^{1}

^{*}

^{1}

Numerical simulations based on a conjugate heat transfer solver have been carried out to analyze various gas quenching configurations involving a helical gear streamed by an air flow at atmospheric pressure in a gas quenching chamber. In order to optimize the heat transfer coefficient distribution at key positions on the specimen, configurations involving layers of gears and flow ducts comprising single to multiple gears have been simulated and compared to standard batch configurations in gas quenching. Measurements have been performed covering the local heat transfer for single gears and batch of gears. The homogeneity of the heat transfer coefficient is improved when setting up a minimal distance between the gears (batch density) and when introducing flow ducts increasing the blocking grade around the gears. An offset between layers of the batch as well as flow channels around the gears plays a significant role in increasing the intensity and the homogeneity of the heat transfer in gas quenching process.

In the automotive industry, in order to achieve a sufficient hardness distribution (corresponding to a martensitic microstructure) or shape quality [^{2}∙K) [

In comparison to quenching techniques using liquids, the impact of the gas flow on the heat transfer between the quenching gas and the element to be quenched is of major importance [

Heat transfer in industrial gas quenching is described by the heat transfer coefficient h, independent of the quenching medium temperature and the specimen temperature and material properties. In industrial gas quenching, the heat transfer coefficient has been derived as [

where c_{2} is a factor related to the quenching chamber geometry.

The heat transfer in gas quenching is analyzed using a local and an integral method to determine the heat transfer coefficient on helical gear specimen. The local determination of the heat transfer coefficient is done through measurement of the local heat flux on the specimen surface using a 5 × 5 mm Captec sensor with integrated thermocouple to monitor the specimen surface temperature [

where l is taken as the streamwise length of the gear [

Quenching conditions involving small heat transfer coefficients for material of high heat conductivity (copper, aluminum) lead the effect of internal conduction to overcome the effect on the external heat transfer by convection so that the temperature distribution is considered uniform at any time in the solid. Thus the integration of the heat equilibrium equation,

can be reduced to a time-depending integration.

In that case, the temperature can be monitored at any point within the specimen and the integral or mean heat transfer coefficient may be calculated. Three specimen helical gears of thicknesses 20, 25 and 30 mm made of EN-AW6082 to minimize the Biot-number have been investigated during the quenching process.

The gas velocity in gas quenching process is measured using a pressure-based 7-hole, pitot probe whose orientation and calibration allow the measurement of the 3 velocity components in a 72˚ cone in relation to the incoming flow [

The simulation of the heat transfer process in gas quenching requires accurate modelling of the quenching environment (flow velocity, turbulence and chamber geometry), at the macro and meso scales. The accurate modelling of the micro scale, that is, the helical gear itself, ensures a valid computation of the heat transfer coefficient on its surface. The simulation library Open FOAM 2.0.1 provides an existing steady-state, coupled heat transfer simulation solver. The computation of the gas flow is based on Reynolds-averaged Navier-Stokes (RANS) equations using the SIMPLE-algorithm (Semi-IMplicit Pressure-Linked Equations) and the energy equation for the fluid. It provides the temperature field around the solid specimen that is used to solve the heat conduction pro- cess for the selected regions to be quenched.

Heat transfer problems simulations at large scale are ( [

While the gas quenching chamber geometry is directly meshed within Open FOAM, flow ducts and the helical gears are implemented as. stl geometries in the mesh with finer resolution, so that the average grid cell number remains below 30 million cells. Inlet boundary conditions are fitted to the complex flow from the heat exchanger as velocity and turbulence components that are extrapolated from measurements in order to accurately model real gas quenching conditions.

Top-to-bottom flow quenching chambers are found in industry for quenching process of gears [

Flow distribution characteristics in the gas quenching chamber are evaluated using the pressure probe. The probe tip is pointed towards the main flow direction (along the z-axis). The probe is mounted on a positioning system scanning the investigated planes (1 - 3) shown in

The comparison of the velocity distributions is presented in

Numerical simulations have been performed based on the gas velocity distribution on Plane 1 as boundary conditions. The main velocity component (along the z-axis), the coefficient of variation, defined as VarCo = Sig/X and the turbulent kinetic energy have been detected and compared to simulation results for the empty flow chamber. A simulation involving a two-layer batch of helical gears has been related to the simulation/measure- ment in the empty chamber. The results are shown in

locity measured in the center area of the chamber, simulation and measurement in the empty chamber show an increase in the mean value as the gas flows to the bottom, while the velocity decreases when the chamber con-

tains a batch. Supposing the mass flow rate to be constant through the chamber, this decreasing value is due to the diversion of the flow on top of the batch, thus transmitting the flow momentum in the transverse directions. The increasing value in the center as no flow resistance is encountered is also described in [

The increase in streamwise velocity for both empty cases in an empty chamber is due to the increase in flow uniformity shown in

Simulations have been performed for a single helical gear in various flow configurations involving flow ducts around the specimen. The influence of the flow duct dimension on the local heat transfer at 4 key positions of the helical gear is investigated. The top surface (disc front) and the bottom surface of the gear (disc gear) as well as the streamed surface (tooth front) and the reverse side of the tooth (tooth rear) are investigated as the intensity and the uniformity of the heat transfer coefficient distribution over those surfaces play a major role on achieved quality of the workpiece in the heat treatment process.

The velocity distribution around the flow duct and the gear is displayed around the gear. Such a recirculation area (

The heat transfer on the specimen surface is evaluated in

Considering a single layer of helical gears in gas quenching, the packing density of the batch has been evaluated, aiming at obtaining a homogeneous local heat transfer coefficient distribution at all gear wheels. At a high batch density, the distance between a gear and its neighboring specimen is low, as represented in

Therefore, the influence of the distance between gears is investigated for the two sides of the critical gear tooth situated closest to a neighboring gear (

cients are plotted in

The flow is producing intense recirculations at higher distances, thus creating intensive heat exchange at the top, low heat exchange into the eddy in the middle of the tooth and eventually reattaching at the bottom, thus re-intensifying the heat transfer. The influence of the gear distance in the batch is displayed in

Three configurations of batches involving staggered helical gears have been investigated using flow and heat transfer simulation as seen in

An offset between parts is used (transverse in one direction) on the second layer for the third configuration of the batch in

The influence of confinement on the specimen heat transfer in gas quenching involving 2 to 3 layers in a batch with second-layer offset is summarized in

Measurements and simulations for evaluation of quenching heat transfer in batch configurations in industrial gas quenching of helical gears have been reported. A combination of adapted flow confinements with a layer disposition involving an offset of the second layer is improving quenching homogeneity and intensity of the quenching for the whole batch. The introduction of confinements improves the local homogeneity of the heat transfer coefficient in the tooth region of the helical gear specimen.

The authors gratefully acknowledge support from the Forschungsgemeinschaft Industrieofenbau (FOGI e.V.) and AichelinGes.mbH, Mödling (Austria) for offering the possibility to perform in-situ test in industrial furnaces.

The project is funded by the AiF (Arbeitsgemeinschaft Industrieller for Schungsvereinigungen “Otto von Guericke” e.V.) through financial resources from the BMWi (Bundesministerium für Wirtschaft und Technologie).