UBC Theses and Dissertations
Plane strain compression testing of aluminum alloy 6061 at elevated temperatures Rehlaender, Saul Martinez
The purpose of this work was to study the effect of different geometrical criteria for samples, in particular the effect of initial thickness on the measured stress - strain curves from hot plane strain compression tests. The recent use of computer models to simulate hot working processes requires experimental values for their appropriate implementation and verification. Therefore, there exists the need to have adequate testing methods to obtain stress - strain data at high temperatures. Plane strain compression tests at elevated temperatures using aluminum alloy 6061 as a testing material were performed using the Gleeble 1500 Thermomechanical simulator. The temperature range of 400 - 500°C, strain rate; of 0.05 - 1 I sec. and true strain of 1 were used as testing conditions. In the present work, three plate shaped sample geometries with different thicknesses were used; a set of compressing platens was designed and fabricated in order to test these samples. The plate - shaped samples were made to test the recommendations given by the ASM [3 1 for obtaining plane strain. An additional cube - shaped sample geometry with its own set of compressing platens was also used; this tool and sample geometry was recommended by Duffers Scientific Inc. (now Dynamic Scientific Inc.), manufacturers of the Gleeble 1500. Stress - strain curves were obtained and it was observed that different stress - strain curves were produced, under the same testing conditions, for each of the tool I sample geometries used. Using optical metallography, the microstructure of the deformed samples from all the different sample geometries was observed. It was intended to observe the deformed grains and their localization in the deformation zone of the samples. It was found that, for a determined tool width, different initial tool width / sample thickness ratios (w / h) produce zones of deformed grains located in different regions of the deformation zone of the samples. Predictions of the strain distribution in the deformation zone of all the different sample geometries were performed by another worker. The predictions were made using the finite element software package DEFORM, utilizing parameters of the constitutive equation of the form ê =À [sinh(oa)]” exp(—Q/RT) [ 68 1 obtained for each of the geometries. A correspondence of the computer predicted strain distribution and the observed deformed grains and their localization in the deformation zone of the different sample geometries was found. From this correspondence, it is observed that, samples with initially large w I h ratios have a more homogeneous strain distribution than samples with initially low w I h ratios. This difference in strain distribution is related to the different stress - strain curves obtained from the different tool I sample geometries and the different initial w I h ratios used. For a specified tool width, initially thinner plate shaped samples (large initial w/ h ratios ) produced a more homogeneous strain distribution than thicker samples. Plate shaped samples with larger initial w I h ratios exhibited a more homogeneous deformation in the entire deformation zone; then these samples would represent better the bulk deformation behavior of the metal. This idea and observations are in agreement with what has been proposed and observed by different researchers.
Item Citations and Data