Correlation of Fatigue Data to Determine Stress Concentration Factors in 76S-T Aluminum Alloys

Citation

Shepherd, John Tyler (1950) Correlation of Fatigue Data to Determine Stress Concentration Factors in 76S-T Aluminum Alloys. Engineer's thesis, California Institute of Technology. doi:10.7907/XMFM-KC84. https://resolver.caltech.edu/CaltechETD:etd-03062009-160046

Abstract

The basic problem of this particular investigation was to determine the stress concentration factors present at a shoulder in the root section of a model propeller blade manufactured from 76S-T aluminum alloy. The particular geometric section considered is found in the type propeller blades such as will be installed soon in the Southern California Cooperative Mind Tunnel, Pasadena, California. These blades were manufactured from forgings of 76S-T by the Hamilton Standard Propeller Division of the United Aircraft Corporation. Stress concentration factors were determined for what will be called the "critical section" of the blade. This section is understood to mean the cross section of minimum area which is located immediately above the stress-raising fillet at the junction of the blade itself and the root flange. A secondary purpose of the investigation was to correlate these data with two other experiments that had already been carried out for this particular blade shape. These earlier experiments determined the magnitudes of surface stresses in the critical section, one utilizing a full-scale, three-dimensional model, and the other a two-dimensional full scale model one inch thick. Both of these experiments utilized static tension loads, the loads being applied over the upper surface of the flange, exactly as they are assumed to act when the blade is rotating in the wind tunnel. Also, an attempt was made to compare the data obtained in this investigation with other fatigue tests of a more general nature. The tests made in this present investigation were all of a fatigue nature, in which one-tenth scale models of the actual blade were used as fatigue specimens. Two types of loading were used. The first consisted of a cyclic loading between varying upper tension limits to zero stress, and the second consisted of a constant upper tension stress limit with varying minimum stresses. Tests were conducted using an upper tension nominal stress limit of approximately 28,000 psi and were extended to tension stresses of lower values which gave a fatigue life of more than 15,000,000 cycles. Due to the slow rate of loading (2,500,000 per day) in the Sonntag Universal Fatigue Testing Machine, the investigation was not continued to the generally accepted value of 500,000,000 cycles which marks the upper cyclic limit of fatigue investigations. It was found that stress concentration factors determined from this type of fatigue testing came surprisingly close to those determined from the full-scale, three-dimensional, static tests. The selection of the type loadings used was not entirely arbitrary. The stress cycle, which varied from zero to a peak tension value and then to zero again, would closely simulate the loadings in a start-stop cycle; while the superposition of a cyclic stress upon a steady tension load would result in approximately the type of loading experienced by the blade when it was rotating at a constant angular speed under aeroelastic axial forces.

Thesis Files