ParaPowerPython

Python adaptation of the (open-source) thermal solver ParaPower (originally written in MATLAB) for rapid evaluation of 3D power module thermal behavior using a lumped thermal resistance network.

Overview

ParaPowerPython builds a discretized 3D representation (features: substrates, layers, components, bond wires, etc.) and computes temperature distribution for (currently) a single static thermal step. Internally it:

1. Loads a user JSON description (geometry, materials, boundary conditions, parameters).
2. Executes a static thermal solution (main(..., mode="thermal", type_="static")).
3. Returns either global maximum temperatures or per‑feature maxima.

Installation

Requirements:

• Python ≥ 3.8
• numpy

Clone or copy this repository and ensure materials.npy stays in the root directory.

Input JSON Structure

Top‑level keys expected by main():

{
    "ExternalConditions": { /* convection & ambient temps */ },
    "Features": [ /* geometric feature list */ ],
    "Params": { /* simulation parameters */ }
}

1. ExternalConditions

Boundary convection coefficients (h) and ambient temperatures (Ta) are face‑labeled. Also include process / internal reference temperature Tproc if applicable.

JSON Key	Meaning	Example
h_Left	Convection coeff at -X face	0
h_Right	Convection coeff at +X face	0
h_Front	Convection coeff at +Y face	0
h_Back	Convection coeff at -Y face	0
h_Bottom	Convection coeff at -Z face	150.0
h_Top	Convection coeff at +Z face	0
Ta_Left	Ambient temp at -X face (°C)	27.0
Ta_Right	Ambient temp at +X face (°C)	27.0
Ta_Front	Ambient temp at +Y face (°C)	27.0
Ta_Back	Ambient temp at -Y face (°C)	27.0
Ta_Bottom	Ambient temp at -Z face (°C)	27.0
Ta_Top	Ambient temp at +Z face (°C)	27.0
Tproc	Internal process / junction temp	230.0

Example:

"ExternalConditions": {
    "h_Left": 0,
    "h_Right": 0,
    "h_Front": 0,
    "h_Back": 0,
    "h_Bottom": 150.0,
    "h_Top": 0,
    "Ta_Left": 27.0,
    "Ta_Right": 27.0,
    "Ta_Front": 27.0,
    "Ta_Back": 27.0,
    "Ta_Bottom": 27.0,
    "Ta_Top": 27.0,
    "Tproc": 230.0
}

2. Params

General simulation parameters (current implementation uses a single static step):

Key	Meaning	Example
Tsteps	(Reserved) list of time steps	[]
DeltaT	Time increment between steps (s)	1
Tinit	Initial temperature (°C)	27.0

Example:

"Params": {
    "Tsteps": [],
    "DeltaT": 1,
    "Tinit": 27.0
}

3. Features

Each feature describes a rectangular prism region (layer/component/etc.). Coordinates are given in micrometers (µm) and converted internally to meters. Mesh resolution is controlled by dx, dy, dz (number of divisions per axis segment).

Key	Meaning
name	Feature descriptor (will map to Desc)
x	[x_start, x_end] in µm
y	[y_start, y_end] in µm
z	[z_start, z_end] in µm
dx	Divisions along X (larger => finer)
dy	Divisions along Y
dz	Divisions along Z
Q	Volumetric heat generation term (units TBD)
Matl	Material name (mapped to internal library)

Example feature list (simplified):

"Features": [
    {
        "name": "base_plate",
        "x": [0, 10000],
        "y": [0, 10000],
        "z": [0, 500],
        "dx": 2, "dy": 2, "dz": 2,
        "Q": 0,
        "Matl": "Copper"
    },
    {
        "name": "mosfet_A",
        "x": [2000, 4000],
        "y": [2000, 4000],
        "z": [500, 800],
        "dx": 2, "dy": 2, "dz": 2,
        "Q": 10,
        "Matl": "SiC"
    }
]

Material names from external sources are auto‑translated (e.g. copper → Cu, Pb-Sn Solder Alloy → SAC405, Air → AIR). If an unknown name is supplied it is passed through unchanged.

Full Minimal JSON Example

{
    "ExternalConditions": {
        "h_Left": 0, "h_Right": 0, "h_Front": 0, "h_Back": 0,
        "h_Bottom": 150.0, "h_Top": 0,
        "Ta_Left": 27.0, "Ta_Right": 27.0, "Ta_Front": 27.0, "Ta_Back": 27.0,
        "Ta_Bottom": 27.0, "Ta_Top": 27.0,
        "Tproc": 230.0
    },
    "Params": { "Tsteps": [], "DeltaT": 1, "Tinit": 27.0 },
    "Features": [
        {"name": "base_plate", "x": [0, 10000], "y": [0, 10000], "z": [0, 500], "dx": 10, "dy": 10, "dz": 2, "Q": 0, "Matl": "Copper"},
        {"name": "mosfet_A", "x": [2000, 4000], "y": [2000, 4000], "z": [500, 800], "dx": 4, "dy": 4, "dz": 2, "Q": 10, "Matl": "SiC"}
    ]
}

Usage

from main import main
result = main("model.json", mode="thermal", type_="static", results_type="global")
print(result)

Arguments:

Parameter	Allowed Values	Description
json_file	Path to JSON model	Input geometry & conditions
mode	"thermal"	(Future: other physics)
type_	"static"	Static thermal step (no transient loop yet)
results_type	"global" / "individual"	Max temperature overall or per feature

Return value:

• global → array of length 2 (initial and computed max temperatures).
• individual → dict: {feature_name: [T_init, T_final], ...}.

Mesh & Performance Notes

• Larger dx/dy/dz values increase resolution but also runtime.
• All coordinates in the JSON are in micrometers (µm); they are automatically converted to meters.