From the Rolls-Royce experimental archive: a quarter of a million communications from Rolls-Royce, 1906 to 1960's. Documents from the Sir Henry Royce Memorial Foundation (SHRMF).
Page from a journal discussing dynamic fatigue tests on rubber samples, including equations and diagrams.
| Identifier | ExFiles\Box 178\2\ img015 | |
| Date | 1st January 1920 | |
| 20 INDUSTRIAL AND ENGINEERING CHEMISTRY VOL. 12, NO. 1 ΔL. The actual fatigue machines used in the authors’ experimental testing are designed so that ΔL and Lmin. can be varied independently. In the following discussion “per cent oscillation stroke” and “per cent minimum strain” are mentioned. In order to comprehend the significance of the fatigue data it is necessary that these variables be clearly understood, for in the final analysis they do determine the fatigue life of a sample. % minimum strain = (Lmin. - L₀) / L₀ × 100 % oscillation stroke = (Lmax. - Lmin.) / L₀ × 100 = ΔL / L₀ × 100 These definitions involve three specific lengths of the sample (L₀, Lmax., and Lmin.) and are merely certain deformations expressed as percentages of the free unstrained length of the rubber. Four different types of oscillation conditions for a constant oscillation stroke, ΔL, are illustrated in Figure 3. (a shows a compression vibration cycle in which the maximum length in the imposed vibration is less than the free unstrained length of the sample. b shows a compression-extension vibration cycle in which Lmin. is less than L₀ and Lmax. is greater than L₀. c shows an oscillation condition in which Lmin. is equal to L₀ and Lmax. is greater than L₀. d shows an extension oscillation cycle in which both Lmin. and Lmax. are greater than L₀. FIGURE 3. TYPES OF OSCILLATION CONDITIONS FOR CONSTANT OSCILLATION STROKE ΔL The two fundamental problems that have been investigated quantitatively are: (1) If ΔL is kept constant, how does the fatigue life of a rubber unit depend on Lmin., for any value of Lmin. ranging from high compression to any possible extension? (2) If Lmin. is kept constant, how does the dynamic fatigue life of a rubber unit depend on the magnitude of ΔL? Linear Dynamic Fatigue Test Samples Figure 4 is a photograph of different types of rubber samples used by the authors in dynamic fatigue studies of rubber. They are cylindrical dumbbells of rubber bonded to metal bolt ends, shown at A and B. The effective rubber lengths varied from 2 inches (sample 1) to 0.125 inch (sample 13), and the ratios of diameter to length varied from 1/16 to 8. Machines for Testing Dynamic Fatigue Life of Rubber Figure 5 (upper) shows rubber samples, R, mounted in a slow-speed fatigue machine designed by the United States Rubber Company. The bottom bars, B, are held fixed and the top bars, T, are vibrated up and down. The rubber samples fastened between the two bars are vibrated through constant oscillation strokes which can have any value up to 3.5 inches. The minimum length of the sample in the vibration cycle can be varied either by adjustment of the separation between the bars or by variation of the point at which the bolt ends are attached to the bars. Both methods are clearly visible in the figure. The vibration frequency is 180 cycles per minute. Figure 5 (lower) shows rubber samples, R, mounted in a high-speed dynamic fatigue machine also designed by the United States Rubber Company. The two outer circular heads, O, are held fixed and the two inner circular heads, I, so-called “wobble plates”, are mounted on a shaft arranged to give a nutational motion to the heads, which in turn vibrate all the rubber samples back and forth along their own lengths. The minimum length of the sample in its vibration cycle can be varied by alteration of the separation of the plates, by the screw adjustment shown, or by variation of the point at which the bolt ends are attached to the plates. The machine is adjusted for a 0.5-inch stroke and vibrates at 3600 cycles per minute. FIGURE 4. TYPES OF RUBBER SAMPLES USED IN LINEAR DYNAMIC FATIGUE TESTS Figure 6 shows a DeMattia fatigue machine with rubber samples, R, being vibrated between a stationary head, S, and an eccentric driven head, C; this fatigue machine has a frequency of vibration of 660 cycles per minute; its stroke can have any value up to 2 inches. The fatigue machine on the right is another high-speed dynamic fatigue machine similar to the high-speed machine shown in Figure 5 (lower) but with a 0.25-inch stroke. General Nature of Linear Dynamic Fatigue Life Curve The general nature of the dynamic fatigue life curve for small oscillation strokes, say of the order of 25 per cent, is illustrated in Figure 7. The per cent linear strain at the minimum length in the oscillation is plotted as abscissa. The number of cycles of vibration necessary to completely rupture the rubber—that is, its dynamic fatigue life—is plotted in arbitrary units as ordinate. The important feature of the dynamic fatigue life curve is that the rubber under linear vibrations exhibits a minimum dynamic fatigue life in the region where Lmin. = L₀—that is, where the return stroke brings the sample back to a condition of zero strain. This minimum life is bounded both in compression and extension by regions of greater fatigue life. The general nature of the curve, a minimum bounded on two sides by maxima, remains the same whether the dynamic fatigue life is plotted as the number of cycles to break the rubber completely into two parts or is plotted as the number of cycles to produce a visible crack in the rubber. Mention has been made in published articles (1-5) that the mechanical fatigue resistance of rubber is less in extension when the minimum of the oscillation falls near zero strain. However, no complete study, as given in the present paper, has been published in which the fatigue lives of rubber have been investigated throughout the compression-extension region as a function of the strain or stress limits. | ||