Multiphysics simulation for when one physics drives another
Multiphysics simulation links fluid flow, heat, structure and electromagnetics in one study, so the effect of each on the others is predicted rather than estimated. We will also tell you when separate studies give the same answer for less effort.
Real products rarely respond to one kind of load at a time. A motor’s losses heat its windings, the hotter copper has a higher resistance and the losses rise again. A valve body carrying hot fluid expands unevenly, and the stress follows the temperature. A duct deflects under the pressure of the air inside it. Each is a coupled problem: the output of one physics is an input to another.
More about multiphysics simulation
Multiphysics simulation carries that link through the analysis. In a one-way coupling, results pass in one direction, such as temperatures from a thermal model applied as loads in a stress model. That is enough when the second physics barely affects the first. When it does, the feedback has to be part of the study. Copper resistance, for example, rises by about 0.4 % per kelvin, so a winding at 100 °C has roughly 30 % more resistance than at 20 °C, and losses calculated at room temperature will understate the heat.
Coupling is not always worth it. A coupled model takes more work to build and more computing to solve, and many problems are well served by separate studies with sensible boundary conditions. Part of our job is to say which case you are in. SixGate™ is the design-to-manufacture process of F-CAD®, and the engineers who set up a coupled study also turn its result into changes to the product.
Typical problems: electric motors and transformers, valves and manifolds carrying hot fluid, power electronics on cooled baseplates, heated tooling, exhaust and process ducting, and busbars and cable joints.
What a coupled study answers
Does temperature change the stress enough to matter?
Expansion stress and distortion from the real temperature field, rather than from a uniform temperature that hides the gradients.
How hot does the fluid make the structure?
Wall temperatures from conjugate heat transfer, which then set material strength, expansion and the temperature of nearby seals and electronics.
Can the structure take the flow loads?
Pressure from the CFD applied to walls, vanes and panels in the structural model, so stress and deflection follow the pressure distribution the flow actually produces.
What temperature do the losses produce?
Winding, core and magnet temperatures driven by electromagnetic losses, and whether they stay inside insulation and magnet temperature limits.
Will running clearances survive operating temperature?
Air gaps, tip clearances and sliding fits at temperature and under load, where uneven expansion and deflection can close a gap that looked generous when cold.
Do we need a coupled study at all?
A clear answer on whether separate studies are enough, based on how strongly each physics affects the others in your product.
Couplings we model
Thermal-structural
Temperature fields applied to structural models for expansion, thermal stress, distortion, and changes in clearance or bolt preload.
Fluid-thermal
Conjugate heat transfer, solving flow and conduction in the solids together so that wall and fluid temperatures are consistent.
Fluid-structure
Loads from the flow study, pressure and temperature, applied to the structural model for stress and deflection, where the deflection is too small to alter the flow.
Electromagnetic-thermal
Copper, eddy current and core losses carried across as heat sources for a thermal study, giving temperatures in windings, cores and magnets.
Electromagnetic to thermal to structural
Losses become temperatures, and temperatures become stress and distortion in a motor housing, transformer tank or mounting frame.
Temperature-dependent materials
Stiffness, expansion and conductivity that vary with temperature, so each part is assessed with the properties it has when hot or cold.
Before you start
What to send us
- CAD of every part involved, including fluid paths or coil and core geometry
- Loads for each physics: flow rates, heat loads, currents, pressures and fixings
- Materials, with properties over the operating temperature range
- Operating conditions and duty cycle, including start-up and shutdown
- The limits that matter: stress, temperature, deflection or clearance
- Any test or field data that shows where the problem appears
When are separate studies enough?
When one physics affects the other only weakly or in one direction. A bracket near a warm motor can often be assessed at a fixed temperature, and a stiff housing under steady flow can take CFD pressures as a one-way load.
What is the difference between one-way and two-way coupling?
One-way coupling passes results in a single direction, such as temperatures into a stress model. Two-way coupling lets each physics update the other, which matters when deflection changes a flow passage or temperature changes a material property enough to alter the first result.
Does a coupled study take more work?
Yes, more set-up and more computing than a single study, which is why it belongs where the coupling changes the answer.
Can we start with one physics and add coupling later?
Yes. A thermal or CFD study often comes first, and the loads from that study are then applied to the structural model once the critical operating conditions are known.
Which physics should be solved first?
The one that creates the load. Losses come before temperatures, temperatures before thermal stress, and flow before the pressure it puts on a structure.
Tell us which physics are involved
