESIDENT: If there is no further discussion, I will ask Prof. Kerr to close the discussion.

PROF. KERR: I find that some points have been brought out in the discussion which show that I have unwittingly recommended, in Fig. 5, a patented article; I was not aware of it, and I see that makes it all the worse if it has been tried and failed.

I purposely commented rather severely on this question of shear, because I wanted to bring out discussion on that very point, and I am obliged to my friend Mr. Clark for the data which he has given in regard to the failures of these arch bar trucks. I presume you gentlemen will remember, in the study of beams, that they were generally considered as solid beams, and that your ideas as to the action of shear were at first, at least, somewhat vague. I really believe, notwithstanding what has been said here, that our experience in testing these beams shows that the ideas of, at least, some designers are still vague, because ultimate failure, as I said, invariably resulted in the shearing of these bolts which affect the moving of one part of the beam over another part. The ideal arch bar truck would be, of course, so proportioned in its different parts that, like the "OneHorse Shay," it would be equally liable to fail or equally strong at all points.

On account of using cheap iron, with a low cross bending strength, it is perhaps advisable in practice to use considerably heavier members in tension and compression than are necessary, at least so far as tension is concerned. Not one of the tests showed any weakness so far as tension of any members of the truck were concerned. It is in the cross bending of these that failure occurs, as Mr. Clark has brought out, and in the shearing of the end bolts.

Adjourned.

List of duplicate copies of the Proceedings of the Western Railway Club which are in the David L. Barnes Library and are for sale or exchange :

1888-Sept. and Oct. (bound in one).

1889-Oct., Dec.

1890-Feb., Sept., Oct., Nov., Dec.

1891-Jan., Mar., Apr., May, Oct., Nov.

1892-Jan., Feb., Apr., May, Sept., Nov., Dec.

1893-Two complete volumes; also single copies of Jan., Feb., Mar., May, Sept., Oct., Nov., Dec.

1894-Jan., Mar., Apr., May, Oct., and Index of subjects discussed by the Club.

1895-Apr., Sept., Oct., Nov., Dec.

1896-One complete volume, also single copies of Mar., May, Sept., Nov.,

Dec.

1897-Jan., Apr., May, Sept., Nov., Dec.

1898-Jan., Feb., Mar., Apr., May, Sept., Oct., Nov., Dec.

1899-May.

Correspondence relating to the Library donations of books, periodicals, etc. for the files, or duplicates sent for exchange should be addressed to H. de K. Woods, Librarian, 1750 Monadnock, Chicago.

Special Notice

The David L. Barnes Library of this Club, at 1750 the Monadnock, Chicago, is open for the use of members and their friends, and we hope it will be used freely It is open on week days from 9 a. m. to 5:30 p. m., except on Saturday, until 3 p. m. Books must not be removed from the Library, but the Librarian will assist visitors in finding information and will promptly reply to letters from out-of-town members desiring information from the Library. Donations of books and technical publications will be gratefully received.

TRUSTEES TO THE LIBRARY.