Back to projects

Small-Scale Wind Turbine

This project developed a two-bladed small-scale horizontal-axis wind turbine through a complete workflow: airfoil screening, BEM blade optimization, QBlade simulation, PMSG matching, prototype manufacturing, generator testing and wind tunnel validation.

Initial Airfoil Screening

The initial design phase reduced a broad airfoil database to a focused set of candidates for low-Re operation. Profiles were first filtered from AirfoilTools using a 12–16% thickness-to-chord range and a reference Reynolds number of about 200,000, then ranked through a weighted score based on aerodynamic efficiency, stall behaviour and relative thickness. The four best-ranked profiles were retained for detailed comparison, with NACA 4415 added as a reference case.

Weighted shortlist

Rank	Airfoil	Eff. score 50%	Stall score 30%	Thick. score 20%	Total
#1	FX 63-137 smoothed	9.84	10.00	9.40	9.80
#2	EPPLER 397	9.89	10.00	9.00	9.74
#3	EPPLER 396	10.00	10.00	8.20	9.64
#4	MH114	9.81	10.00	8.04	9.51
#10	NACA 4415	7.55	10.00	8.00	8.38

Aerodynamic Comparison

The shortlisted airfoils were compared through polar curves at realistic operating Reynolds numbers. This step checked whether the strongest profiles from the ranking also provided a suitable lift-to-drag ratio, delayed stall behaviour and a stable aerodynamic response before moving into BEM blade optimization.

Lift-to-drag ratio versus angle of attack for shortlisted airfoils — **Lift-to-drag comparison**CL/CD curves used to compare aerodynamic efficiency at the expected operating conditions.

Lift coefficient versus angle of attack for shortlisted airfoils — **Stall behaviour**Lift curves used to assess stall robustness and off-design tolerance.

BEM Blade Optimization

Each candidate blade was optimized with Blade Element Momentum theory. The 450 mm radius was divided into 15 radial elements and chord, twist and thickness were recalculated from an initial TSR of 6.5 until the manufacturing limits were satisfied along the span.

Minimum chord constraint: 40 mm.
Minimum thickness constraint: 5 mm.
Final MH114 blade: TSR 4.9, 15 elements plus tip section.

BEM blade geometry with radial airfoil sections — **Radial blade elements**Airfoil profiles positioned along the radius to define chord, twist and thickness distribution.

QBlade Evaluation

The optimized blade candidates were first checked against the operating Reynolds range, then compared at the representative design Reynolds number through lift-to-drag curves. Their QBlade performance was then evaluated using power coefficient and torque curves. MH114 was selected for the final turbine because it combined a strong CP curve with the highest torque peak at the operating conditions.

Operating Reynolds number distribution for shortlisted blade candidates — **Reynolds range**Operating Reynolds number checked along the blade radius.

Lift-to-drag ratio at Reynolds 150000 — **Design-Re comparison**Lift-to-drag performance compared at Re ≈ 150,000.

Power coefficient versus tip-speed ratio from QBlade — **CP vs TSR**Power coefficient curves used to compare peak performance and operating range.

Torque versus tip-speed ratio from QBlade — **Torque vs TSR**Torque peak used to guide generator matching.

PMSG Matching

The aerodynamic design was coupled with a three-phase permanent magnet synchronous generator. For the best blade profiles, two generator configurations were evaluated: two coils per phase and four coils per phase. The four-coil option predicted higher efficiency, while the two-coil layout was chosen for the physical prototype because it was significantly simpler to build and test.

MH114, 4 coils/phase 302.72 W

Predicted electrical power, efficiency 0.832.

MH114, 2 coils/phase 270.8 W

Simpler physical generator layout, efficiency 0.747.

Three-phase permanent magnet synchronous generator model — **PMSG model**Iron-less generator architecture used for sizing.

Manufacturing And Assembly

The MH114 blade was modelled in Fusion 360 from the Excel-based section definition and manufactured using FDM additive manufacturing. The generator was then built manually: six coils were wound, arranged in the stator layout and connected in a three-phase star configuration.

FDM printed wind turbine blade prototype held during assembly — **Printed blade**FDM blade manufactured from the Fusion 360 model.

Hand-wound coil for the generator — **Coil winding**Manual coil production for the PMSG prototype.

Complete hand-built generator assembly with coil connections — **Generator assembly**Six coils arranged in the final stator layout.

Generator Testing

The assembled generator was tested to evaluate the main electrical quantities and manufacturing quality. The measurements included magnetic flux density, open-circuit phase voltages, load power at different resistances and the resulting generator efficiency.

Magnetic flux density measurement on the generator — **Flux measurement**Measured peak flux density was lower than the theoretical value.

Three-phase induced voltage waveform on oscilloscope — **Three-phase waveform**No-load test used to check phase displacement and manufacturing quality.

Induced voltage versus rotational speed — **Voltage vs rpm**Induced phase voltage followed the expected linear trend.

Load power versus load resistance — **Load power**Power output measured for different loads and rotational speeds.

Generator efficiency versus load power — **Generator efficiency**Experimental efficiency evaluated from load power and generator losses.

Design Vs Manufactured Blade

During the Fusion 360 implementation, the hub-blade transition was generated with a 45 mm offset instead of the intended 15 mm offset. The manufactured blade therefore became 480 mm long rather than the optimized 450 mm design. Instead of treating this as a simple manufacturing error, the real geometry was reconstructed in QBlade and used to explain the differences between the intended design, the manufactured blade and the later experimental results.

Comparison between intended and manufactured hub-blade offset — **Hub-blade offset**15 mm intended transition compared with the 45 mm manufactured transition.

Chord comparison between 450 mm and 480 mm blades — **Chord comparison**Same section data shifted outward in the reconstructed 480 mm blade.

Twist comparison between 450 mm and 480 mm blades — **Twist comparison**Radial shift changed the local twist distribution seen by the rotor.

Power coefficient comparison between 450 mm and 480 mm blades — **CP impact**The 480 mm blade showed a slightly lower peak CP and a shift toward lower rpm.

Mechanical power comparison between 450 mm and 480 mm blades — **Power impact**The larger radius increased absolute mechanical power.

Torque comparison between 450 mm and 480 mm blades — **Torque impact**The reconstructed blade delivered higher torque over much of the range.

Slicer comparison of intended and manufactured blade root layer deposition — **Manufacturing consequence**The smoother 45 mm transition reduced local overhangs and likely improved root robustness.

Wind Tunnel Validation

The final prototype was tested in the wind tunnel. Rotational speed, voltage, current and wind speed were measured and used to reconstruct CP, mechanical power, torque and total efficiency. The experimental values were lower than the QBlade predictions but followed the same trends, showing the combined effects of 3D aerodynamics, FDM surface quality, bearing friction and generator losses.

Experimental and theoretical CP comparison versus rpm — **CP comparison**Experimental CP followed the theoretical trend at lower absolute values.

Experimental and theoretical mechanical power comparison — **Power comparison**Measured mechanical power compared with 450 mm and 480 mm predictions.

Experimental and theoretical torque comparison — **Torque comparison**Wind tunnel data used to check rotor-generator behaviour.

Experimental and theoretical total efficiency comparison — **Total efficiency**Theoretical peak efficiencies around 41-42% compared with an experimental peak near 30%.