Managing Variables#

To manage design-level variables, right-click the design entity in the Project Manager window, then select Designer Properties....

Once a variable is created in the manager, it is editable in the Properties window in the bottom-left corner of the workspace, whenever the designer entity is focused.

IMPORTANT

Make sure the variables are on design-level, instead of project-level. The latter is managed via Project -> Project Variables.... It is theoretically possible that project-level variables can also be accessed by Python scripts, but it seems to be quite complex.

Applying Variables to Design#

To refer to a design variable, use its name directly. Notice if you (insist to) use project-level variables, then they should be referred with their names and a ’$’ in front. E.g., this screenshot shows how to create a equation-driven curve with design variables A and B 👇

Configuring Setup and Plot#

I guess you can configure the analysis setup and result plot in PyAEDT, but it is just unnecessary when you can do that with GUI.

1. Configure a setup like you normally do. Notice the time step and stop time can be controlled by design variables as well. Remember the name of setup (by default it should be “Setup1”).
2. Put the simulation data you want to export into a rectangular plot, remember the name of the plot.

If you just began with an empty design, then the plot would be empty like this, it’s fine:

NOTE

Yes it’s possible to export data to multiple plots, and I think it is possible to export field report or others. But currently I have only validated PyAEDT with a single transient rectangular plot as the output.

The parametric model is ready now. We can move on the the Python API system and figure out how to use Python script to overtake those variables and do the simulation.

AutoSim#

It is explained below how to manipulate the parametric model with PyAEDT. First, launch Ansys Electronics Desktop in non-graphical mode, open the Ansys project file (*.aedt) and assign the design to simulate.

from pyaedt import Maxwell2d
from pyaedt import Desktop

Desktop(specified_version="2024.2", non_graphical=True, new_desktop_session=False)
maxwell = Maxwell2d(projectname=projectPath, designname=designName)

IMPORTANT

Here the projectname of Maxwell2d() should be the absolute path of the .aedt file, while designname is just a string of name of the design.

Then read the simulation plan and store each set of parameters in a dictionary.

with open(planPath, mode='r', newline='', encoding='utf-8-sig') as file:
    reader = csv.DictReader(file)
    simPlan = [dict(row) for row in reader]

# An example of simulation plan (*.csv):
# K,wr,Rin,lmin,Next,Nind,SE,phis,DE,phid2,phid1,rpm,Tstep,Tsim
# 1.5,2,20mm,1mm,20,20,0.02mm,0deg,0.02mm,0deg,0deg,100rpm,50us,0.6s
# 1.5,2,20mm,1mm,20,20,0.02mm,0deg,0.02mm,0deg,0deg,-100rpm,50us,0.6s

Then let the simulations begin. For each round of simulation:

1. Set the design variables according to the plan, the variables can be accessed by using maxwell object like a dictionary, with the variable names as keys.
2. Call analyze_setup() to initiate certain analyze setup. If there is an HPC license, the simulation can be set to ran in parallel mode.
3. Call post.export_report_to_csv() to export the data of certain plot. The csv file is named as the plot by default, it should then be renamed immediately to avoid overwrite.
4. Save the project by save_project().

simIndex = 0
for simSet in simPlan:
    simIndex = simIndex + 1
    for variableName in simSet:
        maxwell[variableName] = simSet[variableName]
    
    maxwell.analyze_setup(name=setupName, cores=12, tasks=12)

    fileName = str(simIndex) + '.csv'
    maxwell.post.export_report_to_csv(project_dir=outputDir, plot_name=plotName)
    os.rename(f'{outputDir}\{plotName}.csv', f'{outputDir}\{fileName}')
    
    maxwell.save_project()

    time.sleep(sleepSlot)

The release version of the script can be found at: