Metal Binder Jetting (MBJ) is a non-fusion Additive Manufacturing (AM) technique emerging rapidly due to several advantages such as wider selection of materials, higher production rates and lack of issues such as residual stresses, oxidation and phase changes that exist in other AM processes. Furthermore, MBJ offers the possibility of isotropic material properties due to uniform thermal profile during sintering thereby enabling the microstructures and material state closer to that of conventional processes. Despite these significant benefits, one of the main bottlenecks hindering the potential of this process is the large amount of distortion and shrinkage associated with sintering process. The inability to estimate the distortion and shrinkage of the parts accurately severely restricts the application of this process to larger sized parts than those used in Metal Injection Moulding (MIM) [1].

A reliable numerical simulation tool can predict the distortion and shrinkage of the part on a continuum scale during sintering thereby assisting process control. In addition, the geometries can be pre-compensated, where possible, based on the knowledge of possible sintering distortion and shrinkage through predictions which can aid the design process for MBJ. However, the accuracy of simulations and its applicability as a process control and virtual designing tool hinges greatly on the quality of material data. In the current paper, a phenomenological model for sintering simulation based on Olevsky’s work is presented for industrial sintering applications using hybrid experimental and simulation framework [2].

Extensive material characterisation is undertaken to determine the accurate shrinkage and viscosity of the material for given process parameters to increase the accuracy of predictions. The material under investigation is 17-4PH stainless steel and detailed characterisation includes sinter dilatometry and sinter bend tests in order to evaluate the shrinkage and viscosity of the material as a function of porosity and temperature [3]. The material characterisation is undertaken by IWM, RWTH Aachen. Test specimens for characterisation are printed by Fraunhofer IAPT on their Digital Metal MBJ printer. Furthermore, to investigate the suitability of the presented methodology and evaluate the accuracy of sintering simulations of industrial components, a suitable component with sufficient complexity is designed by Scholz Mechanik as a demonstrator. This demonstrator is printed and sintered by Fraunhofer IAPT using 17-4PH powder on their Digital Metal MBJ printer and Desktop Metal Studio system sintering oven respectively.

The simulation model is analysed in the commercial software Simufact Additive 2021.1 MBJ sintering module based on finite element (FE) solver Marc 2020. The sinter temperature profile is applied on the model as nodal temperature and the mechanical behaviour of the part is then analysed under the influence of temperature, gravity and friction between the part and the baseplate as external boundary conditions. The initial relative density of the part is assumed as 56%. The predicted shrinkage and distortion in the industrial demonstrator part during sintering is shown in Figure 1. The simulation results indicate an average sintered relative density of 96.13% in the part. A comparison of the sintered part dimensions from predictions and measurements indicates a very good agreement, demonstrating the suitability of the suggested approach for accurate prediction of sintering behaviour in MBJ parts.

References:

[1] A. Mostafaei et al., Progress in Materials Science, 2021, 119 100707.

[2] T. Helfer, J. Balaguer Skorohod-Olevsky Viscous Sintering Mdoel, 2019, Techincal report, DOI: 10.13140/RG.2.2.18724.32649

[3] Lee et al., Journal of American Ceramic Society, 2003, 86(6): 877 - 82.