Scripta METALLURGICA Vol. 19, pp. 425-430, 1985 Pergamon Press Ltd. Printed in the U.S.A. All rights reserved LOW TEMPERATURE FATIGUE IN Cu-Zn-Al SINGLE CRYSTALS Marcos Sade and Manfred Ahlers Centro At~mico Bariloche Comisi~n Nacional de Energ~a At6mica 8400 - Bariloche, Argentina (Received October 25, 1984) [Revised February 8, 1985) I. Introduction In an earlier publication Ill, fatigue studies of Cu-Zn-AI single crystals had been reported. These crystals undergo a martensitic transformation from the bcc ~ phase to an 18R structure when a stress is applied. Till now most of the experiments had been performed at or around room tem- perature. A few additional data obtained by cycling at lower temperatures near liquid air indi- cated a substantial increase in fatigue life compared to that at room temperature !i]. More experiments have now been made, in order to compare the fatigue behaviour at low temper- atures (in the following abbreviated "LTF") with that around room temperature (abbreviated "HTF"). In addition to the measurement of the relevant stresses and stress-strain relations, the samples were studied mainly by optical and scanning electron microscopy "SEM". By comparing HTF and LTF more information on the mechanisms for nucleation and propagation of fatigue cracking is obtained. II. Experimental methods The experimental details can be found in Ill, therefore a few remarks are sufficient. Single crystals of Cu-Zn-AI alloys were grown by the Bridgman method in sealed vycor tubes. The alloy compositions are shown in table I. They were chosen with an M s temperature which permitted to induce the martensite at the working temperature Tex p at a stress o sufficiently below the stress necessary to plastically deform the matrix. Most samples were deformed in liquid air at -1860C, but for comparison two samples with the same heat treatment and an M s at -60°C were fatigued at room temperature. All alloys had an electron-concentration of 1.48. Cylindrical samples were spark-machined from the single crystals. They had a gauge length of 6.5 ram, a central diameter of 2.8 nnn and thicker heads (5mm diameter), in order to insert them into the grips of an Universal testing machine, Instron model 1123. All the samples were annealed for 30 minutes at 8000C and then air cooled. After that samples were mechanically (600 grade paper) and electrochemically polished (in a solution of 15% nitric acid in methanol for 4 or 5 minutes at 4 volts). The cycling was performed between a strain corresponding to the B phase and that for a 50% transformation to the 18R martensite. The tests were made in tension with a cycling frequency of 50 cycles per minute on the average (see table i). Samples denoted by the same prior number in table i were cut from one single crystal batch. All tensile orientations are shown within the unit triangle in figure i, with the same symbols as used in table i. III. Experimental results a) The stress dependence of fatigue life: The stress of cycling and the corresponding number of cycles to fracture are plotted in figure 2 for the s~mples fatigued at -186°C , and for s~mples i0-I and 10-2 fatigued at room tempera- ture. The latter were included, since they have the same stress and a similar tensile orienta- tion to samples 1-1,1-2 and i-3. The crosses mark the HTF results of the earlier work [i]. A symbol with an arrow indicates that the cycling was stopped before fracture had occurred. The 425 0036-9748/85 $3.00 + .00 Copyright (c) 1985 Pergamon Press Ltd.