Sizing Axial Pump Impeller Blades for a 25 mm Blood Pump

A 25 mm axial blood-pump concept at 4,000–6,000 rpm needs a blade design based on its actual flow, pressure rise, and geometry—not a single airfoil chosen from a handbook. The stated target is about 6 L/min and 30 mmHg; those values are a starting operating point, not proof that a small axial stage can meet it. First define the operating point and annular geometry, then calculate the flow and blade conditions along the radius.

Define the required flow and pressure rise

Use the stated design target as the first check: approximately 6 L/min at 30 mmHg, with an impeller about 25 mm in diameter and an estimated speed of 4,000–6,000 rpm. Confirm whether 30 mmHg means pressure rise across the pump and whether 6 L/min is the required flow at that pressure. A pump must satisfy both at the same operating point; quoting flow and pressure separately is not enough.

Before selecting a blade section, record the inlet and outlet conditions, fluid properties at operating conditions, rotor and casing dimensions, and the available axial length. Record whether the 25 mm dimension is the impeller tip diameter or the full pump diameter. The hub-to-tip ratio and clearances determine the open annulus and therefore the flow area. Without those dimensions, a two-dimensional velocity diagram cannot settle the blade angles.

Use the pressure-rise form of hydraulic power, P_hyd = Q × Δp, to check the scale of the duty. If the target flow is 6 L/min and 30 mmHg is the pressure rise, the hydraulic power is approximately 0.40 W. This is output power in the fluid, not motor input power; losses make required shaft input higher. Treat the estimate as a scale check, not a motor selection.

Check whether one axial stage fits the duty

The operating point is the first pass/fail question. A small diameter and high speed do not, by themselves, establish that the desired flow and pressure are achievable. Build a pump curve or use a validated design method to assess pressure rise across the required flow range. Compare that curve with the system resistance curve; their intersection is the operating point.

If a single axial stage cannot provide the needed pressure at the target flow, assess alternatives rather than forcing an airfoil choice. Multistaging may add pressure capability but consumes space and adds design complexity. Mixed-flow geometry is another concept worth evaluating when an axial-only arrangement does not fit the flow and pressure requirements. The design must resolve whether the size constraint leaves room for those options.

Ask for the actual head and flow requirements before debating blade profiles. If requirements remain estimates, label them as such and update the design when system-level measurements or specifications become available. That prevents a profile model from creating false confidence in an undefined duty.

Calculate conditions at several blade radii

Do not apply one airfoil section and one velocity diagram across the whole blade. Rotational speed produces a different tangential velocity at each radius, while chord, pitch, local axial velocity, and relative flow angle also vary. The flow regime can therefore change from hub to tip. The design notes identify laminar, transitional, and turbulent conditions along the blade, but the local Reynolds number cannot be established from diameter and rpm alone.

At each radial station, estimate the relative velocity and calculate Re = ρ W c / μ, where ρ is fluid density, W is local relative velocity, c is local chord, and μ is dynamic viscosity. Use fluid properties at the intended operating conditions. State the assumptions used for inlet swirl, axial velocity distribution, and blockage; each affects the local velocity triangle and incidence.

Use a two-dimensional velocity diagram as an initial sizing tool, not as a complete blade design. It helps estimate section angles at a selected radius, but it does not represent radial flow, hub and tip effects, or interaction with the casing. Check several stations and create a spanwise distribution of chord, twist, and section geometry.

Use NACA sections as candidates, not a finished design

The designers had been examining NACA four-series profiles because lift and drag data were available through XFOIL. That is a reasonable screening approach, but airfoil coefficients alone do not choose a pump blade. The coefficients depend on Reynolds number and incidence, while the blade operates in a rotating, confined, three-dimensional flow field. A section that looks suitable in an isolated two-dimensional calculation may not produce the required pump curve.

Evaluate candidate profiles at the calculated local Reynolds numbers and incidence range. Check whether the data cover the expected conditions; do not extrapolate coefficients outside their range without validation. Use the results to compare candidates and inform blade angles, then analyze the assembled rotor and pump geometry with a method that accounts for three-dimensional flow.

A Clark Y profile was mentioned as an example from propeller design, not as a recommended section for this pump. Do not transfer aircraft or marine propeller profiles by analogy alone. Pump-fluid viscosity, rotor confinement, scale, radial variation, and the required pressure-flow duty change the design problem.

Set hub-to-tip geometry before refining the blade

The hub-to-tip ratio affects both flow area and how much the local conditions vary along the span. A large hub can shorten the radial span and reduce the difference in flow regime from root to tip, a design consideration raised for small pumps where fluid damage is a concern. It also reduces the available annular flow area. Recalculate the flow velocity and blade loading when changing hub size; do not treat a larger hub as a free improvement.

Include three-dimensional flow in the decision because the designers themselves flagged hub-to-tip ratio as a reason to look beyond a simple two-dimensional approach. Evaluate the blade from hub to tip, including the root and tip regions, and check casing and hub clearances. A radial section schedule should reflect local velocity and loading rather than repeat a single profile unchanged.

Follow a staged design procedure

1. Confirm that the duty is 6 L/min at a 30 mmHg pressure rise, and establish the operating fluid properties and system curve.
2. Fix the geometric envelope: determine whether 25 mm is the rotor tip diameter, then specify hub diameter, axial length, and clearances.
3. Calculate flow area and local velocity triangles at multiple radii for the proposed speed range of 4,000–6,000 rpm. Document inlet-flow and swirl assumptions.
4. Calculate local Reynolds numbers and screen candidate sections, including the NACA four-series candidates, only within applicable data ranges.
5. Design a spanwise variation in section, chord, and pitch; analyze the complete three-dimensional rotor and assess whether axial, mixed-flow, or multistage geometry meets the duty within the size envelope.
6. Build and test a prototype against measured flow and pressure over the intended speed range. Revise geometry from the measured pump curve rather than tuning from profile coefficients alone.

Verify pressure, flow, and safety before accepting the design

Measure flow and pressure rise at the same time and at controlled speed. Compare the measured pump curve with the target point and system curve. Record fluid temperature and the test geometry so results can be repeated. If the design misses the point, check actual speed, inlet conditions, leakage or bypass paths, and the as-built hub, tip, and blade geometry before changing the profile.

A blood pump adds a separate acceptance problem: avoiding blood damage and demonstrating biological compatibility. The design discussion says that minimizing blood damage was not the project aim at that stage, but a pump intended for blood cannot be accepted as a clinical device on hydraulic performance alone. Treat hemocompatibility and device validation as independent engineering workstreams, not consequences of selecting a familiar airfoil.

Avoid shortcuts that conceal the real design problem

• Do not choose an airfoil from a generic rule of thumb and assume it applies across the blade radius.
• Do not use lift and drag coefficients as a substitute for a pump curve or three-dimensional analysis.
• Do not dismiss multistage or mixed-flow layouts until the size envelope and required pressure rise are quantified.
• Do not infer Reynolds number from rotational speed alone; calculate it using local relative velocity, chord, and fluid properties.
• Do not interpret hydraulic output power as motor input power or as evidence that the target operating point is feasible.

FAQ