Determination of Heat Transfer Coefficients in High Speed Quenching Processes

Abstract

Rapid water quenching of steel specimen in heat treatment processes can evoke compressive residual stresses near the specimen surface. Thereby a significant increase in the fatigue limit of the components may be achieved. For applying this process, knowledge and control of heat transfer coefficients (HTC) in various process conditions during quenching is necessary. For HTC lower than 10 kW/m2K, temperature measurements with thermocouples below the surface of steel components may be performed. However, for measurements of HTC above 10 kW/m2K, the local cooling rates are huge near to and at the surface and the resulting temperature gradients are high. For these rapid temperature changes, the response times of sheathed thermocouples are not sufficient and because of the high temperature gradients also the positioning of the thermocouples within the specimen must be very accurate. The most appropriate possibility to determine HTC above 10 kW/m2K is temperature measurement directly at the surface using thermocouples with a small response time. Another possibility to determine HTC in high speed quenching processes is the use of probe material that provide higher thermal conductivity than steel, such as copper with about 27 times higher heat conductivity compared to austenitic steel. Higher heat conductivity of the specimen to be quenched lowers the temperature gradients within the specimen and the cooling rates. Therefore, measurements with commonly used sheathed thermocouples are feasible. Both techniques for HTC-evaluation in rapid quenching process have been tested. The resulting HTC values will be compared with HTC data from numerical heat and fluid flow calculations and from empirical equations. Results of these evaluations show that heat transfer coefficients of about 50 kW/m2K can be achieved by high speed quenching.

Numerical and experimental investigation of the mesoscale fracture behaviour of quenched steels

“…It is worth noting that the heat transfer coefficient for the current test conditions (Table ) has been calculated outside this work. Frerichs et al solved the inverse heat conduction problem based on cooling curve measurements on copper specimens, having a very good agreement with additional computational fluid dynamics simulations and data from the VDI‐Wärmeatlas…”

“…, a simple scheme of the geometry, symmetry conditions and heat flux applied on the surface is presented. The material (100Cr6) properties adopted have been described previously in the Material section, and the heat transfer coefficient h = 23 kWm − 2 K − 1 employed responds to the aforementioned calculations …”