Investigation of friction and wear behaviors of Al/AlB2 composites materials

In this study, in-situ AlB2/Al bulk composite material was subjected to squeeze casting process, as a secondary operation, to produce composite samples with 5, 10, 20 and 30 % reinforcement ratios. ETIAL8 aluminium was used as the matrix. Scanning electron microscopy analysis showed that the AlB2 reinforcement phase was uniformly distributed within the matrix of the composites. Having performed hardness and density measurements, the samples were subjected to wear tests.Dry sliding wear tests were performed on a pin-on-disk rig in accordance with ASTM G99-95 standard. The amount of humidity and temperature were controlled during the tests. Sliding speeds of 1.83, 3.40, 4.71 m/s, normal loads of 10, 20, 40 N and sliding distance of 500, 1000, 2000 m were the test parameters. It has been observed that the wear rates of the samples were increasing with sliding distance and load. With increasing sliding speed, oxidation took place on the worn surfaces of the samples which resulted in a reduction in the amount of wear rate and coefficient of friction. It has been observed that the increase in the amount of the AlB2 reinforcement phase increased the wear resistance of composite samples. In order to identify the wear mechanisms of the worn surfaces and subsurface deformation, scanning electron microscopy (SEM), optical microscopy and x-ray diffraction (XRD) analysis were carried out. The analysis show that abrasion, adhesion, oxidation, delamination mechanisms were operating on the worn surfaces.At lower sliding speeds and loads, abrasion mechanism was dominant whereas at higher sliding speeds and loads, adhesion and oxidation mechanisms. In addition, at higher loads a mechanically mixed layer was formed on the worn surfaces. Subsurface plastic deformation and cracks were found on the samples tested at high speeds and loads.Finally the wear rate was predicted by using a commercial computer programme, ABAQUS operating based on finite element method. In the computer model, Archard?s and Sarkar?s linear wear equations were adopted. A computer programme was prepared in FORTRAN that was uploaded as a subroutine into the UMESMOTION module of the ABAQUS package. It has been determined that the computer model predicted the wear rate within 70 to 97 % accuracy when compared to the experimental results.