The problem
Traction motors are limited by winding and magnet temperatures, especially in continuous operation. A housing jacket removes most of the stator loss, but its channel shape decides cooling uniformity, hot spots and pump power. Designers also have to keep the jacket castable or machinable. This project builds a method to choose and justify a jacket design.
Objectives
- Estimate stator losses for an example motor and set temperature and pressure-drop targets.
- Model a motor housing with interchangeable jacket layouts that can be cast or machined.
- Run conjugate heat transfer studies with a simplified stator and coolant flow sweep.
- Compare layouts on temperature, uniformity, pressure drop and manufacturability.
What you need
Tools and toolboxes
- SOLIDWORKS Part and Assembly — Housing with jacket channels built from sweeps along helical and circumferential paths, inner sleeve, stator ring and end covers.
- Helix and Spiral curve, Sweep features — Spiral jacket channels with pitch as a parameter.
- SOLIDWORKS Flow Simulation — Internal analysis with Heat Conduction in Solids, user-defined coolant, volume heat sources on stator regions and a parametric flow study.
- SOLIDWORKS Mold Tools (Draft Analysis) — Checking draft and core feasibility for a cast housing.
- Excel — Loss estimate, Dittus-Boelter or Gnielinski heat transfer check and pressure-drop calculation.
Software
- SOLIDWORKS 2023 or later
- SOLIDWORKS Flow Simulation
- Excel
Extra parts and data
- Motor loss data or efficiency map — Copper and iron losses at the continuous rating set the heat sources; without them the study rests on assumptions that must be stated.
Key concepts
- Stator losses
- Copper loss in the windings and iron loss in the laminations make most of the heat in a permanent-magnet motor at continuous load. They can be estimated from efficiency at the operating point.
- Spiral versus axial channels
- A spiral (helical) channel gives one long path with good velocity; axial channels with end turns give shorter paths and lower pressure drop but risk uneven flow.
- Gnielinski correlation
- A correlation for the Nusselt number in turbulent pipe flow, used to estimate the convective coefficient in the jacket for a hand check of the CFD.
- Thermal path to the winding
- Heat travels from the winding through insulation, laminations and the housing interface to the coolant; the jacket only controls the last step, so winding temperature also depends on the stator design.
Step-by-step guide
- 1
Define the motor and losses
Use an example 30 kW continuous PMSM with stator outer diameter 200 mm and core length 120 mm. From an example efficiency at the continuous point, estimate total loss and split it into copper and iron loss with a stated assumption. Choose a 50/50 glycol-water coolant at 65 °C inlet, typical of a shared loop, and a flow range of 4 to 12 litres per minute. Set example targets: housing inner sleeve under 90 °C and a pressure drop limit.
Checkpoint: A loss table and requirements sheet with coolant, flow range and numeric targets, all labelled as example values.
- 2
Model the housing and stator
Model the housing as an outer shell and inner sleeve with the jacket between them. Model the stator as a ring with a slotted region, treating the winding zone as a solid with effective conductivity for copper and insulation, and the laminations with radial and axial conductivities if you have them. Press the stator into the sleeve with a contact resistance interface. Keep rotor, shaft and bearings out of the first study and state that rotor losses are ignored.
Checkpoint: A rebuildable assembly where the jacket channel can be swapped without changing the stator or housing outside.
- 3
Create jacket layouts
Create a spiral jacket by sweeping a rectangular profile along a Helix with pitch and number of turns as parameters. Create an axial layout with straight channels joined by end turns, and a circumferential layout with parallel rings fed by inlet and outlet manifolds. Keep channel cross-section area and wetted area roughly equal. Store each as a configuration and close inlets and outlets with lids.
Checkpoint: Three configurations with similar wetted area and Check Geometry reporting one fluid volume each.
- 4
Set up the conjugate study
Run the Wizard as Internal with Heat Conduction in Solids and gravity optional. Define the coolant as a user-defined liquid in the Engineering Database from reliable property data. Assign an aluminium casting alloy to the housing, electrical steel to laminations and an effective material to windings. Apply Volume Sources to the winding and lamination regions. Set an inlet volume flow and an environment pressure outlet. Give the outer surface a modest heat transfer coefficient to ambient.
Checkpoint: A first run converges and the coolant outlet temperature rise matches the heat removed within a few per cent.
- 5
Goals, mesh and convergence
Add Volume Goals for maximum winding and sleeve temperature, Surface Goals for inlet and outlet pressure and Point Goals at several angular positions on the sleeve to measure circumferential variation. Use a Local Mesh in the jacket with at least four cells across channel height. Run three mesh levels for the spiral case and confirm maximum winding temperature changes under 1 °C and pressure drop under 5 per cent.
Checkpoint: A mesh study table meeting both convergence criteria.
- 6
Hand-check heat transfer and pressure drop
Compute hydraulic diameter, Reynolds number and the Gnielinski Nusselt number for the spiral channel, then a convective coefficient. Estimate sleeve temperature from loss divided by coefficient times wetted area plus mean coolant temperature. Compute Darcy-Weisbach pressure drop with a friction factor and allow for curvature effects. Compare both with CFD; explain differences such as secondary flow in curved channels, which increases heat transfer.
Checkpoint: Hand estimates within about 25 per cent of CFD, with the differences explained.
- 7
Run the flow sweep for all layouts
Use a Parametric Study with inlet flow from 4 to 12 litres per minute for each configuration. Collect maximum winding temperature, sleeve temperature variation around the circumference, pressure drop and pumping power. Plot winding temperature against pumping power. Look for maldistribution in the circumferential layout with surface goals on each ring, and tune manifold sizes if one ring is starved.
Checkpoint: Plots that rank the layouts on temperature for equal pumping power.
- 8
Check manufacturability
For a cast housing, run Draft Analysis on the jacket core surfaces and check that a sand core can be supported and removed. For a two-piece machined design, check that channels can be milled on the sleeve and sealed by a shrink-fit or O-ring outer shell. Note minimum wall thicknesses from casting guidelines and check them with Thickness Analysis if available, or with section views. Reject layouts that cannot be made by your chosen process.
Checkpoint: Each layout marked as castable, machinable or not feasible, with the reason.
- 9
Iterate the chosen design
Take the best feasible layout and refine one or two parameters, such as spiral pitch or channel depth, in a short parametric run. Confirm the refined design still meets the pressure-drop target. Check the effect of a lower flow representing a degraded pump, and of a higher inlet temperature representing a hot day, to show margin. Record each run in the same results table so the margins can be read side by side.
Checkpoint: A refined design that meets temperature and pressure targets with stated margin at off-nominal conditions.
- 10
Drawings and report
Make a drawing of the housing with jacket section views, seal grooves, port threads and critical tolerances such as the stator bore. Write up loss assumptions, materials, set-up, convergence, hand checks, comparison, manufacturability and the final design. State limitations: steady state, no rotor or end-winding detail, assumed losses. Recommend a test with resistance-based winding temperature measurement at the continuous rating.
Checkpoint: A drawing and report with a justified jacket layout and a validation test plan.
Testing & validation
- Mesh convergence on winding temperature within 1 °C and pressure drop within 5 per cent.
- Coolant energy balance closes within a few per cent.
- Gnielinski and Darcy-Weisbach hand checks agree within about 25 per cent.
- Manufacturability checks confirm the chosen layout can be made.
Expected results
A motor housing with three jacket layouts, conjugate heat transfer results across a flow sweep, a manufacturability assessment, and a refined jacket design that meets temperature and pressure-drop targets.
Research angle for your M.Tech thesis
A study can compare jacket topologies at equal pumping power and wetted area, quantify circumferential non-uniformity with a defined index, and test how well standard correlations predict curved-channel heat transfer against CFD. Adding a thermal network model calibrated from the CFD and comparing it with measured winding temperatures on a test motor would make the work more rigorous and publishable.
Take it further
- Add end-winding detail and an oil spray or potting comparison.
- Couple loss data from an electromagnetic tool for each operating point.
- Run a transient peak-power event to find the time to reach the winding limit.
- Compare a cast and a two-piece machined housing on cost with SOLIDWORKS Costing.
Viva and review questions
Why compare layouts at equal pumping power?
Any layout can look better with more flow. Comparing at equal pumping power shows which geometry removes heat most efficiently.
Why does a curved channel transfer more heat than a straight one?
Curvature creates secondary flows that mix the coolant and thin the boundary layer, increasing the heat transfer coefficient, at the cost of some extra pressure drop.
Does the jacket set the winding temperature on its own?
No. Heat must cross the insulation, laminations and the stator-to-housing interface, so these resistances also matter; the jacket controls only the last step.
What would you measure on a test?
Coolant inlet and outlet temperature and flow for the heat removed, sleeve temperatures with thermocouples, and winding temperature by resistance change at the continuous rating.
Learn it at DIYguru
These DIYguru courses and programmes teach what this project needs.