Automate Design Trade Studies Using Simcenter HEEDS and Simcenter 3D

Automate Design Trade Studies Using Simcenter HEEDS and Simcenter 3D

Learn how Siemens Simcenter HEEDS and Simcenter 3D can be used to automate design trade studies. This blog post will focus on highlighting an example workflow and give insight into what the user experience looks like. Take advantage of this knowledge to understand what type of problems this workflow can help you solve efficiently.

Challenges:

  • Hand-calculations not readily applicable to design problem. Need finite-element model (FEM).
  • Exploring design space manually with FEM is time-consuming with slow turnaround
  • Traditional optimization routines require significant subject matter expertise

Values:

  • Simcenter HEEDS and Simcenter 3D enable automation of FEM-based analysis studies, improving turnaround.
  • Developed process can be quickly updated to perform alternate trade studies.
  • Developed process used for design space exploration can be leveraged to perform optimization studies.

Example Geometry and Problem Description

To explore the utility of automating trade studies with Simcenter HEEDS and Simcenter 3D, imagine a rectangular orthotropic fabric material with a circular hole cutout at the center, as shown below. The fabric can only react tension loads. The hole is locally stiffened with a cord-reinforced grommet, and the grommet is contained within a square box boundary that is used to support a cover flap for the hole.

Further, if we imagine that the hole is to serve as an access point for the system internal hardware, then we might be interested in seeing how large we can make this hole while still showing positive margins. So, let’s consider the case where the fabric material is pinned at the four corners and a uniform vertical tension is applied along the top and bottom edges of the fabric, as shown below. We will subject our model to this load case as part of a trade study that aims to track the minimum margin in the fabric and cord with varying hole diameter.

Step 1: Defining Trade Study Parameters in Simcenter 3D

The CAD geometry is created in Simcenter 3D. For each step in the CAD generation process, the inputs to the geometry tools can be tied to user-defined parameters, thereby creating a parametric geometry model that automatically updates when the user-defined parameters are changed. These automatic updates propagate forward into the resulting mesh made in Simcenter 3D and the Nastran analysis-deck it writes out. As shown in the image below, multiple user-defined parameters are used to build our fabric-with-hole geometry. In particular, we will be interested in varying the hole_diameter parameter and seeing the impact of this change on the resulting margins.

Step 2: Setting Up Trade Study in Simcenter HEEDS

Setting up the Simcenter HEEDS study begins by laying out the process required to automate the evaluation of a single design point. The process is outlined with the use of modules, or what Simcenter HEEDS calls “portals,” that handle the runtime execution of tools available to the user (both commercial and custom) as well as the exchange of input and output files between the tool and Simcenter HEEDS. The process used to evaluate a single design point in our current study is shown below; it requires a single portal to complete the trade study. We will use Simcenter HEEDS to perform the following workflow:

  1. Pass updated hole_diameter parameter value into Simcenter 3D
  2. Execute Simcenter 3D to update geometry, update mesh, update analysis-deck, and run Nastran
  3. Read in stress and strain information from the Nastran OP2 files and compute margins
  4. Repeat steps 1-3 for all hole_diameter values of interest

With the process laid out, the next step is to tell Simcenter HEEDS what variables to track and how to pass the variable value into the design evaluation process. The only trade study variable will be the hole diameter, which is defined in the Simcenter 3D CAD part file. To have Simcenter HEEDS drive changes to this variable, first we define the variable name and type (as shown in the first image below) and then we use the Simcenter 3D portal to link the Simcenter HEEDS hole diameter variable to the Simcenter 3D hole diameter CAD expression (as shown in the second image below).

jan-7

The next step is to tell Simcenter HEEDS what responses to track and how to read them from output data generated by the design evaluation process. To have Simcenter HEEDS read and process output data, first we define the response names and types (as shown in the first image below) and then we use the Simcenter 3D portal to link the Simcenter HEEDS responses of type Tag to the Simcenter 3D Nastran solution data (as shown in the second image below for Fabric_Grommet_Max_Principal_Strain). The Simcenter HEEDS responses of type Formula are determined within Simcenter HEEDS, leveraging Python and linked variable/response data for determining additional responses of interest, such as the margin values for the fabric and cord.

Jan-9

Finally, the study definition is provided to Simcenter HEEDS. In our case, we will choose to evaluate multiple design points with a sweep in the hole diameter trade variable, as shown below. Lastly, kick-off the trade study.

jan-10

Step 3: Visualizing Results

After performing the trade study analysis runs, the next step is to visualize the results. Simcenter HEEDS provides powerful built-in visualization tools that help to automate this step of the process. Simcenter HEEDS can also be used to synchronize multiple result plots and images to help the user gain insight into the mechanisms underlying the design performance trends.

For example, a synchronized set of plots and images are shown below in order to give an overall summary of the trade study results. The results shown include:

  • Top left: minimum strain margin in fabric vs. hole diameter
  • Top right: minimum stress margin in cord vs. hole diameter
  • Bottom left: contour of max principal strain in fabric (outside grommet) for current hole diameter selection
  • Bottom middle: contour of max principal strain in grommet for current hole diameter selection
  • Bottom right: contour of max stress in cord for current hole diameter selection

Altogether, these synchronized results demonstrate the power of our automated workflow: Simcenter HEEDS updates the geometry parameters in Simcenter 3D, Simcenter 3D updates the geometry and associated mesh and writes out an updated analysis deck, the analysis case is run, and the post-processing of these results is performed and compiled by Simcenter HEEDS for user review. After setting up this process once, little effort is required to vary alternate geometric parameters and regenerate the same plots.

Looking at the margin plots first, we note that the minimum margin associated with the fabric and cord both decrease with increasing hole diameter, as expected. The minimum strain margin in the fabric is negative for hole diameters above 1150, while the minimum stress margin in the cord remains positive for all hole diameters considered. Interestingly, the shape of the minimum fabric margin plot does not appear to be as smooth as the minimum cord margin plot, and the sensitivity of the fabric margin to hole diameter does not vary monotonically.

The next set of plots and images shown below are provided in order to better understand the mechanisms underlying the fabric strain response behavior. The results shown include:

  • Top left: minimum strain margin in fabric vs. hole diameter
  • Top right: max principal strain in fabric (both inside and outside grommet region) vs. hole diameter
  • Bottom left: contour of max principal strain in fabric (outside grommet) for current hole diameter selection
  • Bottom right: contour of max principal strain in grommet for current hole diameter selection

Looking at the top right plot, we note that the region seeing the peak max principal strain in the fabric varies with hole diameter. Initially, the peak strain occurs outside of the grommet region and increases logarithmically with hole diameter towards the lower range of hole diameter values considered. Within the grommet region, however, the peak strain increases linearly with hole diameter. The peak strain in the grommet surpasses strains outside the grommet region for a hole diameter of 800, explaining the sudden shape change in the minimum margin plot shown in the top left. Both fabric regions exhibit exponential variation in the observed peak strain with increasing hole diameter towards the upper range of values considered. The exponential sensitivity is greater outside the grommet region, surpassing the strains observed within the grommet for a hole diameter of 1200, explaining once again the shape change in the minimum margin plot. Reviewing the strain contours in the bottom row as the hole diameter increases makes clear that the exponential growth observed is due to interaction of the hole with the fabric edges. The presence of edge effects also explains why the peak strain observed outside of the grommet and near the edges is more sensitive to changes in the hole diameter toward the upper range of values considered.

The next set of plots and images shown below give a more detailed summary of the stress and deformation response of the cord reinforcement along the hole perimeter. The desire is to quantify the level of distortion observed in the hole when fully loaded. The results shown include:

  • Top left: minimum stress margin in cord vs. hole diameter
  • Top right: contour of max stress in cord for current hole diameter selection
  • Bottom left: relative change in deformed hole vertical/horizontal radius vs. hole diameter
  • Bottom right: ratio of relative change in deformed hole vertical/horizontal radius vs. hole diameter

The bottom left plot indicates that the cord deforms more in the vertical direction than in the horizontal direction. This matches our expectation that the larger deformations will be aligned with the direction of applied tension. The ratio of these directional deformations is computed and plotted in the bottom right, thereby quantifying the extent by which the circular hole ovalizes. This ratio increases by less than 0.5% for a hole diameter of 1150 mm, at which point the minimum fabric margin goes negative. Thus, little hole distortion is observed and does not drive design decisions for this particular example.

Summary

We looked at the utility of leveraging Simcenter HEEDS and Simcenter 3D in order to automate a FEM-based trade study. As an example, we set out to quantify the largest hole diameter that can be cutout from an orthotropic fabric material while still ensuring positive operational margins. As part of the automated process, Simcenter HEEDS was used to automatically post-process and visualize results data in order to develop insight into the mechanisms underlying the fabric strain response and the cord-reinforced hole deformation response. Finally, the automated process highlighted in this post can be altered with little user effort, allowing for quick evaluation of alternate trade studies or even an optimization problem that seeks to maximum the zero-margin hole diameter by varying multiple design parameters.

Written by Jose Marquez

José Márquez is an Aerospace Engineer with a background in loads analysis, multi-disciplinary optimization, and software development. Outside of work, he enjoys gardening and rock climbing.