412 



THE INDIA RUBBER WORLD 



[May 1, 1913. 



Notes on Tension Tests of Rubber. 



By P. L. ll'ormclcy. United States Bureau of Standards. 

 A PAPER READ AT THE TIIIkD INTERNATIONAL RUBBER CONFERENCE, IIEr.D IN NEW YORK. 1912. 



THE wonderful development in the rubber industry and the 

 enormous demand for rubber products with their constantly 

 multiplying and broadening fields of application, make 

 it evident that the time is now ripe for the heartiest co-opera- 

 tion between manufacturers and consumers, with a view to estab- 

 lishing uniform methods of testing and, as far as possible, uni- 

 form specifications for goods intended for the same purpose. 

 Uniformity of methods is absolutely essential to an intelligent 

 comparison of the work of different laboratories, and a more 

 general interest in the standardization of specifications and 



' ^"S r. L. W'oK.MELEV. 



methods of testing would result in a substantial benefit not 

 only to the reputable manufacturer and large consumer, but 

 also to the public at large. 



In testing materials in general, the refinement of the methods 

 used should be determined in each case by the nature of the 

 material tested, and in the speaker's opinion there is no reason 

 why rubber should be an exception to the rule. 



Now as regards testing machines and apparatus used in deter- 

 mining the physical properties of rubber, a most interesting 

 field is presented, in which there is room for many improvements. 

 However, when one considers the rather large and apparently 

 unavoidable inherent variation in the physical properties of rub- 

 ber, it seems that in the present state of our knowledge, uni- 

 formity of methods is more to be desired in testing than extreme 

 accuracy and refinement in the matter of taking measurements. 

 This of course is particularly true in the case of commercial 

 or routine testing where the goods have been manufactured 

 to meet the requirements of specifications. Unfortunately speci- 

 fications are not always as definite as might be desired, never- 

 theless, the adoption by both parties concerned of uniform meth- 

 ods of testing, would remove the cause of numerous misunder- 

 standings between manufacturer and consumer. 



To discuss in a comprehensive way the various methods that 

 have been proposed for testing the physical properties of rubber 

 would require a more thorough knowledge of the subject than 

 I possess, and far more time than is at my disposal today. 

 There are, however, certain points in connection with methods 



and machines used in testing the tensile properties to which I 

 should like to ask your attention, with the hope that sufficient 

 interest may be aroused to bring out a discussion and an ex- 

 change of ideas and opinions on the subject. 



In regard to the most desirable form of test piece, the ring 

 and the straight specimen, each has its particular advantages, and 

 considerable difTerence of opinion is being expressed as to the 

 relative merits of the two. The method which is usually fol- 

 lowed in preparing rings involves several distinct operations 

 which, if carefully carried out, consume quite a good deal of 

 time. On the other hand the straight specimen may be cut out 

 by a single stroke of the die, the operation requiring only an 

 instant. Rings, when accurately cut, undoubtedly show uniform 

 results, but on account of the varying stress over the cross- 

 section, this method does not give the true value for tensile 

 strength. An analysis of the distribution of stress in a ring 

 shows that near the breaking point the tension decreases at an 

 approximately uniform rate from the inside to the outside surface. 

 If this variation in stress were the same for all rubbers, the 

 ring method might be accepted as giving a true relative measure 

 of strength, but tension curves plotted for different rubber com- 

 pounds show that this is not the case. 



The ring lends itself very rapidly to the automatic measure- 

 ment of elongation, and in this respect its advantage over the 

 straight specimen is apparent. It may be noted, however, that 

 since the strength of rubber is not the same in all directions, 

 being greatest in the direction in which the sheet has been rolled, 

 the ring gives a smaller ultimate elongation than is shown by 

 a straiglit specimen. This difference in strength and elongation 

 by the two methods is not always found, but the results of tests 

 which I have made on twenty different compounds show a gen- 

 eral tendency toward higher values for straight specimens, par- 

 ticularly in the case of high-grade compounds. 



Objection to the straight test piece is often raised on the 

 ground that specimens fail by tearing at or near the grips. This 

 difficulty may be overcome by gradually increasing the width of 

 the test piece at the ends, and by using a grip that will not 

 bruise or injure the rubber by excessive pressure. A form of 

 grip which the writer has found to be exceedingly convenient 

 and perfectly satisfactory consists of a series of circular discs 

 mounted eccentrically, and in such a way that each disc acts 

 independently in pressing against the test piece. By this means 

 the gripping pressure holding the specimen is automatically in- 

 creased in proportion to the applied tension, and furthermore, 

 the pressure is uniform over the entire width of the grip even 

 in cases where the thickness of the rubber varies. 



For measuring the applied load it would appear that prefer- 

 ence is usually given to some form of dead weight or pendulum 

 machine, the spring balance or dynamometer being often looked 

 upon with disfavor. 



It is the policy of the Bureau of Standards to encourage as 

 much as possible the testing of rubber goods and for this reason 

 our testing machines have been constructed with a view to the 

 greatest simplicity consistent with a reasonable degree of 

 accuracy. For determining tensile strength and elongation we 

 use an ordinary twin spring dynamometer attached to the upper 

 end of a metal column. This column is slotted to receive a rack 

 which carries the movable grip and which may be operated 

 either by hand or by motor, with stepped pulleys for different 

 speeds to meet the requirements of experimental work. 



It is necessary, of course, to provide means for cushioning or 

 preventing the recoil of the springs when a specimen breaks. 



