Altair HyperXtrude 2021.2 Release Notes

Updates to H3D Writer Module

H3D files exported from the solver now use version 2021.0 of the H3D writer libraries. This change enabled the interfaces to create on-demand results in SI units from the exported results in user units. However, this did not support plotting the results on boundary faces. In this version, an additional FEM file containing the BC data is also exported from the solver to enable plotting the results on BCs. With these two changes, the solver will no longer have to write two H3D files to support post-processing in Inspire and HyperView. It will export only results in user-specified units. Interfaces will automatically generate the results in SI units on-demand at the time of post-processing.

Binder jet printing is a 3D additive manufacturing technique and it is becoming increasingly popular. It is used to make complex metal parts with excellent mechanical properties. It costs significantly less than other AM methods. This process is followed by a heat treatment process in an oven known as sintering. This is a long process and typically happens for 10+ hours. During this process, the metal particles in the part come together leading to part densification and significant part shrinkage. This heat treatment process has three stages: Debinding, Sintering, and Cooling. The part before the debinding stage is called the green part, and after the debinding stage is called the brown part. The process of debinding removes the binding agent and it can be done chemically. The main design questions are centered around the sintering part of the heat treatment process. During this stage, the metal particles come together and the part densification occurs. The key questions a designer of this process may have are:

• For a given brown part and the oven temperature curve, how much does the part shrink?
• What should be the size of the green part to achieve the final part shape and tolerances?
• How to evaluate the efficiency and improve the oven temperature curve?

The goal of the simulation is to answer these questions. The binder jet sintering solver helps to simulate part densification, shrinkage, and distortion during the sintering process; and to answer the above-mentioned design questions. It helps the user to perform virtual trials to determine the dimensions of the green/brown part that will ensure that the sintered part dimensions are within acceptable tolerances.

The solver supports two types of analysis:

• Shrinkage prediction analysis
• Geometry-based shrinkage compensation analysis

Shrinkage Prediction Analysis

The goal of this analysis is to predict the part shrinkage for the given green part and the oven temperature curve.

Supported Physics:

Densification and shrinkage are computed by solving time-dependent conservation equations that govern the change in relative density, temperature, and deformation during the sintering process. The solver can use the given oven temperature curve and has an additional set of control parameters to efficiently start and end the solution. The solution can be started from the initial point of the curve or at a desired point in the curve before part densification The key features of this solution type are:

• Temperature can be either computed or assumed to uniformly follow the curve.
• The solution can start at any temperature below the densification start temperature.
• It can end when the part densification is achieved or continue until the cooling cycle is completed.
• The solver is implicit and can compute accurately with large time steps for a quick solution.

Supported Results:

• Displacement
• Velocity
• Temperature
• Relative Density
• Sinter Deviation
• Volume Change

Figure 1.

Geometry Based Shrinkage Compensation Analysis

The goal of this analysis is to determine the brown part geometry from the desired final part shape and tolerances. This is an efficient design optimization analysis and is completely performed by the solver without any external optimization tool.

Supported Physics:

• This module uses the physics and algorithm implemented for shrinkage prediction analysis and adds a design optimization module.
• For the given sinter deviation (the maximum allowed part size deviation) and the maximum number of design iterations, the solver performs optimization iterations.
• In each design iteration, shrinkage prediction analysis is performed on the shape corrected from the previous iteration, then the sinter deviation is computed. Based on this, the part size is corrected to perform the next design iteration.

Supported Results:

• Results of shrinkage prediction analysis for each design iteration
• Corrected brown part shape in STL format

User-Defined Functions for Material Model - Element ID

The solver sends all relevant data to the interface to compute the flow stress and this previously did not include the element ID. In this version, based on a customer request, the correct element ID is also included in this list.

User Control in Partitioning Dead Cycle Time

The solver has added a new parameter DeadCycleTimeInContainer to control how much of the dead cycle time (DeadCycleTime) is inside the Container. The rest of the time, it will be outside the Container. If this parameter is not specified, the solver will use the current default of 50% of the time outside the Container and the remaining 50% of the time inside the Container.

Symmetric model gives Asymmetric OptiStruct results

The HyperXtrude solver writes the loads on the workpiece - tool interfaces to perform the stress analysis using OptiStruct. These loads are computed by performing the extrusion analysis until the specified convergence. The loss of symmetry could happen for a few different reasons, such as asymmetric mesh and numerical oscillations. In this particular case, the projection and mapping algorithm was improved to address this issue in the reported model.

Incorrect Heating During Dead Cycle Time

The dead cycle time is divided into two parts. In the first part, the hot billet is outside the container. Therefore, it should lose heat, and the container also should lose heat. The solver was showing the container gaining heat due to incorrect convection reference temperature. This issue is resolved by correctly setting the reference temperature.

Random Memory Errors

An underlying cause for a random memory error was identified and resolved.

Pin Deflection Analysis - Post Processing in Inspire FSW

The exported FEM file is now organized into components using HM comments to enable post-processing of the pin deflection analysis results in Inspire FSW. In this release, Inspire will automatically launch the OptiStruct run and enable post-processing.

Initial Conditions using Extrusion Results

The solver is enhanced to automatically map extrusion flow analysis data, temperature and grain size as the initial conditions for the quenching analysis. This feature is also available for use in Inspire Extrude.