Altair Manufacturing Solver 2022.1 Release Notes
General
Altair Manufacturing Solver is a state-of-the-art solver suite for manufacturing
applications built on a parallel, modular, and extensible framework that is suitable
for simulations of manufacturing processes. This release contains solutions for the
following modules:
- Additive Manufacturing
- Injection Molding (3D Solution)
- Injection Molding (Shell Solution)
- Metal Casting
- Welding
- Polymer Material Data Analytics (PMDA)
Highlights
Highlights for this release include:
- A new shell solver (AMS-Shell) that performs injection molding analysis is introduced in this release. This solver is used to compute molding process window, gate location optimization, and real-time flow pattern analysis using the surface mesh of the part. This solver has an interface in Inspire Mold.
- The 3D injection molding solver now supports multistage analysis.
- The welding solver has added residual stress analysis with springback.
- The casting solver can now do thermomechanical analysis in a parallel distributed environment using MPI.
Additive Manufacturing (3DP for SLM)
New Features
- Detection of inverted elements due to large part displacements
- During the printing process, the part accumulates a significant amount of deformation, especially at the top surface. For each time step, the solver checks if the Z displacement of any node in the model becomes larger than the element size. If such scenarios are detected, the solver prints an error message stating that elements are inverted due to exceedingly large deformation. The solver stops the computation and exits after issuing the error message. (AMSLVR-418)
- Implementation of the layered integration for lumping process
- A new lumping algorithm is added in this release. Instead of relying on a universal lumping factor for all the elements, this method computes the deposition effect locally for each individual element. As a result, the final displacements in the parts are more accurate for the given mesh size, thus allowing use of coarser discretizations without compromising the accuracy. (AMSLVR-425)
- Prediction of part collision with the recoater during the printing process
- During the printing process, if the Z displacement of the part during the printing stage becomes larger than the powder layer thickness, the part collides with the recoater which spreads the powder across the top surface of the part. For each time step, the solver checks if the Z displacement of any node in the model is larger than the powder layer thickness specified in printer settings. If this is detected, the solver prints a warning message stating that a recoater collision will occur. The solver does not stop the computation when such a recoater collision is detected. The location of the collision is exported in an .h3d file and can be visualized in the GUI. (AMSLVR-417)
Enhancements
- Wall roughness - A new surface quality result
- Due to the nature of the layer-by-layer deposition process, non-smooth surfaces of the printed part are common. The solver calculates the roughness of the surface at each node by obtaining the magnitude of the displacement vector when the top layer of elements is activated. This result is exported in the .h3d file and can be visualized in the GUI. (AMSLVR-419)
Metal Casting
New Features
- Performing thermomechanical analysis in a parallel distributed environment
- The MPI support is now extended to the thermomechanical analysis and runs in a distributed environment. This can save computational time by running in parallel. (AMSLVR-416)
- Mesh quality checks
- The casting solver checks for mesh quality at the very beginning of the analysis and prints the summary of the mesh quality results in a separate file. For some checks such as the aspect ratio bounds, you can choose to proceed with analysis even if the quality is not good. However, for checks such as elements that should be connected by face and not just by edge, the solver prints error messages and stops the run. (AMSLVR-392)
- User-defined custom python processes
- The solver supports custom processes and in order to enable user-defined custom processes, now they can be located in any user folder location. This is a very advanced feature with no user interface. Using these custom processes you can now compute additional derived results, advanced BC specifications, specify custom bounds for variables used in other computations, and so on. (AMSLVR-393)
Enhancements
- Improved linear solver for the thermomechanical analysis
- A new linear solver is implemented that greatly improves calculation times for thermomechanical analysis and the benefits extend to both SMP and MPI parallel computations. (AMSLVR-388, AMSLVR-441)
- Improved MPI load balance for solidification analysis
- A new partitioning procedure is implemented to improve the workload balance in a parallel distributed computing environment for solidification analysis. (AMSLVR-434)
- Improved contact algorithm for Thermomechanical Analysis
- Several minor improvements are implemented in the contact algorithm between the part and the mold. This results in faster convergence and savings in computational time.
- Additional information on filling Analysis
- The solver prints additional information for filling analysis. The information printed varies based on the analysis. For example, the inlet velocity/flow rate in a pressure-driven filling is printed and this result is difficult for you to determine directly. The additional information presented includes step number, elapsed time, % completed, imposed velocity, fill time, and flow rate. (AMSLVR-336)
Resolved Issues
- Issue due to larger fill times in a constant level filling
- In a constant level filling, the inlet radius is adjusted dynamically to maintain a constant level/head. The inlet opening and closing are inversely proportional to fill time. When the fill time is very large, this was leading to a slow adjustment of the opening and causes difficulties in maintaining the level. This issue is resolved by improving the function used to adjust the inlet opening radius. (AMSLVR-432)
- Error in multicycle analysis in MPI runs
- The issue with multicycle simulation not working properly in MPI (distributed computing) is now resolved. (AMSLVR-437)
- Inlet computation in MPI for a constant level filling
- The issue in computing the inlet radius and the center for filling analysis in a distributed computing environment using MPI is now resolved. This was for the casting analysis that maintains the level/head in the sprue. (AMSLVR-489)
Injection Molding (3D)
New Features
- Multistage analysis
- The solver is now enabled to perform multistage analysis. Using this option, you can now perform multiple cooling analyses before the final filling and packing analysis. For example, a sequence of analyses such as cool + cool + cool + fill + pack + cool + warp can be performed. This helps with understanding the cyclic thermal behavior of the mold during the multiple injection cycles. In the subsequent injection cycles, more realistic thermal conditions of the mold can be specified based on the results of these simulations. (AMSLVR-356)
- Part weight computation during packing analysis
- During the packing stage, additional material is packed into the part cavity making use of the compressibility of the material to compensate for shrinkage. The part weight increases due to the additional packed material until the gate freezes. This increase in the part weight during the packing stage is shown as an additional column in the *.out file and also in the CSV file. (AMSLVR-469)
- Fiber orientation computation during packing analysis
- In this release, fiber orientation is computed during the packing stage as a continuation of the computation during the filling stage. At the end of the pack analysis, fiber orientation data is written to an XML file similar to the one written for fill analysis in earlier releases. (AMSLVR-467)
- Exporting warpage data deck in MPI
- The solver now writes the OptiStruct data deck used for warpage analysis even when parallel computation is used for analysis. This enables you to run the entire analysis: fill, pack cool, and warp in MPI. (AMSLVR-478)
Enhancements
- Injection pressure profile after V/P switchover
- During the filling stage, the molding process switches over from velocity control to pressure control; and this process is called the V/P switchover. In the molding machines, pressure can be specified as a profile (changing with time) during the pressure control phase. The solver is now enhanced to allow the pressure specification using time-dependent table data. (AMSLVR-470)
- Improvements to automatic timestep estimation in packing analysis
- The molding solver supports a detailed 3D packing solver and a fast hybrid 3D packing solver. The automatic time step determination used by the detailed packing solver is enhanced to perform packing simulation more efficiently. (AMSLVR-464).
- Improvements to the fast packing simulation performance
- The linear equation solver used by the fast hybrid 3D solver is changed to an efficient direct solver to improve the efficiency and the convergence of the fast packing analysis. (AMSLVR-463)
- Enforcing gate freeze in packing analysis
- During the packing process, the packing of the additional material in the part cavity ceases after the polymer in the gate freezes and this determines the end pressure-hold. The molding solver determines the gate freeze during the packing simulation. However, this was not enforced in ending the pressure holding phase of the packing stage, and pressure-hold was maintained for the specified duration of the holding phase. In this release, the solver ends the holding phase as soon as the gate freezes and then switches to the computation of pressure relaxation on the mold wall to compute sink marks. This results in saving computational time. (AMSLVR-459)
- Determination of no-flow temperature
- From this release, the NO_FLOW_TEMEPRATURE is not expected in the material data. The solver now internally computes the no-flow temperature based on the ejection temperature and Tait data. (AMSLVR-460)
- Fiber-fiber interaction
- The solver now considers fiber-fiber interaction in the fiber orientation computation. This improves the accuracy of fiber orientation computations. (AMSLVR-164)
Injection Molding (Shell)
The AMS-Shell Injection Molding solver is a very fast shell solver. It computes the
solution using the surface mesh created on the part and gates. The features made
available in this release are:
- Fast filling analysis
- Gate location optimization
- Molding process window computation
To compute the results, the solver needs only a triangular surface mesh that accurately captures the part geometry. This solver has an interface in Inspire Mold version 2022.1.
New Features
- Fast filling analysis
- This feature includes a quick filling analysis for the given gate locations for a single part to see the flow pattern in real-time. To use this feature, only one part cavity needs to be modeled and it can have multiple gates. The solver does not require the entire runner system and it is not supported. (AMSLVR-461)
- Optimization of gate locations
- This feature includes quick determination of the optimal gate locations for the given part. The optimal gate locations are found iteratively to satisfy the balanced filling of the part. This analysis is focused only on a single cavity and multiple cavities are not supported.
- Molding process window simulation
- This feature is used to simulate the molding process window using fast variable strip analysis. This is achieved through implementation of thickness analysis, flow pattern analysis, and variable strip analysis for various process conditions. The solver requires only a surface mesh on the part and it computes all other mesh/data structures internally. The solver uses the same material data file as provided for the AMS 3D solver. Molding window results are shown for a range of fill times, melt temperatures, and mold temperatures. (AMSLVR-214)
Welding
New Features
- Residual stress analysis with springback
- In this release, the welding solver adds support for residual stress analysis for the welded parts. Nonlinear elastoplastic material models are implemented for the welding solver in order to predict plastic deformation induced by the welding process. Springback analysis is supported through inertia relief in order to reach the final deformation state with minimal external constraints. (AMSLVR-310)
- Quadratic tet element in thermal and coupled analysis
- The welding solver adds support for second-order tet elements for both the thermal and coupled analysis. Quadratic tets offer better properties than linear tets in resolving displacement and stress results, thus requiring fewer elements and/or layers in the heat-affected regions than a regular linear tet. (AMSLVR-413)
Enhancements
- Write peak temperature as a static result instead of transient
- Peak Temperature was previously written in .h3d files as a transient result. With this enhancement, the solver stores peak temperature in memory during the simulation and write the result to the .h3d file only at the end of the last step of the simulation. This avoids creating large .h3d files without losing any useful information. (AMSLVR-452)
- Add total time and time step summary in .out file
- In the .out file, the solver prints the total run time and the time step information as a summary. For example, the total welding time, time step size, and the total number of time steps are now exported to the .out file. These time step statistics are helpful for understanding the run time. (AMSLVR-289)
Resolved Issues
- Incorrect calculation of heat source orientation
- This is a bug fix regarding the incorrect calculation of the orientation vector for the double ellipsoid heat source as it is moving along a user-defined path. The fix implements piecewise linear interpolation of the orientation vector based on the current heat source center location and the orientation vectors of its neighboring nodes. (AMSLVR-475)
Polymer Material Data Analytics (PMDA)
New Features
- New viscosity models
- PMDA now supports the power-law model and Herschel-Bulkley fluid.
- New temperature dependence models
- PMDA now supports Arrhenius and the exponential model.
Enhancements
- Additional level-1 testing rules
- Improvements to current rules and additional common-sense rules (for example, B6 has to be in the order or less than 1.0e-07 K/Pa or less) have been added.
Resolved Issues
- Either table or scalar value for thermal conductivity and specific heat
- PMDA was expecting the scalar (average) value even when the table data was present. This requirement was removed so the material JSON file can have either the scalar value or the table data.
- Displaying messages/images during a batch operation
- PMDA is mainly used in batch mode. The material report generation no longer pops up images when invoked from SimLab.