Multi-Element Side Wings and Rear Diffuser

Bachelor's Thesis | UIRT-03 Formula Student Chassis | Jan–Jun 2026

← Back to Projects
2D | One element
2D | One element
2D | Two elements
2D | Two elements
2D | Initial three elements
2D | Initial three elements
2D | Three elements refined
2D | Three elements refined
3D | Surface pressure on the final assembly
3D | Surface pressure on the final assembly
3D | Streamlines through the side wings
3D | Streamlines through the side wings
Why

Formula Student is won in tight corners at 40–60 km/h, where downforce matters more than top speed. Our team designed its side wings by trial-and-error iterations. I wanted to leave the next generation a repeatable, numbers-based method instead.

The problem
  • The bare UIRT-03 chassis made lift, not downforce: −20.55 N at 60 km/h.
  • Could a ground-effect package alone (three-element side wings and a rear diffuser) turn that around, and could it replace a front and rear wing altogether?
#
What went wrong
What I changed
01
What went wrong

Adding a third element made the wing worse: my first three element sweep peaked at CL = 1.730, below the two-element best of 3.143.

What I changed

The contours showed flow failing to reach the downstream elements. My initial sweep had been too narrow (0-3 % overlap); widening it to 0-6 % found CL = 4.044 (+134 %) at 5 % overlap and spacing, same airfoils. Cp plots traced the gain to the main and second elements.

02
What went wrong

Fine polyhedral meshes ran past the workstation's 64 GB of RAM.

What I changed

Switched to a trimmed hex-dominant mesh aligned with the freestream, and a symmetry plane that halved the 3D cell count; spending the saving on refinement around the wings.

03
What went wrong

The two-element grid study didn't reach mesh independence: the finest mesh came in 10.3% below the chosen one, though both converged to 10−5.

What I changed

Every design in each sweep was meshed with the same settings, so comparing them is fair, but the exact CL values carry that uncertainty (about 7 % in 3D).

04
What went wrong

Lowering the side wings 10 mm for more ground effect cut the gain from 128.9 N to 116.8 N.

What I changed

Read it as separation rather than ground effect and kept the standard ride height for the final Maximum Area design.

The result

The 20° diffuser added 40.8 N of downforce over the bare chassis. The Maximum Area side wings added 128.9 N for 6.2 N more drag. On its own, it's unlikely to replace a full front-and-rear wing package at 60 km/h. Whether it's enough depends on the team's lap-time targets. Each was simulated on its own, so testing them together is the next step. The team now has a step-by-step method for multi-element side wings, from airfoil choice through slot geometry to placement on the car.

351
Simulation cases
+134 %
CL from slot geometry
+128.9 N
Side wings
+40.8 N
Rear diffuser
6.2 N
Side-wing drag penalty
Method
Solver
Siemens STAR-CCM+, steady RANS
Turbulence
SST k-ω, 2nd-order upwind
Mesh
Trimmed hex + prism layers, y+ ≈ 1
Grid independence
5 studies, 45 mesh densities
Domain
3L up · 8L down · 3L sides and top
Boundaries
Velocity inlet, rolling road, 60 km/h
Reynolds
9.8 × 105 (c = 0.86 m, V = 16.67 m/s)
Hardware
64 GB workstation, ~16 M cell ceiling