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

TAE Marine REVIEW Fig aa. AUTHORS ALTERNATING STRESS. CURVE OF COPPER .ALUMINIUM ALLOYS i) z Q 1400|\--= a z ° Oo ie 1200; Qa z < - wo 1000}--# a =z = oO w | § 800}-2-- z Ww "o < ° 600 | os c uw ke al q- 400|--S- Hs o- e s ee 200 : : 1 2 3 4 § 6 10 PERCENTAGE. OF ALUMINIUM IN COPPER use. Its most noticeable chemical "A" steel were 0.26 per cent, "B" steel feature was an excessive percentage of phosphorus. In alternation it be- haved as shown in 'Table XIII. TABLE XIII. Ratio of reversals below the speci- Alternations endured under a standard tests. fication, per cent. 2 46 85 3 46 85 4 50 83 Some time ago, to test this ques- tion, the author made the following experiments on acid steel 34 in. round bolts, both as rolled and in the rolled and drawn condition. ~The chemical composition of the four steels employed was practically constant, except in carbon. The sili- con averaged 0.12 per cent, manga- nese 0.85 per cent, the phosphorus 0.044 per cent, and: the sulphur 0.041 per cent. The carbon contents of the Fig. 13. AUTHORS STANDARD ALTERNATING STRESS AND YIELD POINT CURVES ON ROLLED AND ROLLED AND DRAWN 341N. BOLT STEELS. Cae aE 0-26 % CARBON 1G ST CARBON STEELS 0}41% CARBON ROLLED 049% CARBON 300 ie acne | ARBON BON STEELS 4.1% CARBON 200 Te ROLL AND Dh % CARBON pt Nee ee ama Saree REVERSALS OF STRESS ENOURED 160 re ed } 40 : Ls 20 2s 30 - 35 APPARENT ELASTIC LIMIT IN TONS PER SQ. INCH. e 0.31 per cent, "C" steel] 0.41 per cent, and the "D" steel 0.49 per cent. The tests are set forth in Table XIV. TABLE XIV, HOT ROLLED SET. n ne 5 fe bd : ' os Ho a, % = id oe i 3 27 4 dg 3 u a. g g . 3 © 5 = : . Ea Sa Od od a oe ¢ ob 30 8 oo BR eS ee ee eS S) Se ee a 4 A 0.26 24.1 35:0 34.5 61.6 B 0.31 O537: 37.6 31.8 59.6 c 0.41 28.9 42.7 26.5 >. 480 D 0.49 29.9 46.9 25.5 50.8 ROLLED AND DRAWN SET, A 0.26 36.7 38.9 19.0 =50.8 B 0:31 Sa: 41.9 17.5 47.2 C 0.41 38.9 46.6 16.0 40.4 D 0.49 38.0 52.4 15.0 42.0 The yield points had. thus~ been raised by the drawing operation to the following extent: A 12.6 tons per sq. in. B 9.4 tons per sq. in. C 10.0 tons per sq. in. D_ 8.1 tons-per sq. in. Under -the author's standard alter- nating tests the results embodied in Table XV were registered: Hot rolled. ; 156 Rolled and drawn. TABEE Sev. wn ' 5 3 4 od 2 6 Oo q . og So g a ao 3a 2 ay a mH: 5 28 S oS Ba ee Sa LO > < oO i 0.26 24.1 346 Hot rolled. A 0.26 36.7 248 Rolled and drawn. B 0.31 PAN Ip 344 Hot rolled. B 0.31 35.7, 222 Rolled and drawn. Cc 0.41 28.9 336 Hot rolled. Cc 0.41 38.9 206 Rolled and drawn. D 0.49 D 0.49 & DO 90.5 ow Ww bo oOo Thus the mean endurance of the hot-rolled steels was 337 alternations of stress, that of the rolled and drawn 43 steel was 208 reversals, a mean differ- ence of 129 alternations. The rolled and drawn below the specified minimum endur- ance, while the hot-rolled steels were 12 per cent above the specification. It should also be remembered that the test bars were turned to 3% in. diam- eter, thus completely removing the skin of the bolts on which the me- chanical injury caused by drawing would be at its maximum. The tests are plotted in Fig. 13, in which it will be seen that the erratic effect of. the cold drawing has disturbed anything like regular co-ordination 'between the percentage of carbon. present and the alternations endured. In drawing to a close his reference to the Wohler test and its indications, the author cannot do better than quote a. re- Fig, 14. markable paragraph from an able pa- per on "The Production and Thermal Treatment of Steel in Large Masses," by Cosmo, Johns, of Messrs. Vickers,- Son & Maxim, Ltd, read before the May meeting of the Iron and Steel Institute, 1904 (Journal of the I. S. I., No. 1, page 72, 1904): "Tt may have been noticed that en- durance tests have not been included in the tabular results -given above. The reason is that in no instance have they shown any want of agreement with the static tests, while the mere recital of the number of revolutions a particular test-piece ran before it broke, would convey no information unless they could be compared with the results obtained on other machines under similar conditions, with the same ratio of load to elastic limit. The endurance is proportional to, and depends upon, the elastic limit." ~ Finally, if steels were liable to frac- ture only under the fatigue phenom- enon or Wohler, it is hardly an un- reasonable hypothesis to put forward steels were 43 per cent -