Hot extrusion of Ti-6Al-4Valloy has been studied using 4nite element simulation and the results are compared with those obtained experimentally. First, the constitutive behavior of the material and friction at the extrusion temperatures are established based on the results obtained through cylindrical and ring compression tests, respectively. While the 8ow stress below βtransus temperature is expressed as a strain-dependent function, it is taken as strain-independent one at higher temperatures. The distribution of strain, temperature and effective stress has been simulated under different design and processing conditions. Simulation results show that heat generation due to deformation is signi4cant (as much as 160℃) in the hot extrusion of Ti alloys, and it mainly occurs at the beginning of the extrusion process. This leads to reduction in flow stress which, in turn, leads to enlarged deformation zone. A fair agreement has been found between the experimental results and those obtained through simulations.

In the present work, an attempt was made to predict the temperature evolution during the extrusion of 7075 aluminium alloy by means of 3D FEM computer simulation. Results show that ram speed has a significant influence on the temperature distribution in the billet, which continuously changes throughout the process, as a result of complex heat generation and heat loss. The thermal effect results in characteristic variation of extrusion pressure. At a higher ram speed, the decrease of extrusion pressure is faster during the steady-state extrusion, due to more heat generation, less heat loss and thus more steeply decreased flow stress, as the process proceeds. The temperature inhomogeneity on the cross-section of the workpiece entering the die bearing is more pronounced when ram speed is higher. While going through the die, the extrudate undergoes a process of temperature redistribution and the temperature becomes more homogeneously distributed at the die exit. The present simulation does not render support to the general statement that the corner of the extrudate is hotter than the flat surfaces. Incipient melting is predicted to occur after a half of the billet is extruded at a ram speed as low as 1 mm/s corresponding to an extrusion speed of 0.48 m/min, if the billet contains the phases with low melting points. However, if the billet of the same alloy is in an improved metallurgical condition, no melting-related defects would be expected to occur to the extrudate running at a speed eight times faster. The results also confirm the linear relationship between the increase of the maximum temperature and logarithmic ram speed during the steady-state extrusion.

The conventional aluminium extrusion process is run at constant ram speed, leading to quality inconsistency along the length of the extruded product and even to hot shortness as a result of continued temperature evolution. In the present work, computer simulation of the process at varying ram speed was performed in order to determine the conditions to prevent the extrudate temperature from rising excessively. To maintain the maximum workpiece temperature around 500 and 480℃ corresponding to two initial microstructural states of 7075 aluminium billets, two ram speed profiles were derived from the simulation results of a series of conventional extrusion runs. The predetermined ram speed profile commenced at a relatively high value at the beginning of an extrusion cycle and decreased exponentially with ram displacement as soon as the maximum workpiece temperature reached the target value. The simulations showed that with these ram speed profiles the continued temperature increase normally occurring during conventional extrusion could be effectively inhibited. This was verified experimentally by measuring the extrudate temperature continuously using a thermocouple in the die close to the bearing and also using a multi-wavelength pyrometer behind the die. With the predetermined ram speed profiles, the fluctuations of the maximum workpiece temperature could be controlled within a range of 10℃. The time taken to extrude each billet could be significantly shortened. In addition, the die face pressure remained stable, which would also favour the consistency of the quality of the extruded product.

Samples of Ti-3Al-5V-5Mo alloy were compressed in both b and aqb phase region on a Gleeble 1500 Simulator. Compression tests were carried out in the temperature range of 700-1000℃ and strain rate range of 0.05–15 sy1. Experimental results show that the flow stress of Ti-3Al-5V-5Mo alloy decreases with the increase of temperature and the decrease of strain rate. At high strain rate, typically 5 and 15 sy1, discontinuous yielding followed by flow oscillations was observed in both b phase region and aqb phase region; at low strain rate, the flows display single peak stress. The flow stress at a strain of 0.2 was analyzed with a stand constitutive equation. Activation energy parameters were obtained, and they are 133.46kJymol for the b phase region and 661.90kJ/ymol for the α-β phase region. Microstructures of the compressed specimens in water-quenched conditions were critically observed. High temperature deformation mechanisms have been elucidated. In the b phase region, the operative deformation mechanisms are dynamic recovery at high strain rates and grain boundary sliding at low strain rates. In aqb phase region, the α phase undergoes dynamically recrystallization at both high and low stain rates.

Received 19 July 2002; accepted 11 December 2002 Abuminium extrusion involves the generation of free surface, thermal effects, large deformations and complex geometries. The establisbed finite alement method (FEM)-based 3D simulation tools using the updated Lagrangian approach, or the Eulerian approach or the arbitrary Lagrangian Eulerian approach all have limitations in describing the process that develops from the transient state to the steady state before reaching the end when the steady state is disturbedi As a result, the simulation of aluminium extrusion performed so fro has been restricted to simple geometries, small length to diameter (L/D) ratios, the beginning stage or steady state conditions. Tiris paper reports on an unprecedented attempt to simulate an entire cycle of alumimium extrusion from a billet with an L/D ratio of 4 to a solid cross-shaped profile, using the DEFORM 3D software based on the updated Lagrangian approach. Simulation successfolly predicts a complete extrusion pressure/ram displacement diagran that begins with a pressure breakthrough and ends with another pressure rise due to the inhihition of metal flow by the rigid dinmmy block. The developments of velocity, effective strain and temperature inside the deforming billet indicate that the processisnon steady, evenin the steady state, as a result of continuous heat generation and sticking condition at the billet-container interface. The non steady characteristics are reflected in the expanding defomation zone and shiinldng dead metal zone. Simulation also reveals the patterns of the maximum temperature variations in the workpiece and in the tooling, due to heat generation and exchange. Even at a relatively low ran speed of 2mm/s, the maximumn temperature of the workpiece, after an initial steep rise, increases gradually till the end of the process, which may well lead to the occurence of hot sbortness. On the basis of these results, a change of the conventional mode of aluminium extrusion is recommended, which at present operates almost all at a constant ram speed and often begins with a uniform billet temperature across the aluminium extrusion industry in the world.