:IRON AND STEEL UNDER CYCLICAL VARIATIONS OF STRESS. 53 



tension until within a short distance of the yield. Specimen No. 2, fig. 1, showed, 

 also, that in compression the elastic limit was higher than 17 tons per sq. inch. The 

 apparent elastic range was therefore greater than 42 tons per sq. inch as against the 

 26 tons per sq. inch which was maintained against repetitions. 



Experiments Nos. 2, 3, 4, 5 and 6, Table III., refer to a specimen in the condition 

 of Specimen No. 2 after 6000 reversals. 



Following a final load of 18 '26 tons per sq. inch in tension, the specimen was 

 heated in boiling water for 15 minutes. Test No. 2 showed that the specimen was 

 still elastic at the same load. In compression the limit was reached at about 8 '5 tons 

 per sq. inch and the load was continued to 13'28 tons per sq. inch. Recovery was 

 again produced, and it was then found that the elastic limit in compression had been 

 raised at least to 13 - 28 tons per sq. inch, but that the tensile elastic limit had fallen 

 to 13'0 tons per sq. inch. By alternately heating and testing, the elastic limits were 

 moved about very considerably, but always with the condition that if the tension 

 limit was raised the compression limit was depressed, and vice versa. 



Both limits cannot be determined at the same time, but Test 6 shows that the 

 elastic limit was found at the last stress imposed before heating, and the elastic 

 ranges for the four previous cases can now be seen to be as nearly as possible 26 tons 

 per sq. inch, i.e. equal to the elastic range found by repetitions. The agreement is 

 not complete, as no extension occurs to correspond with the " permanent extension " 

 of figs. 2 and 3. 



Specimen 5 was immersed in boiling water for an hour and a half, without any other 

 treatment between this heating and the repetitions of stress recorded in fig. 3. The 

 elastic limit in tension now appeared at 20 tons per sq. inch, the loading being carried 

 to 26 '6 tons per sq. inch. On heating and retesting the limit was again found at 

 20 tons per sq. inch, and so on for Tests 9 and 10. The compression limit found 

 in Test 11 is zero, but is uncertain, due to the smallness of the readings. The elastic 

 limit of 20 tons per sq. inch, approximately, is somewhat less than the 22 tons per 

 sq. inch estimated from the repetition experiments. 



The two specimens just described illustrate BAUSCHINGER'S two theorems very 

 clearly, the first showing that for nearly equal stresses the elastic limit in tension can 

 only be raised by a corresponding drop in the compression limit, whilst the second 

 shows that a limit exists above which the tension limit cannot be raised, so long as 

 the stress is entirely removed in each cycle. 



Specimen No. 7, fig. 5, of the axle steel should agree with Specimen No. 5, as the 

 last load applied was zero. Test 12 did not show any limit until 26'6 tons per 

 sq. inch had been reached, and then the departure from the elastic line was small. 

 The compression limit appeared to be 5 tons per sq. inch, or the elastic range was 

 apparently 31 or 32 tons per sq. inch. As the method of repetition gave 19 tons per 

 sq. inch range, the specimen was no better than in its primitive condition. 



The experiments for Swedish iron, made with nearly equal stresses, agreed with the 



