How the new Tesla Roadster fans and thrusters work
Original technical analysis: suction load, thrust and operating limits · Editorial methods
The October 2026 Roadster combines aerodynamic surfaces, suction fans and compressed-air thrusters. They produce forces in different ways. Wings use airflow over the car, fans create a pressure difference near a surface, and gas jets provide propulsion or rotation. Keeping those mechanisms separate explains far more than calling the car “flying.”
Three systems and three jobs
Section titled “Three systems and three jobs”| System | Where the force comes from | Reported role on the Roadster |
|---|---|---|
| Wings and other aero surfaces | Air flowing around shaped surfaces | Downforce, with movable elements in Track Mode |
| Suction fans | Lower pressure beneath the car relative to surrounding air | Additional force toward a nearby surface |
| Cold-gas thrusters | Reaction to gas expelled through nozzles | Assisted launches; broader maneuvering with Rocket hardware |
This is a general explanation of the mechanisms, not a reconstruction of Tesla’s undisclosed engineering. The conceptual diagram below illustrates force directions; its shapes and arrows are not measured components or load vectors.
Active wings: changing the aerodynamic surface
Section titled “Active wings: changing the aerodynamic surface”Tesla’s October 16 Track Mode post says the wing actuation uses the Tesla Optimus actuator system. That identifies an actuator attribution, not the wing’s measured load, deployment speed or control logic. The Verge separately reports a retractable rear wing on the displayed car.
The post calls the package the best passive downforce package of any road car, “approaching F1 levels.” Tesla provides no comparable speed, vehicle setup or force measurement in that post. It is therefore a manufacturer comparison, not a measured ranking. Movable surfaces and the term “passive downforce” are also worth separating: an actuator can change a surface’s position while the resulting aerodynamic load still comes from the airflow.
The cover is labelled Road Mode in Tesla’s original media. It does not show the full Track Mode deployment and should not be used to infer its final geometry.
Suction fans: what “2g” does and does not mean
Section titled “Suction fans: what “2g” does and does not mean”A suction system lowers pressure under the vehicle. The surrounding higher pressure then pushes it toward the surface. Unlike a wing relying on forward airflow, powered suction can generate load at low vehicle speed. Its effectiveness still depends on the airflow path, underbody clearance and how well the system maintains the pressure difference.
The Verge reports Tesla’s claim of up to 2g of suction force. If that phrase means a suction force twice the vehicle’s weight, the force would be 2 × mass × standard gravity. On level ground, under a simple static model with no other aerodynamic forces, it would add to the vehicle’s weight to give a tire normal load of roughly three times that weight. This is a conditional interpretation of the wording, not a measured Tesla load case.
It does not automatically mean 2g lateral acceleration. Cornering also depends on tire behavior and road conditions, and tire grip does not scale perfectly with added normal load. The reported inverted-ramp demonstration illustrates why suction matters, but it does not establish grip over potholes, gaps or rough public roads. A fan load map and an operating envelope would answer those questions more directly than a stunt clip.
A force budget for the suction claim
Section titled “A force budget for the suction claim”The conditional interpretation above can be checked without knowing the car’s mass. Let W be its weight. If fan suction adds 2W downward on level ground, the tires support W + 2W = 3W, before other aerodynamic or dynamic loads are considered. The force described as suction is the additional 2W; the tire load in this simplified case is 3W.
This also shows why the figure cannot establish an upside-down safety margin. In an ideal static inverted position, a 2W attraction toward the surface would oppose a weight W pulling away, leaving a contact load W. That model assumes the suction claim remains valid in that position, with no leakage, transient motion or system failure. It says nothing about the car’s actual reserve capacity or safe operation.
Pressure difference and effective sealed area determine the suction force: force = pressure difference × area. A usable load map would therefore need ground clearance and surface conditions alongside the headline force. With those inputs missing, the sensible engineering question is how much load survives a changing surface, rather than whether a quoted g figure alone makes every maneuver possible.
Cold-gas thrust: stored air becomes propulsion
Section titled “Cold-gas thrust: stored air becomes propulsion”Cold-gas thrusters expel compressed gas without combustion at the nozzle. A rear-facing jet applies a forward reaction force to the car; jets positioned elsewhere can apply vertical force or a turning moment. That is the mechanical distinction between Thrust’s rear assistance and Rocket’s broader maneuvering system.
The Verge’s event report gives two high-pressure tanks with 600 liters combined volume, approximately 250 kilograms of air, and ten nozzles for the demonstrated setup. Those are reported event figures, not standard equipment for Performance. Tank volume alone does not tell us how long full thrust lasts: that also needs pressure, temperature, mass flow and nozzle-performance data.
The report says the car can replenish its air supply. It does not establish refill time, energy use or how many full-power launches a fill supports. Stored air is finite even when an onboard system can replace it later.
NASA Glenn’s thrust equation explains why nozzle performance requires mass flow, exit velocity and a pressure term. For a simplified constant-flow discharge, duration = usable gas mass ÷ mass flow rate. A tank-volume figure supplies neither input to that duration calculation. This is a general propulsion reference, not NASA validation of Tesla’s system.
Why suction is not hovering
Section titled “Why suction is not hovering”The fan mechanism described here relies on a pressure difference near a surface. It should not be credited with free-air hovering. Road & Track attributes the jumps and rotations to the gas boosters. Brief airborne maneuvers do not supply endurance, controllability or an approved flight envelope for a customer vehicle.
For which version receives each system, see Performance, Thrust and Rocket. For what an assisted launch time actually establishes, see the acceleration analysis.
Tesla’s Track Mode post and the two event articles were rechecked October 16, 2026. The physical explanations are this publication’s analysis; no pressure, aerodynamic or thrust measurements were performed here.