September 1, 1921 



THE INDIA RUBBER WORLD 



891 



Pig. 11 



(iioral Ur* 7*«(. 



3000 



ITSO 



Sis. I 



Fl«. IS 



3000 

 CTSO 

 (900 



tooo 



J?M 

 ISOO~ 



72 79 84 90 96 



T«n*ll> 



itio 

 3000 



e7so 

 eaoo 

 uso 

 eooo 



17*0 

 1900 



Fig. 13 



12 le 24 30 3« «3 4Q S4 



L 



temperatures and in the dark, and test strips removed at inter- 

 vals over a period of a little more than five years. From these 

 curves it is to be seen that the accelerated-life test is some- 

 vkfhat more severe than the natural-life test, although there is 

 some degree of correspondence in the general shape of the 

 curves. 



The curves of Fig. 5 were made from data obtained as those 

 of Fig. 3, and, like Fig. 3, the curves show sharp deterioration. 

 In each of these cases, we should judge from the experience of 

 our laboratories that the compound thus made and vulcanized 

 would decay too rapidly in natural life. 



Fig. 6 shows a different compound and a different cure. It 

 is also chosen to show a wholly different type of aging curve. 



In this case the tensile strength and elongation have remained 

 at high levels, and one would judge this compound to be a good 

 one. In point of fact it is a remarkably good one, and while we 

 cannot show you the curves from natural life it has stood the 

 test of time in a highly satisfactory manner. 



The next diagram to which your attention is invited is Fig. 7. 

 Here have been plotted the data from the natural and short-life 

 tests and tensile strengths only. The compound is one of the 

 older type composed of natural grades of rubber and litharge 

 as the accelerator. The accelerated-life test again shows more 

 severe than the natural aging. These curves may be considered 

 typical, and one of the general conclusions may be anticipated 

 at this point, which is that when an accelerated aging curve 

 shows a type where the tensile strength is nearly constant for 

 the first three or four days and no severe decline following, it 

 is pretty general to find the natural-life test is at least as good, 

 and usually better. One must here warn those who may use 

 this method to gain experience in judging curve data, to follow 

 the method of comparison of unknown with known compositions 

 before predicating any considerable volume of output upon un- 

 compared data. 



Fig. 8 is intended to describe a relatively close comparison 

 of accelerated and natural-life tests of a lower-grade composi- 

 tion. 



Fig. 9 is the data of a lower-grade compound the natural-life 

 test for which was run but 24 months. The parallelism is fairly 

 close. The fact of the natural tensiles being under the others is 

 probably due to difference in the original mixings. 



The next four. Figs. 10, 11, 12 and 13, are from a compound 

 of the type, plantation 50, caucho ball 50, sulphur 8. The first 

 shows how one may be misled, due to irregularities in tensile 

 strength. Of course, in making any curve to relate functional 

 data where irregularities exist one should follow a physical 

 method of averaging, but the average rubber chemist has neither 

 the time nor the facilities for so elaborate a method. We, there- 

 fore, are in the habit of disregarding the obviously low figures. 

 In the four diagrams one judges that as the time of vulcaniza- 

 tion increases, the rate of decay increases and the parallelism 

 with the natural life becomes more apparent. This we conclude 

 to mean that the short or accelerated aging test can and does 

 show us, if nothing else, how to determine the correct degree 

 of vulcanization necessary to insure the maximum possible 

 length of natural life. 



Fig. 14 illustrates the effect of the life test, described in the 

 foregoing, upon the stress-strain diagrams that have been of 

 late discussed so widely. The compounds are of the inner tube 

 variety, that is to say, composed of rubber, sulphur and ac- 

 celerator. This chart is of a high-grade type and shows the high 

 resistance to aging. There is, as will be observed, but little 

 change in the form of the curves. The compound has become 

 stiffer and weaker. 



The last diagram is Fig. IS, in which are shown the results of 

 the aging test upon a similar compound, but one which resisted 

 the effect of the heating to a lesser degree than the one shown 

 in F\(i. 14. There is very little that we can say regarding the 

 effect of aging on the stress-strain curves. Our work has had 

 to do more largely with the effect upon that more basic property, 

 tensile strength. 



CONCLUSIONS 

 In the foregoing paper an attempt has been made to describe 

 a test the use of which might permit one to predict the probable 

 effect of natural aging upon culcanized rubber. No one realizes 

 better than the authors, the shortcomings of the presentation 

 and the incompleteness of the proofs offered. So much in our 

 work upon rubber goods is approximate and incapable of any 

 real degree of scientific proof that it is with a feeling almost akin 

 to apology that it has been presented at all. Yet we have worked 



