Joule heating
Electrothermal analysis couples steady current flow to heat transfer. The resistive loss
is the volumetric heat source, and the temperature changes the
conductivity, which changes where the current flows. RunMat solves the two together, steady or over
time, from a .fea study:
| Procedure | What runs |
|---|---|
electrothermal steady {} | Current flow and steady heat transfer iterated to the two-way fixed point |
electrothermal transient { start_time; end_time; time_step } | Current flow at every time step and backward-Euler heat transfer, iterated over the whole history |
What equations are solved?
with the material's resistive_heating_coefficient (resistivity linear in temperature,
) and its electrical reference temperature.
Both fields live on the same canonical mesh, so the loss reaches the thermal solve element by element with no interpolation, and the integrated heat source equals the electrical power.
Coupling. Each element's conductivity is evaluated at its mean temperature. A sweep solves current flow with the current temperatures, then the heat transfer with the resulting Joule density. Sweeps repeat until the largest change of any element temperature is below of the temperature rise. For a transient study each sweep solves current flow at every time point (sources such as a fault-current signal are sampled at that time) and then the whole transient with the tabulated Joule source. With the first sweep is already exact (one-way coupling) and the second confirms it.
How do I author an electrothermal study?
model heated_conductor {
thermal { reference_temperature = 293.15 K }
assign material.copper to geometry.busbar.region.body
constraint ground {
electrical.potential { target = geometry.busbar.region.ground_terminal; value = 0 V }
}
load current {
electric_current on geometry.busbar.region.input_terminal = 400 A
}
load cooling {
thermal.convection {
target = geometry.busbar.region.exterior
ambient_temperature = 298.15 K
coefficient = 10 W/m^2/K
}
}
}
study heated_busbar {
geometry = geometry.busbar
mesh = mesh.busbar
model = model.heated_conductor
analysis operating_point { electrothermal steady {} }
output temperature = analysis.operating_point.field.node.temperature
output current_density = analysis.operating_point.field.element.electric_current_density
}What does a run publish?
Two cases:
<analysis>: the heat transfer at the converged operating point (temperature, temperature gradient, heat flux; the full history for a transient), with the converged Joule source bound.<analysis>/conduction: current flow at the converged temperatures (potential, electric field, current density, Joule power density), at the last time point of a transient.
Each authored output belongs to the case that produces its quantity. The run log records every sweep's temperature change and the Joule power and terminal power (), which agree.
What material and boundary data are required?
Every conductor needs conductivity, resistive_heating_coefficient and reference_temperature
(electrical), and conductivity, specific_heat and mass_density (thermal). Terminals need at
least one prescribed potential per conductor; thermal boundaries need a sink (fixed temperature,
convection or radiation). Heat crossing a joint can use a thermal contact interface; an electrical
contact resistance is modeled as a thin body with its own conductivity.
How accurate is it?
A bar with fixed end temperatures carrying a fixed current has the closed-form solution with . RunMat reproduces it to the tolerances in electromagnetic validation, and its Joule power matches integrated along the bar.
See electromagnetics, heat transfer, electromagnetic loads, thermal loads, and current status.