The previous sections describe how to set up and perform a normal analysis in which all the input is known (or assumed) prior to the analysis, and one result is obtained. The properties of the model are known and entered, and the analysis is performed. Based on the results, you can change the input to reduce the stress to an acceptable level (obtain a larger factor of safety) or to reduce the weight (obtain a smaller factor of safety). In some cases, several iterations may be required before an acceptable design is reach.
Design Optimization is a stand-alone program that automates the process to obtain such an iterative result. Based on the objective (minimize the stress, deflection, and so on) and the constraints (plate thickness, total weight, and so on), Design Optimization will perform the iterations to determine the optimum design. The outcome of a design optimization analysis is a new model with the design variables set to the optimum values, the results for the optimized model (same as any stress analysis), and charts showing the history of the optimization.
A Design Study is similar to an optimization analysis. A design study will only create charts that show how the objectives are affected by changing each of the design variables.
The remainder of this document will refer to design optimization for both optimization and studies unless indicated otherwise.
The capabilities of the Design Optimization and Studies are these:
- Analysis Types:
- Static Stress with Linear Material Models
- Natural Frequency (Modal)
- MES with Nonlinear Material Models
- Static Stress with Nonlinear Material Models
- MES Riks Analysis
- Natural Frequency (Modal) with Nonlinear Material Models
- Design Variables:
- Plate/shell element thickness
- Membrane element thickness (MES/nonlinear stress only)
- Beam element cross section dimensions (enter from cross section library)
- Either continuous or discrete values for each design variable
- Objectives:
- Volume
- Maximum stress (von Mises or beam worst stress)
- Displacement magnitude
- Frequency
- Constraints:
- Minimize
- Maximize
- Set a lower limit
- Set an upper limit
The general method of setting up and performing the analysis is as follows:
- Build the complete model as normal.
- Enter a reasonable value for the design variables. While entering the values for beams and nonlinear shell elements, right-click in the field and choose Set as Design Variable. The
symbol will appear next to the field. For linear plate elements, a check box appears in the Design Variable column to indicate that the plate thickness is a design variable.
- Run the analysis. Although not strictly necessary, this step is recommended to get realistic seed values (starting values) for the design optimization input.
- Access the Design Optimization dialog using Analysis
Analysis
Optimization.
- Specify the lower and upper limits of each design variable. See the page Design Variables for details.
- Specify the objectives and constraints of the analysis. See the page: Objectives & Constraints for details.
- Specify the solution parameters. See the page Solution Parameters for details.
- Run the design study or optimization. See the page Running the Analysis for details.
- Review the results. See the page Reviewing the Results for details.
- Design study results consist of a series of graphs showing the sensitivity of the objectives to each design variable.
- Design optimization results consist of a series of optimization history graphs for each objective and a complete FEA model and results of the optimized design which can be reviewed in the Results environment.
Tip: The units used within the Design Optimization interface are based on the active Display Units when
Analysis
Analysis
Optimization is given. Changing the Display Units in Autodesk Simulation when the Design Optimization dialog is opened will have no effect on Design Optimization.