Measurement of temperature and determination of heat transfer coefficient in backward can extrusion

Ph.D.: Poul Henningsen
To develop reliable models of metal forming processes it is important to know the temperature distributions in the boundary between workpiece and tool. As earlier work on determination of the heat transfer coefficient have mostly focused on simple upsetting, this work is measuring the heat transfer between workpiece and tool in a backward can extrusion.
The process is characterised with large deformations and very high surface pressure, more than 2000 N/mm2 . The large deformations result in a rise of temperature of more than 200 degrees Centigrade. Because of large surface expansion, the process is very dependent on effective lubrication. In the experiments a can in low carbon steel is formed, with a lubrication layer of phosphate soap. The workpiece is formed with a reduction of 56%
The temperature is measured by thermocouples in the die insert. The die insert is divided into two halves, where 8 grooves have been milled, in the lower part. Thermocouples are welded to the end of the grooves. The end of the grooves are located at 4 different distances from the tool surface (0.4, 0.8, 1.2, 2,0 mm). After the welding of the thermocouples, the die insert is put together and placed in the container.
The punch is equipped with three thermocouples mounted at a distance of 0.5 mm from the surface. The thermocouples are welded to the end of grooves milled in a small plug. the plug is then pressed into a hold in the punch nose.
Finally the temperature of the workpiece at the free surface is measured, by welding a thermocouple directly onto the surface. It is then possible to follow the heating and cooling of the workpiece during the process.
All of the temperature measurements in the tool and workpiece are compared with a number of FEM simulations computed with different heat transfer coefficients. From the simulations the current heat transfer coefficient is determined.