A Student Team Cut Their Car's Drag by 71%, and Jumped Four Places in the Race

Last year, a student team’s car finished 11th. This year, the same team, with a redesigned car, finished 7th, and went more than twice as far on the same amount of energy.

The difference wasn’t a bigger engine or a lighter frame. It was air.

“It allowed a handful of students to complete seven full design iterations… a programme that would otherwise have required far more hardware, time, and people than a student team has available.”, Michail Angelos Gkouvelis, Mechanical Lead, Prometheus Eco Racing (NTUA)

The race that rewards efficiency, not speed

Shell Eco-Marathon doesn’t crown the fastest car. It crowns the one that goes farthest on the least energy, like a fuel-economy contest, engineered to the extreme. In that race, every bit of drag pushing back against the car is energy the team doesn’t get to use.

Prometheus Eco Racing, a student team from the National Technical University of Athens, competes in the “Urban Concept” class: a small, real-world-shaped car with doors, lights, and a windshield, not a stripped-down speed pod. That’s a much harder shape to make slip cleanly through the air.

A problem the team could feel, but couldn’t see

Their previous car, PYRFOROS III, had a drag problem, measured by a single number engineers call the drag coefficient, or Cd, that says how much the car fights the air moving around it. Lower is better. The team knew their number was too high. What they didn’t have was a fast way to find out where the drag was coming from, or what design change would fix it.

Their old simulation setup, the software used to predict how air flows around a shape before you ever build it, was slow. Every question took too long to answer, and a one-season competition build doesn’t wait around.

Final result: 11th place, at 80 km per kilowatt-hour.

Turning a season of guesswork into a season of answers

For the next car, PYRFOROS IV, the team switched their entire aerodynamics workflow to AeroCloud, running in the cloud instead of on their own limited hardware. That one change didn’t just make things faster, it changed how they worked.

Instead of committing to a shape and hoping, they could test an idea, see the result, and try again. Across one season, they ran seven full redesign rounds on the body, the canopy, and the underside of the car. When a design was worth taking seriously, they ran a highly detailed simulation to confirm it, precise enough to model airflow around the wheels and underbody, not just the car’s rough outline.

And it wasn’t just a number back. AeroCloud showed where on the car the air was struggling, like a heat map for wasted energy. That’s the part that actually changes decisions: not just knowing you have a drag problem, but seeing exactly which part of the car is causing it (what CFD engineers call force decomposition, breaking total drag down piece by piece).

Here’s the twist: the team had already faired and sealed the wheels from the very start of the project, before a single CFD run. When force decomposition came back, it showed that even after that upfront work, the wheels and wheel wells still accounted for roughly 40% of the car’s total drag, the single largest contributor on the entire vehicle. That’s not a discovery that triggered a fix. It’s proof of how much drag a shape most teams assume they’ve already handled can still be hiding.

The numbers, and what they meant

That’s a 71% cut in drag, and a 2.5x jump in how far the car could go on the same energy. Once you account for the car’s actual size, what engineers call CdA, the real-world improvement was even bigger.

PYRFOROS IV also won 1st place in the event’s Vehicle Design Award, a category judged not on speed, but on whether the team could explain and justify their engineering decisions. Having simulation data behind every design choice, instead of a guess, was central to that case:

“AeroCloud was a decisive factor in that timeline: it allowed a handful of students to complete seven full design iterations, refined in detail where the physics demanded it, a programme that would otherwise have required far more hardware, time, and people than a student team has available.”, Michail Angelos Gkouvelis, Mechanical Lead, Prometheus Eco Racing (NTUA)

Why this matters even if you’re not building a race car

This wasn’t a car company with a wind tunnel and a dedicated CFD department. It was a small student team with one season and a deadline that couldn’t move. If that’s enough to find a 71% improvement, the limit was never the team’s skill, it was how fast they could test an idea and get a real answer.

If you’re currently guessing at where your own design loses efficiency, instead of seeing it, that’s the same problem PYRFOROS III had.

	PYRFOROS III	PYRFOROS IV
Drag coefficient (Cd), lower is better	0.299	0.111
CdA (frontal-area-corrected)	0.343	0.098
Race result	11th place	7th place
Efficiency	80 km/kWh	202 km/kWh