Example: Rotating Machinery (Fan / Pump)#
This example demonstrates simulating a rotating fan or pump impeller using the Multiple Reference Frame (MRF) approach. MRF is a steady-state approximation that avoids the cost of a full transient simulation.
Objective#
Generate a multi-region mesh with rotating and stationary zones
Run an MRF simulation of a fan or pump
Analyze performance (pressure rise, mass flow, efficiency)
Visualize flow patterns through the rotor
Background: MRF Approach#
The Multiple Reference Frame method:
Divides the domain into a rotating zone (containing the impeller) and a stationary zone (inlet/outlet ducts)
Solves the flow equations in a rotating reference frame within the rotating zone (adding Coriolis and centrifugal source terms)
Uses a frozen rotor interface to couple the rotating and stationary zones
This gives a steady-state approximation of the time-averaged rotating flow. It’s fast (single steady solve) but doesn’t capture blade-passing transient effects.
Note
MRF is best for design-point analysis and comparative studies. For detailed transient analysis (noise, vibration, blade-to-blade interaction), use transient sliding mesh (available on Pro tier and above).
Step 1: Create a Meshing Project#
Dashboard → New Project → Meshing
Name: “Fan Simulation”
Upload your fan/pump geometry (STEP format)
Step 2: Domain Configuration#
Select Rotating Machinery domain type
Studio creates two zones:
Rotating zone — Cylindrical region around the impeller
Stationary zone — Surrounding duct/enclosure
Rotating Zone Setup#
Parameter |
Description |
Example |
|---|---|---|
Rotation axis |
Direction of rotation |
Z-axis: [0, 0, 1] |
Rotation origin |
Center of rotation |
[0, 0, 0] |
RPM |
Rotational speed |
3000 RPM |
Zone shape |
Cylindrical region enclosing the rotor |
Radius slightly larger than blade tips |
Zone Sizing
The rotating zone should:
Fully enclose all rotating parts (blades, hub, shroud)
Have a small gap (5–10% of blade span) between the blade tips and the zone boundary
Extend axially to include the blade passages
Do not include stationary parts (duct walls, volute) in the rotating zone.
Step 3: Multi-Region Setup#
Regions#
Region |
Type |
Description |
|---|---|---|
|
Rotating (MRF) |
Contains the impeller blades |
|
Fluid (stationary) |
Upstream duct or plenum |
|
Fluid (stationary) |
Downstream duct or volute |
Interfaces#
Interface |
Between |
Type |
|---|---|---|
|
rotor ↔ inlet_duct |
Frozen rotor |
|
rotor ↔ outlet_duct |
Frozen rotor |
Step 4: Surface Naming#
Surface |
Name |
Condition |
|---|---|---|
Duct entrance |
|
Velocity inlet or total pressure |
Duct exit |
|
Pressure outlet |
Blade surfaces |
|
No-slip wall (rotating) |
Hub |
|
No-slip wall (rotating) |
Shroud / duct wall |
|
No-slip wall (stationary) |
Step 5: Mesh Settings#
Parameter |
Value |
|---|---|
Target cell size |
Blade span / 20 |
Min cell size |
Blade span / 100 |
Refinement levels |
10 |
BL layers |
8 |
BL first layer height |
Target y+ ≈ 30 for RANS |
BL growth rate |
1.2 |
Refinement Zones#
Zone |
Location |
Purpose |
|---|---|---|
Blade passage |
Between blades |
Resolve blade-to-blade flow |
Leading edges |
Blade leading edges |
Capture flow around blade entry |
Tip gap |
Between blade tips and shroud |
Resolve tip leakage flow |
Step 6: Generate Mesh#
Expect 5–20 million cells for a medium-resolution fan simulation. Check:
Quality in the rotating zone (blade passages should have uniform, fine cells)
Interface quality (cells should be similar size on both sides)
Boundary layer quality on blade surfaces
Step 7: Simulation Setup#
Setting |
Value |
|---|---|
Turbulence model |
k-ω SST |
Rotating zone RPM |
3000 RPM (match your design speed) |
Inlet velocity |
Based on operating point (e.g., 5 m/s) |
Outlet pressure |
0 Pa (or set based on system curve) |
Turbulence intensity |
3% |
Max iterations |
1500 |
Algorithm |
SIMPLE |
Step 8: Results Analysis#
Performance Metrics#
Total pressure rise — Measure total pressure at inlet and outlet:
ΔP_total = P_total_outlet - P_total_inlet
Mass flow rate — Integrate velocity across the inlet face
Efficiency — Compare actual pressure rise to ideal:
η = (ΔP_total × Q) / (τ × ω)
Where Q is volume flow rate, τ is torque on blades, ω is angular velocity
Flow Through Blade Passages#
Add a slice plane perpendicular to the rotation axis, at mid-blade height
Color by velocity magnitude
Look for:
Uniform flow through all passages (good)
Separation on blade suction side (may indicate stall)
High velocity near blade tips (tip effects)
Meridional View#
Add a slice plane through the rotation axis (meridional plane)
Color by axial velocity or pressure
Shows how the flow develops from inlet through the rotor to the outlet
Blade Loading#
Color blade surfaces by pressure
The pressure side (concave) should show higher pressure
The suction side (convex) should show lower pressure
The pressure difference drives the torque and pressure rise
Tip Flow#
Add a slice plane at the blade tip radius
Color by velocity
Check for tip leakage flow (flow from pressure to suction side through the tip gap)
This is a key source of loss in unshrouded impellers
Design Point vs. Off-Design#
To map a performance curve:
Run simulations at several flow rates (vary inlet velocity)
Record pressure rise and efficiency at each point
Plot ΔP vs. Q (fan curve) and η vs. Q (efficiency curve)
Identify the best efficiency point (BEP)
Off-Design Caution
MRF becomes less accurate far from the design point, especially in stalled or highly separated conditions. If you need off-design performance mapping with high accuracy, consider transient sliding mesh simulations.