Electric Current, Ohm's Law, and Electro-Thermal FEM
Table of Contents
1. Electricity in Modern CAE: Beyond Structural Analysis
With the rapid growth of electric vehicles, power electronics, and electrified systems, CAE engineers increasingly need to analyze electrical current flow and its thermal consequences — not just mechanical stress. Electro-thermal FEM couples electrical analysis (finding the voltage and current distribution) with thermal analysis (computing the resulting Joule heat and temperature field) to design safe, reliable electrical components.
The mathematical structure is remarkably similar to structural analysis — Ohm's law has the same form as Hooke's law, and the FEM equations for electrical conduction are structurally identical to thermal conduction. An engineer fluent in thermal FEM can quickly learn electro-thermal coupling.
EV fast charging puts 200 amps through the cable — that seems like an enormous current. Why doesn't the cable just melt?
The cable doesn't melt because the current density J = I/A is kept below the safe limit by making the conductor large enough. For 200A in a copper cable with cross-sectional area 70 mm²: J = 200/70e-6 = 2.86 MA/m². Power dissipation: P = J²ρL = 2.86e6² × 1.72e-8 × 5m ≈ 70W per 5 meters. That 70W heats the cable, but the thick insulation (rated to 90°C) and the copper's low resistivity keep the temperature rise within safe limits. FEM electro-thermal analysis verifies the temperature never exceeds the insulation limit, and also checks connector contacts where local resistance can be 10× higher than the wire — those are the hotspots that actually fail.
2. Ohm's Law: V = IR
Ohm's Law relates voltage (electric potential difference), current, and resistance:
This is the most powerful tool for quick circuit analysis. In FEM, the continuum form of Ohm's Law:
Where $\mathbf{J}$ is current density (A/m²), $\sigma_e$ is electrical conductivity (S/m), $\mathbf{E}$ is electric field (V/m), and $\phi$ is electric potential (V). Combined with current continuity ($\nabla \cdot \mathbf{J} = 0$ for steady state), this gives the governing equation for electrical FEM:
This is mathematically identical to the heat equation ($\nabla \cdot (k\nabla T) = 0$) — replace conductivity $\sigma_e$ with thermal conductivity $k$, and electric potential $\phi$ with temperature $T$. Every FEM thermal solver can solve electrical conduction problems with trivial material property substitution.
3. Joule Heating: P = I²R
When current flows through a resistor, electrical energy is dissipated as heat (Joule heating or I²R loss):
In the continuum form, the volumetric heat generation rate (W/m³) from Joule heating:
This Joule heat source term feeds directly into the thermal FEM equation — the source $\dot{q}'''$ from the electrical analysis drives the temperature rise in the thermal analysis. This is the coupling between the two physics.
| Application | Current (A) | Typical Resistance (mΩ) | Joule Heat (W) |
|---|---|---|---|
| EV fast charge cable (1m) | 200 | 0.3 | 12 |
| EV battery cell interconnect | 50 | 0.5 | 1.25 |
| PCB trace (1mm wide, 35µm thick, 10mm) | 1 | 50 | 0.05 |
| Spot weld in car body (during welding) | 10,000 | 0.2 | 20,000 |
| Bus bar in industrial panel | 1000 | 0.05 | 50 |
4. Resistivity and Current Density
Electrical resistivity $\rho_e$ (Ω·m) is a material property characterizing how strongly a material opposes current flow. Resistance of a conductor:
Current density $J = I/A$ (A/m²) must be kept below safe limits in every conductor. Maximum safe current density depends on cooling conditions — typically 1–5 A/mm² for air-cooled conductors, 5–20 A/mm² for liquid-cooled.
| Material | Resistivity ρe (×10⁻⁸ Ω·m) | Conductivity σe (MS/m) | Temp Coeff (×10⁻³/°C) |
|---|---|---|---|
| Silver | 1.59 | 62.9 | 3.8 |
| Copper (annealed) | 1.72 | 58.1 | 3.9 |
| Gold | 2.44 | 41.0 | 3.4 |
| Aluminum | 2.82 | 35.5 | 3.9 |
| Tungsten | 5.60 | 17.9 | 4.5 |
| Steel (carbon) | ~10 | ~10 | 5.0 |
| Solder (SnAgCu) | 12.3 | 8.1 | 4.0 |
Temperature Dependence of Resistivity
Resistivity increases with temperature for metals (phonon scattering): $\rho_e(T) = \rho_0[1 + \alpha_R(T-T_0)]$. This positive temperature coefficient creates thermal runaway risk: higher temperature → higher resistance → more Joule heating → even higher temperature. FEM must model this nonlinear coupling to correctly predict maximum operating temperature.
5. Skin Effect at High Frequency
At DC, current distributes uniformly over the conductor cross-section. At high frequency (AC), the electromagnetic skin effect concentrates current near the conductor surface. The skin depth:
Where $\omega = 2\pi f$ is angular frequency, $\mu$ is magnetic permeability. At 50 Hz in copper: $\delta = \sqrt{1.72 \times 10^{-8} / (\pi \times 50 \times 4\pi \times 10^{-7})} \approx 9.3$ mm. At 10 kHz: $\delta \approx 0.66$ mm. At 1 MHz: $\delta \approx 66$ µm.
For high-frequency power electronics (EV inverters, DC-DC converters operating at 10–100 kHz), skin effect significantly reduces the effective conductor cross-section and increases AC resistance. Litz wire (multiple thin strands, each thinner than skin depth) is used to recover low AC resistance.
For an EV inverter operating at 20 kHz, should I worry about skin effect in the busbars? They're 5mm thick copper bars.
Definitely worth checking. At 20 kHz in copper: δ = √(1.72e-8/(π×20000×4πe-7)) ≈ 0.47 mm. Your 5mm busbar is about 10 skin depths thick — current is concentrated in the outer ~0.47mm on each side, using only about 2×0.47/5 = 18% of the cross-section effectively. The AC resistance is roughly 5× higher than the DC resistance, meaning 5× more Joule heating than a DC analysis would predict. You absolutely need to include skin effect in your busbar thermal analysis. Use FEM with the full electromagnetic equations (Ansys Maxwell, COMSOL) or at minimum apply an AC resistance correction factor. Missing this can mean designing a busbar that overheats at operating frequency even though it passes a DC thermal test.
6. EV Fast-Charging Cable: Worked Example
Design a DC fast-charging cable capable of continuously carrying 250A at 800V (200 kW charging). Maximum conductor temperature: 90°C. Ambient temperature: 40°C. Cable length: 5m.
Step 1: Required Conductor Area
Assuming copper ($\rho_e = 1.72 \times 10^{-8}$ Ω·m) and maximum current density 3.5 A/mm² for air-cooled flexible cable:
Step 2: Joule Heat Calculation
Step 3: Temperature Rise Check (lumped)
Cable surface area: $A_s = \pi D L = \pi \times 0.022 \times 5 = 0.346$ m². With $h = 10$ W/m²·K (natural convection): $\Delta T = P/(hA_s) = 56.6/(10 \times 0.346) = 16.4$°C above ambient. At 40°C ambient: conductor reaches ~56°C — well below 90°C limit. FEM verification confirms this is conservative (internal heat must conduct through insulation), but the hand calc gives confidence.
7. FEM Electro-Thermal Coupling
Electro-thermal FEM simultaneously solves for voltage/current distribution and temperature field, with Joule heating coupling them:
The coupling is: electrical conductivity $\sigma_e$ depends on temperature (material nonlinearity), and the Joule heating term $\sigma_e|\nabla\phi|^2$ drives thermal analysis. Sequential decoupling is common: solve electrical first, extract Joule heat, input to thermal. Fully coupled iteration is needed when $\sigma_e(T)$ variation is significant.
Abaqus Electro-Thermal Setup
*MATERIAL, NAME=COPPER
*ELECTRICAL CONDUCTIVITY
5.8E7, 20. ! Conductivity S/m at 20°C
5.0E7, 100. ! Temperature-dependent
*THERMAL CONDUCTIVITY
385.,
*COUPLED THERMAL-ELECTRICAL, DELTMX=10.
1., 1000., 0.1, 10.
*CFILM (surface convection)
CABLE_OUTER, 1, 10., 40. ! h=10, T_ambient=40°C
8. EV Battery Pack: Electrical and Thermal Co-Simulation
An EV battery pack combines hundreds or thousands of cells in series/parallel configurations. The electrical and thermal behavior are tightly coupled: cell resistance increases with aging and temperature, affecting current distribution; current distribution affects local heat generation; heat affects aging rate. This creates a full electro-thermal-electrochemical coupling problem.
Key analysis objectives:
- Current distribution: Which cells/modules carry the most current due to resistance variation?
- Thermal hot spots: Which cells reach highest temperature during fast charging?
- Cell-to-cell temperature variation: Gradient > 5°C between cells accelerates differential aging
- Thermal runaway propagation: If one cell enters thermal runaway, does it propagate?
How do battery management systems (BMS) use FEM simulation? I thought BMS was purely electronics control...
FEM doesn't run in the BMS in real-time — it's used during design to create the thermal and electrical models that the BMS algorithms are based on. For example: FEM electro-thermal simulation of the pack at various charge rates gives the thermal response data. This data is used to calibrate a reduced-order model (a thermal equivalent circuit with maybe 20–50 nodes) that the BMS can run in real-time to estimate cell temperatures from a few strategically placed sensors. Without FEM, you couldn't know where to place those sensors, or trust that the simple model captures the hotspot locations correctly. FEM is the "ground truth" that validates the simple models safe for real-time deployment.
9. Cross-Topics
| Topic | Connection | Link |
|---|---|---|
| Heat & Temperature | Joule heating P=I²R is the thermal source in electro-thermal FEM | Heat and Temperature |
| Thermal Expansion | Joule heating causes thermal expansion in conductors and PCBs | Thermal Expansion |
| Wave Properties | Electromagnetic waves — skin effect is a wave propagation phenomenon | Wave Properties |
| Electromagnetics Theory | Maxwell's equations, induction, magnetic FEM, antenna analysis | Electromagnetics |
| Fluid Dynamics | Liquid cooling of EV batteries couples electrical heat generation with CFD | Fluid Dynamics |