Metal Flow Optimization Within The Deformation Zone In Continuous Rotary Extrusion (cre) Of Az91 Magnesium Alloy

Numerical modelling of continuous rotary extrusion (CRE) of magnesium alloy AZ91was carried out using commercially available DEFORM® 3D software, to propose ideal deformation chamber tool geometry with lower metal velocity (strain rate) and temperature gradients within the deformation chamber resulting in optimal extrudate velocity. The analysis began by developing the numerical model of the CRE process and calibrating it with the help of a well-documented in literature aluminum alloy AA6063 in DEFORM® 3D. Calibration was verified with data from the published work and helped in modelling aspects such as the friction conditions, mesh elements, heat transfer conditions at workpiece and tool interface and time step of magnesium alloy simulation. To understand how the magnesium AZ91 alloy response to the process conditions in the CRE process, a number of compression tests were carried out at higher strain rates and temperature which are closer to the actual encounter in the physical CRE process and which are not available in literature. Measured flow stress data of magnesium AZ91 alloy was then corrected for friction and deformation heating through inverse analysis. Zener-Hollomon a temperature compensated strain rate model Z was used to calculate the material flow stress for magnesium AZ91 alloy. It was found that the Zener Hollomon model was matching well with the measured flow stress data with a regression coefficient R2 of 0.96. This material model was then incorporated into the calibrated numerical model to analyse the magnesium alloy AZ91 behaviour at different wheel speeds of 3 rpm and 7.5 rpm when a 9 mm rod is extruded in CRE process. CRE of AZ91 was then verified by grid pattern analysis and comparing the metal flow and die temperature with the physical experimental results. The numerical model was able to replicate the actual physical process and predict the state variables of the extrudate. The predicted state variables gave insight into the sources which influence extrudate quality. Hence to reduce the state variable gradient and have a uniform metal flow within the deformation chamber an additional feeder plate was introduced within the 2 deformation chamber between the original feeder plate and the extrusion die. H/D ratio (Hheight, D- diameter) of the additional feeder plate was varied from 0.7 to 1.2 to study which ratio provided lower state variable gradient and therefore optimal extrudate velocity. By comparing various simulation results obtained in this study it was found that additional feeder with H/D ratio of 1 tends to have optimal extrudate velocity and reduced state variable gradients. These results are proposed to our research collaborators the Institute for Non-Ferrous Metal in Skawina, in Poland to help them narrow down the number of physical experiments performed to optimize the extrudate quality thereby saving time and cost.