CENOS Python package

This is a simple package that allows to interface with the cenos backend via a python API.

Quick-start

To use this package you will need:

• Python (> 3.10)
• CENOS simulation software installed.

Once you have those, you can install this package:

pip install cenos-py

How it works

While you use CENOS normally using the GUI, most of your actions are written out to a python_trace_script.py file inside of the case itself (File -> Show case in file explorer). It will look something like this:

import cenos_py

case = cenos_py.CenosCaseIH()

case = CenosCase("INDUCTION")
case.set_pre_processor('geomWizard', load_recent=True)
case.update_template_shape_type(1, 'workpieceTube')
case.update_template_property(1, 'diameter3', 32)
case.update_template_property(1, 'diameter4', 10)
case.update_tab_property('simulation', 'tend', 3)
case.update_tab_property('simulation', 'isAdaptive', True)
case.assign_group_material('workpiece', 'alloy_34Cr4')
case.update_physics_property('physicsThermal', 'boundary', 'workpiece_surface', 'hr', 0.9)
case.entered_meshing_window()
case.set_meshing_global_density('rough')
case.set_meshing_grading_surface(0.29)
case.set_meshing_domain_element_size('workpiece', 2)
case.set_meshing_domain_skin_layer('workpiece', True, 4.6, 10, 1.4)
case.generate_mesh()
case.calculate()

This trace file is also the best reference for the API: if you're not sure which function to call to change something, make that change once in the GUI and check what got added to the trace file.

How to use it

Let's look at two ways to run a case with different parameters.

Option 1: Build the whole case from scratch

The most obvious approach is to take the trace file and turn the parameter you care about into a variable, then repeat the whole thing in a loop:

import cenos_py

case = cenos_py.CenosCaseIH()

for i in [20, 30, 40, 50]:
    case.set_pre_processor('geomWizard', load_recent=True)
    case.update_template_shape_type(1, 'workpieceTube')
    case.update_template_property(1, 'diameter3', i)  # <-- parameter goes here
    case.update_template_property(1, 'diameter4', 10)
    case.update_tab_property('simulation', 'tend', 3)
    case.update_tab_property('simulation', 'isAdaptive', True)
    case.assign_group_material('workpiece', 'alloy_34Cr4')
    case.update_physics_property('physicsThermal', 'boundary', 'workpiece_surface', 'hr', 0.9)
    case.entered_meshing_window()
    case.set_meshing_global_density('rough')
    case.set_meshing_grading_surface(0.29)
    case.set_meshing_domain_element_size('workpiece', 2)
    case.set_meshing_domain_skin_layer('workpiece', True, 4.6, 10, 1.4)
    case.generate_mesh()
    case.calculate()

This works, but for a large case the script gets long, and it's easy to end up repeating steps that didn't actually need to change (like generating the mesh multiple times).

Option 2 (recommended): Open an existing case and only change what you need

Instead, you can open a case you've already built in the GUI, and only touch the properties you actually want to vary:

import cenos_py

case = cenos_py.CenosCaseIH()

your_case_path = r"C:\path\to\your_case"

for i in [20, 30, 40, 50]:
    case.open(your_case_path)

    # Create a copy first, so we don't overwrite the original case
    case.save_case_as(f"C:/path/to/your_case_{i}")

    case.update_template_property(1, "workpiece_height", i)
    case.calculate()

This is the recommended approach: shorter scripts, and no wasted steps.

It also makes the base case easy to keep up to date. If you want to make a static, non-parametric change — a different material, a physics property, swapped-out geometry — just open your_case_path in the GUI, make the change, save it, and re-run your script. There's no need to touch the Python at all; every run will pick up the new base case automatically.

Result analysis

Actions performed in the CENOS GUI result viewer do not get written to the python trace script. Instead you can access all the results via case.results., for example:

case = cenos_py.CenosCaseIH()

case.open(r"C:\path\to\your_case")

case.results.get_active_power()
case.results.get_average_temperature("Solid1")

Each of the functions will have docstrings and argument and return type annotations, meaning if you start typing case.results.get_dielec then you should see the "parameter hints":

def get_dielectric_losses(
    entity: str = "Total",
    port: int = 1
) -> list[float]
Return the dielectric losses in the entity, in watts (W).