Mr. Wickhorst.

Mr. Kinkead.

Mr. Wickhorst.

MR. MAX. H. WICKHORST.-Such tests as I have made indicate likewise that piled iron will stand more vibrations than bloom iron. It is a question though, whether if both materials were subjected to the right kind of heat treatment, that difference would actually exist in the finished material. In other words, it is just possible that one material may have been finished at a much lower temperature, although in that case we ought to be able to discover the difference in the reduction of area.

MR. J. A. KINKEAD.-I put in a vibration-test machine some time ago, reducing all the variables I possibly could. I found that I could not duplicate results sometimes within 200 per cent. noticed the same thing in the experience of others on material I furnished. Some would get very good, others very poor results. The iron was as nearly the same as it was possible to get it. I have taken five bars, and made five tests on each, getting 200 to 800 vibrations on pieces from the same bar. I think, therefore, that the matter of vibration tests should be taken up very thoroughly and check tests made on the same bar, or other means adopted to find out whether we can duplicate results within satisfactory limits, of, say, a few per cent. On the tensile strength we can get check results within very small limits. Until we can devise a machine that will give results on check tests within satisfactory limits, I think it would not be well to adopt a vibration requirement. Otherwise every testing engineer will have a vibratory-test machine of his own, the design of which he thinks better than any other. The result will be that anybody proposing to sell iron to a railroad company who wants to know how the material will test will have to have a machine of each kind, and may require twenty different machines. At present I think there are about five. The results from one machine do not compare at all closely with the results from any other.

MR. WICKHORST.-For the purposes of investigation an elaborate vibratory machine may be required, but for simple

inspection purposes a machine can be devised almost anywhere, Mr. Wickhorst. and, as I look at it, it is one of the functions of our Committee M

to recommend some standard, simple method of making vibratory

tests.

As regards variations in results, some ten or twelve years ago the Burlington Road made quite a series of vibration tests, and the first tests were made by simply screwing the staybolt into a plate and hammering over the end to simulate service conditions. We then made special clamps to hold the staybolt ends, and even under these conditions there was a certain amount of play and the axis of the bolt where clamped was moved through a small angle. In order to overcome or minimize that we simply put an extension on the bolt so as to hold it at two places, six inches between the clamps, and, in that way, even if the clamps were not perfectly tight, still the axis of the bolt could not move very materially at the place where the bolt was held. With this it was possible to get somewhat more consistent results, but even then there was considerable variation. I think these variations are due mostly to variations in the material, which, perhaps, are local, but nevertheless represent actual variations in the material.

MR. KINKEAD.-In my vibration tests I eliminated all variables Mr. Kinkead. as far as possible. I used 6-in. lengths, and held the bar between tool steel dies. The dies were held by four bolts set by means of a long wrench and spring balance as recommended by the Baltimore and Ohio Railroad Company. To give an even stress on the bolts, the back bolts were set, while the stress was off. The pieces were then reversed, and the front bolts set. The pieces were cut by the same die and all taken from the same bar of iron. The iron used did not show an average of more than 1.5 per cent. elongation, nor more than 300 or 400 lbs. in tensile strength. Even with iron of such uniform character, results varied from 200 to 800 vibrations.

One more point. Under service conditions the vibratioen in the boiler will not produce stresses in staybolts exceeding ths elastic limit of the metal. Iron can not be stressed beyond the elastic limit without injury. With the vibratory test-machine the stresses exceed the elastic limit.

MR. WICKHORST.-Mr. Kinkaid states that in service the Mr. Wickhorst. staybolt can not be bent beyond the elastic limit. I think he is

Mr. Wickhorst. much mistaken. In the ordinary process of firing up a locomotive boiler, starting with cold water, the water at the crown sheet is heated first, and I have found by actual tests a difference of temperature of as much as 200° F. during the process of firing up. Again, during the process of injecting cold water into the boiler with the ordinary check-valve, the cold water flows in underneath the water already in the boiler causing a drop of, perhaps, 100° F. at the mud ring. Keeping these facts in mind, we can readily imagine movements of the staybolts amounting at times to to in. in some places.

Mr. Wille.

MR. WILLE. The vibratory tests are as specific as the drop tests, but they have hitherto been made in a crude way. The chief difficulty has been the difficulty of getting a very tight clamp so that the bolt does not vibrate in the clamp. Inasmuch as bolts are vibrated in service through a definite angle, it has been customary to make the vibratory test by specifying a vibration through a definite angle in place of specifying that they should be vibrated under a given fiber stress in flexure. This corresponds with specifying a drop test without limiting the maximum deflection. Some irons may require a load of 3,000 lbs. to deflect in., while others may require 6,000 lbs. The latter will be subjected to a higher fiber stress than the former for a given deflection.

One curious thing about the vibratory test of staybolts is, that iron that shows a maximum vibration at a small deflection shows a minimum vibration at a large deflection; and, vice versa, iron that shows a large number of vibrations at a large deflection shows a small number at a small deflection. This points to the fact that it will be necessary to specify the fiber stress and the amount to which the bolt should be strained in bending. If this be done, the results will be concordant, and even more definite than many of the bending, drop tests or impact tests.

A PRELIMINARY REPORT ON THE EFFECT OF COMBINED STRESSES ON THE ELASTIC

PROPERTIES OF STEEL.

BY EDWARD L. HANCOCK.

For some time engineers and others have been interested in the effect of combined stresses on the materials of construction. Lord Kelvin in the preparation of his article on Elasticity for the Encyclopedia Britannica, had a series of tests made on piano wire. The wire having sufficient weight to hold it straight while suspended was subjected to a torque at the bottom; when an additional weight was added it was found that the elastic limit of the wire in torsion was lowered. From these tests it seemed possible that the same thing would be true in the case of compression-torsion (loaded columns), although no tests of this kind were made.

Mr. J. J. Guest, in England (see Proc. Physical Soc. of London, Sept., 1900), carried out a series of tests to obtain the effect of combined stresses on ductile materials, the materials used in the tests being wrought iron, mild steel, copper and brass. The results of these tests also lead to the same conclusion, that the elastic properties are lowered when the materials are subjected to combined stresses.

The slight knowledge on this subject and the entire absence of any information available for engineers have inspired the writer to carry out a series of investigations that should give some information immediately available for practical work. The series contemplated includes the following:

(a) Tests of steel and iron solid rounds and hollow tubes in tension while under torsion.

(b) Tests of steel and iron solid rounds and hollow tubes in torsion while under tension.

(c) Tests of steel and iron solid rounds and hollow tubes with increasing tension and torsion.

(d) Tests of steel and iron solid rounds and hollow tubes in compression while under torsion.

Thus far only a part of series (a) on solid steel rounds has been carried out and the results given in this report are taken from these

tests.

Apparatus for Testing.-The difficulty of making tests of materials under combined stresses lies, principally, in the fact that no machines are available for such tests. The apparatus used in the tests under consideration is shown in Fig. 1.* It consists

of two specially constructed heads fitted to an ordinary 100,000pound tension testing machine. Each of these heads consists of a flat cast iron base fitting into the slot in the head of the testing machine provided for the insertion of the wedges, as ordinarily used. The outer side of this casting is finshed to provide for three * Acknowledgment is made to the Engineering News for the cuts used in this paper.