September 1, 1921 



THE INDIA RUBBER WORLD 



889 



strain curve. In must of their experiments a normal and not where a lesser number of tests are to be made and where 



an accelerated aging period is used, the time in most cases be- 

 ing at least a year. De Vries, however, heated the vulcanized 

 sample at 72 degrees C. or 65 degrees C. for 12 to 16 days. 

 The action of light as a factor in accelerating oxidation was 



studied by dc N'ries and more recently by Porritt". 



standard type of electric drying oven is available, this can be 

 adapted to such work, provided a mechanical agitator is placed 

 inside the oven to insure an even temperature throughout, and 

 a constant supply of fresh air heated to 160 degrees F. is main- 

 tained throughout the test. As an example, the type known as 



Fig. 2. Plan .'\Nn Elevation of Age-Testing Apparatu.s 



The accelerated life test as worked out and put into practice 

 at The B. F. Goodrich Co. was described at the September, 1916, 

 meeting of the American Chemical Society"". 



The first work done in the laboratories of The B. F. Goodrich 

 Co. vn accelerated age tests was in the fall of 1907, at which 

 time a suggestion came to us from Dr. Van Der Linde, of the 

 Gutta Pcrcha & Rubber Co., Toronto, Canada, who used a very 

 fast aging test. He performed it upon three small pieces of 

 rubber. The samples to be tested were put in an air bath of 

 heated air about 140 degree? C. for a period of one, two and three 

 hours, taken out, and examined for cracking, hardening, or to 

 ascertain whether they were becoming soft. This method was 

 not parallel with actual aping in any particular and we came 

 to the following : 



PRESENT METHOD OF MAKING TESTS 



The age-testing apparatus used in the present method is shown 

 in Figs. 1 and 2. Into this oven, heated air was blown at a 

 temjieraturc of 160 degrees F., care being taken to see that a 

 continuous supply of hot fresh air was added during the entire 

 aging period. A number of samples 3.'32-inch thick were pre- 

 viously cut in the form of our standard test strip and placed in 

 this air bath. The test strips were suspended by punching a 

 hole in the enlarged end, passing a wire through and placing 

 between each strip a rubber disk about the same thickness as 

 the test piece, to keep the individual strips separated and allow 

 a free circulation of air so tliat the entire surface was exposed. 

 The air was then started circulating and the test continued for 

 a period of two weeks, taking out three samples each day. These 

 were allowed to stand for 24 hours at room temix-rature until 

 they reached a state of equilibrium, after which they were test- 

 ed for tensile strength and elongation. 



While we have referred to a particular type of oven in 

 which to perform these aging tests, we wish to point out that 



=1 Journal of the Society of Chemical Industry. Vol. 3i. 872-4, 1916; 

 Vol. 37, 280-4T, 305-6T, 340-2T. 191S; Vol. 39, 826A, 192;1. The India- 

 Rubher Journal. Vol. 52, 679-80. 79-1. 1916; Vol. 53. 220-2. 366, 1917. 



=2 The India-Rubber Journal, Vol. 53, 101, 1917; Vol. 57, 77, 1919; Vol. 

 61, 87-90, 1921. 



=3J(mrnil of the Societv of Chemical Industry, Vol. 38. 339-45T, 1919. 



=<The India-Rubber Journal, Vol. 52, 615-20, 757, 795, 1916; Vol. S3, 

 327-8, 1917. 



»The India-Rubber Journal, Vol. 60, 1159-62. 1920. 



=«The India Rubber World, Vol. 55, 127. 1916. 



the Freas electric ovens, equipped with electric agitator on the 

 inside, has proved most satisfactory. 



The data plotted in curve form give a time-decay curve of the 

 compound. For a long time these were always run in com- 

 parison with a standard compound, that is, at least two sets of 

 samples were run at the same time. Thus, by comparing the 

 curve of a compound the age of which we knew, and the curxe 

 of a compound which we did not know, we obtained a reliable 

 indication of the service to be expected from this new com- 

 pound. It is worth while to point out that we consider this 

 purely a practical and not an ultimate test. Our experience, 

 however, based on many thousands of these curves, indicates 

 to us that this method of determining the approximate aging 

 of a commercial rubber product is reliable, and is of great Vrilue 

 as an aid to a compounder in a manufacturing plant, pemiitting 

 him to study the aging of the composition of compounds. 



No attempt has been made until recently to determine the 

 probable chemical reactions involved. It seems reasonable to 

 assume, however, that at the temperature employed and with the 

 current of fresh air passing over the samples, an tixidation oc- 

 curs, forming one of the unstable peroxides and at the same time 

 causing an increase in the coefficient of vulcanization, similar 

 to that resulting from the aging of rubber under normal condi- 

 tions. 



Having now described the method by which the test is per- 

 formed, let us proceed to examine a few curves taken from the 

 records of our laboratory. We have many thousand curves, and 

 we have chosen a few of those which illustrate certain typical 

 characteristics and of which natural aging data were available. 



CURVES SHOWING TYPICAL AGING CHARACTERISTICS 



The compound used in Fig. 3 is one of the older types so com- 

 iiKiti in rubber thread. This curve shows the change in the 

 elcingation and in the tensile strength over a period of seven 

 d;..\s. The curves are sharp and the deterioration was rapid. 

 Fig. 4 shows the decay of the same compound and cure, although 

 tested at a different tiine and upon different strips. The heavy 

 line is the curve plotted from the data taken from the strips that 

 had been subjected to the aging cabinet described in the fore- 

 going. The broken line is the plot from the data obtained from 

 strips that had been filed away in a dark room kept at ordinary 



