The following text may have been generated by Optical Character Recognition, with varying degrees of accuracy. Reader beware!

level with the face of- the die, the number of alternations endured would have been sensibly less, and this + error may be approximately corrected by setting out a diagram like the low- ee sketch in Fig..9.. M is the line of maximum stress, Z the zero of stress. Draw a line along 'the plane of rup- Fig .7. DIAGRAM SHOWING NATURE AND STRESS- CYCLE OF THE WOHLER TEST. (7 Ox Ts [w | ture parallel with the plane of maxi- mum stress, then at any convenient angle draw a line from the zero of stress to the plane of maximum stress, on the point of intersection erect a perpendicular reaching the plane of tapture, Let the line: 4 2. be propor- tional to the alterations endured, then the result which would be registered, had the fracture taken place on the plane of maximum stress, will equal x T---v y, y being the point of intersection of the an- gular line with the plane of rupture. The author here hastens to admit that theo- retically the Wohler test is perfect, while that of the author is theoretically hope- lessly wrong, since it reproduces no known condition in engineering practice, nevertheless in many cases this 'test pre- . dicts the liability of material to fail in Fig. 8. DIAGRAM SHOWING NATURE AND STRESS edly LINES IN AUTHOR'S TEST MCS$s Practice when the Wohler test would com fidently indicate that the metal was relia- ble. This apparently improbable enuncia- _ tion, the author will presently prove ex- Perimentally by results obtained on un- known metals, over the preparation of which the author had no control, and dis- _ safe. TAE Marine Review 41 tinguished from each other at the time TABLE xX, the tests were made only by certain let- ee endured LeLS: ani ae and numbers. As comparative bend- (Under standard Ratio of ing tests made by the method of Woh- ea coe conditions.) Beet eel phorus, to alternat- ler and that of the author accumulated, No. percent, 1. 2. Mean. ing stress. it became evident that the former test 5 an ee re ais was merely a reflection of the elastic 3 -509 oe 128 100 37 limit; thus certainly reflecting danger of Speaking in round numbers, the Wohler test indicated that a mild steel containing 0.5 per cent phosphorus was, with equal stresses, 10 times as capable of resisting alternating stress as a steel containing 0.04 per cent phosphorus. The author's test, on the other hand, indicated that a steel con- taining 0.5 per cent phosphorus had about one-third the endurance of a steel containing only 0.04 phosphorus. The curves of the yield points of the two alternating tests of the three steels are plotted in Fig. 10, where it will be seen that the Wohler curve is sim- fraenire under excessive stresses, but in- sidiously concealing the danger from other causes, which determine fractures generally attributed to fatigue. In tact, no matter how: dangerously brittle the steel may be from chemical or physical causes, if such causes have co-produced a high elastic limit, the Wohler indicates the steel to be quite safe if stressed well short of that limit, when, as a matter of practical fact, it is inevitable that the steel must suddenly rupture' sooner or later under stresses theoretically quite For, perhaps, the strongest proot of the soundness of the foregoing state- ment, the author is indebted to his friend, J. E. Stead, F. R. S., who determined to test the points at issue on steels concern- ing whose highly dangerous character there could be no possible doubt. Mr. Stead accordingly made and tested at Middlesborough, a series of mild steels with ascending phosphorus up to about 0.5 per cent. Of course, to deliver such a steel as that last named for use in the construction of engine parts, could be the act only of a metallurgical madman. Mr. Stead's steels were forged from 6 in. to 2 in. square, and were practically identical in composition, except for their phos- phorus contents. -Their general analysis registered: Fig. 9. I ; 2 3 ilar in type to that registered by the yield point or apparent elastic limit, the author's curve being in an oppo- © site direction and indicating what is Per cent. ' Per cent. : Carbone: i... 030°. Silicon, 1223/2. 0.21 well known to be the mechanical ef- Miner eee 0.40 Silphuts.i3 es 0.058 In No. 1 steel the phosphorus was 0.041 fect of phosphorus on steel. per cent, in No. 2, 0.303 per cent, and in No. 3, 0.509 per cent. The static tests are embodied in Table THE -WOHLER TEST APPLIED TO COPPER | ALUMINUM ALLOYS. In the eighth report made to the Vill, : alloys research committee of the Lf ELE oe stitution of Mechanical Engineers by | Ss a bo a Dr HC. oH. Cappenter aug cB y S $3 "> 3. Edwards, from the National Physical fe 2 g 8 g< g Laboratory, it was observed by these a v4 a ae SG authors in connection with rolled bars @ as ae 8 8 2s that the added aluminum remained dis- @ Fa 504 $31 23.0 ~~ $30 _~--s Solved in the cold alloy till the cop- 2, 302 gd ae af ; ae per became saturated, namely, when "The 'Wobler tests. made by Mr. Stead about 7.5 per cent of aluminum was give the results set forth in Table IX, present. The micro strictarcs Of ee the stresses being + and -- 15 tons per @lloys, up to and including the sat- urated alloy, all consisted of a single constituent. presenting the ordinary al-_ lotrimorphic crystals of metallography, square inch, 7. e., a range of 30 tons. TABLE IX. : Ratio of resist- Reversals of ance to alter- Steel Phosphorus = but on the addition of aluminum in . nating stress. oe ae Le $1,000. 1.0 excess of 7.5 per cent a new dark--- 2 oe oe ae Ge etching constituent, presumably richin 3 aluminum, appeared, and more or less perfectly enveloped the crystals -of the saturated solution in a relatively brit- tle cement.. A set of these alloys, an which the copper contained from 0.10 _ The author's tests made by him at Sheffield University in complete ig-_ norance of the nature of the steels. registered the figures embodies in Ta- ble X.