No abstract

No abstract

Similar vulcanized rubber bands were subjected to four different tensions which produced immediate strains of approximately two to five times their original length. During the life of the bands this elongation was augmented an additional one to two hundred percent by the so-called after-strain or drift. Observations were taken at constant temperature. The average life of the bands was about 33 weeks. No correlation was found between the magnitude of the load and the time that the specimen could bear its load. Drift for all stresses continued to the breaking point. In all cases the initial time rate of drift was comparatively high. It declined with time to a positive minimum, passed through an inflection, and then steadily increased for the remaining life of the band. The inflection occurred after some weeks or months of drift. No correlation was found between the magnitude of the load and the time of inflection in the drift rate. During the interval between one and eighty minutes, the magnitude of the drift in at least one specimen was directly proportional to the log of the time. It then deviated from the log law and for three weeks closely followed a decreasing power law.

When a suitable weight is supported by suspension from a piece of rubber, as a stationer's band, the rubber may be stretched any amount up to several times its original length, but its new length is not constant; it increases with time. The increase in length with time is variously known as “after-effect, #x201D; “creep,” “drift,” “flow,” or “time-yield.” This phenomenon, which seems to have been very incompletely investigated, was probably first recognized by Dietzel in 1857. Kohlrausch in 1875 made an entensive study of both torsional and linear after-effect in metal, glass, and rubber. The load used by Kohlrausch on rubber were, however, exceedingly small, and the duration of drift was limited to one day. He came to the conclusion that drift in rubber followed a power law for the first sixty minutes. During the next decade Pulfrich slightly extended the work of Kohlrausch by experimentation on a red rubber tube, using elongations up to 150 per cent, with maximum observation time of 15 days. He concluded that the power law of Kohlrausch held for at least thirty minutes of drift. In 1903, Bouaase and Carrière observed the after-effect (in pure gum and sulfur cords of 4 mm. diameter and of specific gravity 0.984) under a great variety of experiments, and concluded that drift was to be expressed by an exponential or logarithmic law rather than by a power function. Both Phillips and Schwartz arrived at similar conclusions. More recently Ariano reported that the drift proceeded at a decreasing rate which finally assumed a constant value either finite or zero. Van Geel and Eymers found that for milled rubber the drift continued until the specimens broke, but that for rubber obtained by evaporation of latex all after-effect ceased within three minutes. Shacklock noted that creep took place for some hours and then reached a limit. Evidently more light needs to be cast on the probem of the drift effect in rubber. Two preliminary experiments on drift are herein considered; Part I on general trends, and Part II on more specific analysis of the effect as observed in one specimen.

Specimens of each of two vulcanized rubber compounds of known composition were subjected to various constant gravity loads in a controlled temperature chamber for several months. The lengths of the stretched bands were observed from time to time while the temperature was altered in gradual steps within a range of about thirty degrees centigrade centering around room temperature. A projection method for observing small changes in length was devised for the experiment. The contraction of stretched rubber with rise in temperature, first observed about the middle of the eighteenth century, was verified, as also was its positive elongation when unstretched or when but slightly stretched. The present experiment shows that the expansion or contraction for each specimen was constant when measured in cm per degree temperature change, but not when measured in percent of the prevailing length of the band. That is, the thermal elongation was most simply expressed when considered independent of the after effect or drift. For molded bands originally 1 mm in thickness, 4 mm in width, and 10 cm in mean circumference, and with loads ranging from 30 to 836 grams producing elongations up to about 700 percent, the rates of expansion were found to lie between plus 0.0037 and minus 0.0550 cm per degree. The relation between the rate of thermal elongation, dl/dθ, and the stretching load, w, was found to be linear within the limits of the experiment, and the constants c and d in the empirical equation, dl/dθ=cw+dhave been evaluated, the former being negative in sign. The critical tensions for which there appeared no thermal change in length were 63 and 97 grams, respectively, for the particular shapes and compositions used. These tensions are equivalent to critical stresses of 7.72 and 11.9×105 dynes/cm2 (based on the cross section of the unstretched band) or of 11.2 and 17.2 lb./in.2, respectively, for the two compounds. Extrapolation by placing w =0 does not seem justifiable because when the resulting rate is divided by one-half the circumference of the bands the magnitude of the coefficients of linear expansion turn out to be several times the generally accepted values based on volumetric determinations.