Note

Go to the end to download the full example code.

Multiphysics: Circuit-HFSS-Icepak coupling workflow#

This example demonstrates how to create a two-way coupling between Circuit-HFSS designs and Icepak.

Let’s consider a design where some components are simulated in HFSS with a full 3D model, while others are simulated in Circuit as lumped elements. The electrical simulation is done by placing the HFSS design into a Circuit design as a subcomponent and by connecting the lumped components to its ports.

The purpose of the workflow is to perform a thermal simulation of the Circuit-HFSS design, creating a two-way coupling with Icepak that allows running multiple iterations. The losses from both designs are accounted for: EM losses are evaluated by the HFSS solver and fed into Icepak via a direct link, while losses from the lumped components in the Circuit design are evaluated analytically and must be manually set into the Icepak boundary.

On the way back of the coupling, temperature information is handled differently for HFSS and Circuit. For HFSS, a temperature map is exported from the Icepak design and used to create a 3D dataset; then the material properties in the HFSS design are updated based on this dataset. For Circuit, the average temperature of the lumped components is extracted from the Icepak design and used to update the temperature-dependent characteristics of the lumped components in Circuit.

In this example, the Circuit design contains only a resistor component, with temperature-dependent resistance described by this formula: 0.162*(1+0.004*(TempE-TempE0)), where TempE is the current temperature and TempE0 is the ambient temperature. The HFSS design includes only a cylinder with temperature-dependent material conductivity, defined by a 2D dataset. The resistor and the cylinder have matching resistances.

The workflow steps are as follows:

  1. Solve the HFSS design.

  2. Refresh the dynamic link and solve the Circuit design.

  3. Push excitations (HFSS design results are scaled automatically).

  4. Extract the resistor’s power loss value from the Circuit design.

  5. Set the resistor’s power loss value in the Icepak design (block thermal condition).

  6. Solve the Icepak design.

  7. Export the temperature map from the Icepak design and create a new 3D dataset with it.

  8. Update material properties in the HFSS design based on the new dataset.

  9. Extract the average temperature of the resistor from the Icepak design.

  10. Update the resistance value in the Circuit design based on the new resistor average temperature.

Perform required imports#

Perform required imports.

from pyaedt import Circuit
from pyaedt import Hfss
from pyaedt import Icepak
from pyaedt import downloads
import os

Download and open project#

Download the project archive. Save it to the temporary folder.

project_path = downloads.download_file("circuit_hfss_icepak", "Circuit-HFSS-Icepak-workflow.aedtz")

Open the project and get the Circuit design.

circuit = Circuit(
    project=project_path,
    new_desktop_session=True,
    specified_version=242,
    non_graphical=False
)

C:\actions-runner\_work\_tool\Python\3.10.9\x64\lib\subprocess.py:1072: ResourceWarning: subprocess 4592 is still running
  _warn("subprocess %s is still running" % self.pid,
C:\actions-runner\_work\pyaedt\pyaedt\.venv\lib\site-packages\pyaedt\generic\settings.py:437: ResourceWarning: unclosed file <_io.TextIOWrapper name='D:\\Temp\\pyaedt_ansys_e6365f2f-3b8c-40bb-8a68-3f4e6e7c8dd5.log' mode='a' encoding='cp1252'>
  self._logger = val

Set the name of the resistor in Circuit.

resistor_body_name = "Circuit_Component"

Set the name of the cylinder body in HFSS.

device3D_body_name = "Device_3D"

Get the Icepak design#

Get the Icepak design.

icepak = Icepak(project=circuit.project_name)

Set the parameters for the iterations#

Set the initial temperature to a value closer to the final one, to speed up the convergence.

circuit["TempE"] = "300cel"

Set the maximum number of iterations.

max_iter = 5

Set the residual convergence criteria to stop the iterations.

temp_residual_limit = 0.02
loss_residual_limit = 0.02

This variable will contain the iteration statistics.

stats = {}

Start the iterations#

Each for loop is a complete two-way iteration. The code is thoroughly commented. Please read the inline comments carefully for a full understanding.

for cp_iter in range(1, max_iter + 1):
    stats[cp_iter] = {}


    # Step 1: Solve the Hfss design
    # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    # Solve the Hfss design.
    hfss.analyze()


    # Step 2: Refresh the dynamic link and solve the Circuit design
    # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    # Find the HFSS subcomponent in Circuit.
    # This information is required by refresh_dynamic_link and push_excitations methods.
    hfss_component_name = ""
    hfss_instance_name = ""
    for component in circuit.modeler.components.components.values():
        if (
            component.model_name is not None
            and hfss.design_name in component.model_name
        ):
            hfss_component_name = component.model_name
            hfss_instance_name = component.refdes
            break
    if not hfss_component_name or not hfss_instance_name:
        raise "Hfss component not found in Circuit design"

    # Refresh the dynamic link.
    circuit.modeler.schematic.refresh_dynamic_link(name=hfss_component_name)

    # Solve the Circuit design.
    circuit.analyze()


    # Step 3: Push excitations (HFSS design results are scaled automatically)
    # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    # Push excitations.
    circuit.push_excitations(instance=hfss_instance_name)


    # Step 4: Extract the resistor's power loss value from the Circuit design
    # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    # Evaluate power loss on resistor.
    r_losses = circuit.post.get_solution_data(expressions="0.5*mag(I(I1)*V(V1))").data_magnitude()[0]

    # Save the losses in the stats.
    stats[cp_iter]["losses"] = r_losses


    # Step 5: Set the resistor's power loss value in the Icepak design (block thermal condition)
    # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    # Find the Solid Block boundary in Icepak.
    boundaries = icepak.boundaries
    boundary = None
    for bb in boundaries:
        if bb.name == "Block1":
            boundary = bb
            break
    if not boundary:
        raise "Block boundary not defined in Icepak design."

    # Set the resistor's power loss in the Block Boundary.
    boundary.props["Total Power"] = str(r_losses) + "W"


    # Step 6: Solve the Icepak design
    # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    # Clear linked data, otherwise Icepak continues to run simulation with the initial losses.
    icepak.clear_linked_data()

    # Solve the Icepak design.
    icepak.analyze()


    # Step 7: Export the temperature map from the Icepak and create a new 3D dataset with it
    # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    # Export the temperature map into a file.
    fld_filename = os.path.join(
        icepak.working_directory, f"temperature_map_{cp_iter}.fld"
    )
    icepak.post.export_field_file(
        quantity="Temp", output_file=fld_filename, assignment="AllObjects", objects_type="Vol"
    )

    # Convert the fld file format into a dataset tab file compatible with dataset import.
    # The existing header lines must be removed and replaced with a single header line
    # containing the value unit.
    with open(fld_filename, "r") as f:
        lines = f.readlines()

    _ = lines.pop(0)
    _ = lines.pop(0)
    lines.insert(0, '"X"    "Y" "Z" "cel"\n')

    basename, _ = os.path.splitext(fld_filename)
    tab_filename = basename + "_dataset.tab"

    with open(tab_filename, "w") as f:
        f.writelines(lines)

    # Import the 3D dataset.
    dataset_name = f"temp_map_step_{cp_iter}"
    hfss.import_dataset3d(
        input_file=tab_filename, name=dataset_name, is_project_dataset=True
    )


    # Step 8: Update material properties in the HFSS design based on the new dataset
    # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    # Set the new conductivity value.
    new_material.conductivity.value = (
        f"{old_conductivity}*Pwl($TempDepCond,clp(${dataset_name},X,Y,Z))"
    )

    # Switch off the thermal modifier of the material, if any.
    new_material.conductivity.thermalmodifier = None


    # Step 9: Extract the average temperature of the resistor from the Icepak design
    # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    # Get the mean temp value on the high resistivity object.
    mean_temp = icepak.post.get_scalar_field_value(
        quantity="Temp", scalar_function="Mean", object_name=resistor_body_name
    )

    # Save the temperature in the iteration stats.
    stats[cp_iter]["temp"] = mean_temp


    # Step 10: Update the resistance value in the Circuit design
    # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    # Set this temperature in circuit in variable TempE.
    circuit["TempE"] = f"{mean_temp}cel"

    # Save the project
    circuit.save_project()


    # Check the convergence of the iteration
    # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    # Evaluate the relative residuals on temperature and losses.
    # If the residuals are smaller than the threshold, set the convergence flag to `True`.
    # Residuals are calculated starting from the second iteration.
    converged = False
    stats[cp_iter]["converged"] = converged
    if cp_iter > 1:
        delta_temp = abs(stats[cp_iter]["temp"] - stats[cp_iter-1]["temp"]) / abs(stats[cp_iter-1]["temp"])
        delta_losses = abs(stats[cp_iter]["losses"] - stats[cp_iter-1]["losses"]) / abs(stats[cp_iter-1]["losses"])
        if delta_temp <= temp_residual_limit and delta_losses <= loss_residual_limit:
            converged = True
            stats[cp_iter]["converged"] = converged
    else:
        delta_temp = None
        delta_losses = None

    # Save the relative residuals in the iteration stats.
    stats[cp_iter]["delta_temp"] = delta_temp
    stats[cp_iter]["delta_losses"] = delta_losses

    # Exit from the loop if the convergence is reached.
    if converged:
        break