Optimization of the debinding process

The debinding of green bodies is measured in situ with the ThermoOptical measurement systems (TOM) as well as with the thermal analysis method at the Fraunhofer-Center HTL and subsequently optimized. The optimization is undertaken for all types of green bodies regardless of the component size and geometry, the inorganic components (metals or ceramics) and the type of organic binder and/or the binder content. Thus, for example, debinding processes for the in general somewhat difficult debinding of green bodies from an injection molding process can be developed.

The result of debinding is influenced by the furnace atmosphere. While debinding in air or other furnace atmospheres containing oxygen, the combustion of organic additives dominates, pyrolysis processes are relevant for debinding in inert or reducing atmospheres. The latter processes frequently occur with debinding in air in the interiors of green bodies, because oxygen only reaches this area when the binder burn out has been completed on the peripheral layers.

There are adequate in-situ measurement stations available at the HTL center for all relevant furnace atmospheres. Inert gases, oxidizing or reducing gases – as well as 100% hydrogen – can be used as furnace atmospheres. Even the furnace atmosphere in gas-fired furnaces can be exactly simulated. The calibration of the furnace atmospheres between the manufacturing furnace and the corresponding ThermoOptical measurement system is vital, because the debinding conditions optimized using the TOM system can be transferred to the manufacturing furnace.

By measuring the test body weight, the debinding rate can be measured with a very high level of reproducibility^[1]. The reproducibility is 0.1%. This enables the development of a meaningful database for the calculation of the debinding kinetics. For this purpose, several debindings are performed at the HTL using differing temperature-time cycles on the same green body samples. A robust numerical process calculates a kinetic model from the measured data, facilitating the prediction of the debinding rate for any random temperature-time cycle within the scope of the measured range^[2]. With this kinetic model, simple optimization of the debinding cycle is already possible. Thus, temperature-time cycles can be calculated where the debinding rate is almost constant. This leads to lower stresses on the components than temperature-time cycles with constant heating rates^[3]. The maximum still safe debinding rate is then determined by experimentation with the corresponding calculated temperature-time cycles. For this purpose, the debindings are undertaken on larger samples or smaller components in the TOM system, and during sintering, the damages to the samples are registered in-situ. Measurements of acoustic and gas emissions are used primarily at the HTL for registering the damage, as they can sensitively register slight damage.

Further in-situ measurements are necessary for more precise testing, e.g. for upscaling to other component geometries and for consideration of effects in industrial furnaces. Therefore, the endo-/exothermic effects with pyrolysis or binder burn out must be quantified, which is performed at the HTL using dynamic differential scanning calorimetry (DSC) in controlled atmospheres. The thermal conductivity of the green bodies is determined during debinding using a Laser-Flash method. Furthermore, the permeability of gases is measured through the pores of the green body. Together with the kinetic model, the measured data is used in a coupled finite element (FE) model, which has been developed at the HTL to optimize the debinding process. The model determines the temperature distribution in the green body during debinding for each time step, considering the reaction heat. From the local temperature and locally available oxygen, the kinetic model calculates the local debinding rate. The resulting gas phase reactions lead to concentration and pressure gradients, which are dissipated in the pores by diffusion and flow processes. These processes are also FE simulated. Finally, the mechanical stresses for the respective time step resulting from temperature differences and the gas over-pressures are calculated. The simulation is then repeated for the subsequent time step until debinding has been completed. The debinding conditions are adjusted with the FE model to minimize the mechanical stresses on the green body. This facilitates selective optimization of the debinding conditions for the individual components. Debinding cycles can be drastically reduced in comparison to empirically optimized cycles.