Running Parametric Analyses
After preparing your model, you are ready to run a parametric analysis.
Beginning a Design Study, DOE, Temporary Settings Sweep, or Optimization Analysis
The first steps in running a parametric analysis are the same for all types of parametric analyses. Follow the instructions below to begin a parametric analysis and then follow the instructions at the end for completing the different types of parametric analysis.
Note: | You may want to change the display, output, and optimizer options before you run a parametric analysis. Learn more about Optimization Settings. |
To begin a Design Study, DOE, Temporary Settings Sweep, or Optimization Analysis:
1. On the ribbon menu, click the Design Exploration tab. From the Design Evaluation container, click the Design Evaluation Tools icon 
.
. The Design Evaluation Tools dialog box appears.
2. In the Model text box, enter the name of the model to analyze.
3. In the Simulation Script text box, enter the name of the simulation script to use. Learn about Creating a Simulation Script.
4. Select either Measure or Objective to define the type of objective you are using. Learn about computing a measure of performance (objective).
■If you selected Measure, select Last, Minimum, Maximum, or Average from the pull-down menu, and enter the name of the measure in the text box.
■If you selected Objective, enter the name of the objective in the Objective text box. Optimizations are limited to one objective. You can monitor more than one objective in a design study or DOE, however, by entering more than one name separated by a comma.
5. Select Design Study, Design of Experiments, Temporary Settings Sweep, or Optimization. For a comparison of the types of parametric analyses, see About Parametric Analysis Tools.
6. Follow the instructions in the following to complete the parametric analyses:
Tip: | Before running the parametric analysis, you can preview it by selecting Preview. Preview shows you each configuration of your model for every design variable. It displays an alert box asking you if you want to pause after each configuration. Select YES to pause. |
Tip: | By default Adams View writes out a parasolid file of the model geometry for each run in the parametric analysis. This can take up time and disk space, especially for models with large geometry. If the model geometry is not changing run-to-run, then it may be advantageous to instruct Adams not repeat this geometry writing process for each run. This can be done by clicking the "Output" button for the Design Evaluation Tools dialog, setting "Output Category" to "Database Storage" and selecting "Yes" for "One Parasolid". |
Completing a Design Study
Learn about how to complete a Design study and how Adams View stores the results of an analysis:
Completing and Executing a Design Study
To complete the information for a design study and execute the analysis:
1. Begin the analysis as explained in Beginning a Design Study, DOE, or Optimization Analysis.
2. In the Design Variable text box in the Design Evaluation Tools dialog box, enter the name of the design variable that you want to vary.
Note: | You cannot enter more than one design variable in the text box. If you need to enter more than one design variable, then perform a DOE. |
3. If you specified only a range for the design variable, enter the number of levels (values) you want to use in the Default Levels text box.
4. Select Start.
Adams View runs a simulation for each level of the design variable. When the simulations are done, Adams View returns the variable to its original value.
■If you specified only a range for the design variable, Adams View uses equally spaced levels across the range. You specify the number of levels in the Default Levels text box.
■If you specified a list of values for the design variable, Adams View runs a simulation using each value, ignoring the Default Levels text box.
For more information about variable ranges and values, see Controlling Variable Values.
Analysis Results
Adams View creates an analysis object named Last_Multi under your model. The analysis contains a result set named Design_Study_Results. The result set contains the following components:
■Trial, which contains the number of each run (one through the number of runs).
■Component with the same name as the design variable, which contains the values used for the variable for each run.
■For each measure or objective, a component with the same name as the objective or measure, which contains the values of the performance measure for each run.
Note: | If the analysis Last_Multi already exists from a previous parametric analysis, Adams View deletes the previous results and replaces them with the new results. Learn how to save results permanently at Saving Results. |
Completing a DOE
Learn how to complete a Design of experiments (DOE) analysis and how Adams View stores the results of the analysis:
Note: | The enhanced DOE capabilities found in Adams Insight, provide you with more sophisticated experiments and improved results in an easy-to-use interface. For more information, see the Adams Insight online help, or contact your MSC sales representative. |
Completing and Executing a DOE
To complete the information for a DOE and execute the analysis:
1. Begin the analysis as explained in Beginning a Design Study, DOE, or Optimization Analysis.
2. In the Design Variable text box in the Design Evaluation Tools dialog box, enter the name of the design variables that you want to vary.
3. If you specified only a range for one or more of the design variables, enter the number of levels (values) you want to use for those variables in the Default Levels text box.
4. Set Trails defined by to Built-In DOE Technique, Direct Input, or File Input.
■If you selected built-in techniques, use DOE Technique to select the technique. If you want to check that the variables have the same number of levels and display the required number of runs, select Check Variables, Guess # of Runs.
■If you selected direct input, enter the number of trials (simulations) and the trial matrix.
■If you selected file input, enter the name of the file containing the trial matrix.
5. Select Start.
Adams View runs a simulation for each trial that the DOE technique or trial matrix defines. When the simulations are done, Adams View returns the variables to their original values.
The DOE technique or trial matrix controls the number of simulations and the combination of variable values to use for each simulation. For example, the Full Factorial technique simulates every possible combination of levels. If you use two variables with three levels each, Adams View runs nine simulations. For more information on the DOE techniques and entering your own trial matrix, see About Design of Experiments.
The DOE technique or trial matrix selects values for a variable based on the range or list of values you defined for the variable.
■If you specified only a range for a design variable, Adams View selects from equally spaced values across the range. You enter the number of values in the Default Levels text box.
■If you specified a list of values for a design variable, Adams View selects directly from those values, ignoring the value in the Default Levels text box.
■For more information about variable ranges and values, see Controlling Variable Values.
Analysis Results
Adams View creates an analysis named Last_Multi under your model. The analysis contains a result set named DOE_Results. The result set contains:
■Component named Trial that contains the number of each run (one through the number of runs).
■For each design variable, a component with the same name as the variable, which contains the values used for the variable in each run.
■For each measure or objective, a component with the same name as the objective or measure, which contains the values of the performance measure for each run.
Note: | If analysis Last_Multi already existed from a previous parametric analysis, Adams View deletes the previous results and replaces them with the new results. For instructions on permanently saving previous results, see Saving Results. |
Completing an Optimization
Learn how to complete an Optimization and how Adams View stores the results of an analysis:
Note: | If you simultaneously run optimizations of two Adams View models, from the same working directory, Adams View may crash. |
Completing and Executing an Optimization
To complete the information for an optimization and execute the analysis:
1. Begin the analysis as explained in Beginning a Design Study, DOE, or Optimization Analysis.
2. In the Design Variable text box in the Design Evaluation Tools dialog box, enter the name of the design variables to vary.
3. If you want to automatically save the original values of the design variables before starting the optimization analysis, select Auto. Save.
4. If you want to immediately save the current values of the design variables, select Save.
5. Set Goal to Maximize or Minimize.
6. If you want to add constraints, select Constraints and enter the names of the constraints in the text box that appears.
7. Select Start.
If you selected Maximize as the goal, Adams View adjusts the design variable values to increase the measure or objective as much as possible. If you selected Minimize, Adams View reduces the objective as much as possible.
If you specified value ranges on any of the design variables, Adams View increases or decreases the objective as much as possible without exceeding the value limits. For more information about variable ranges, see Controlling Variable Values.
If you specified constraints, Adams View increases or decreases the objective as much as possible without violating the constraints. For information about using constraints to control the optimization, see Creating Constraints (Optimization Only).
Adams View iteratively adjusts the design variable values, attempting to improve the model performance with each iteration. Adams View may need to backtrack to avoid violating a constraint or limit on a variable value. Therefore, the model performance does not necessarily improve with each iteration. At each iteration, Adams View runs several simulations to approximate derivatives and converge on the next iteration.
The last iteration will be the best values that the optimization could find without violating constraints or limits. Adams View normally leaves the design variables set to the optimized values. If you interrupt the analysis or Adams View encounters an error during the analysis, Adams View resets the variables to their original values.
If you do not want to keep the optimized values, and you selected the Auto. Save check box or used the Save button to save the original values, you can select the Restore button to return the variables to their original values.
Analysis Results
Adams View creates an analysis object named Last_Multi under the current model in the Modeling database. The analysis contains a result set named Optimization_Results. The result set contains:
■Component named Iteration that contains the number of each iteration (one through the number of iteration). Iteration zero is the initial model configuration.
■For each design variable, a component with the same name as the variable that contains the values used for the variable for each iteration.
■Component with the same name as the objective or measure that contains the values of the performance measure for each iteration.
Note: | If the analysis Last_Multi already existed from a previous parametric analysis, Adams View deletes the previous results and replaces them with the new results. Learn about Saving Results. |
Completing a Temporary Settings Sweep
A temporary settings sweep allows one to specify and execute a number of trials, each of which is defined by applying one or more Temporary Settings files to the baseline model.
Completing and Executing a Temporary Settings Sweep
1. Begin the analysis as explained in Beginning a Beginning a Design Study, DOE, Temporary Settings Sweep, or Optimization Analysis.
2. Populate the table with the desired number of rows, one per trial
3. In each cell enter one or more temporary settings file(s) to apply to the model. Each row/trial represents a single simulation that will be run
4. Select Start and Adams will run one simulation for each row in the table and then revert the model back to its baseline configuration
Analysis Results
Adams View creates an analysis named Last_Multi under your model. The analysis contains a result set named Temporary_Settings_Sweep_Results. The result set contains:
■Component named Trial that contains the number of each run (one through the number of rows in the table).
■For each measure or objective, a component with the same name as the objective or measure, which contains the values of the performance measure for each run.