Scope Views Home



Follow @scope_views


Can SpaceX Really Make a Tesla Fly?



Tesla has always promised a flying (!) ‘SpaceX’ version of the Roadster. Most assume it’s just an idle dream, but occasional updates suggest not, with the latest coming from Tesla’s design chief in early 2022.

So is this even possible? Could SpaceX technology really make a Tesla Roadster fly (apart from on the end of Falcon Heavy)?

No journalistic opinion is required, it’s just a basic Physics and rocketry problem. So I thought I’d whip out my digital envelope and scrawl some calculations on the back to see if I could get a more quantitive answer. Here goes …

First of all, what does ‘fly’ mean here? Probably just short hops, like across a car park. But even this is pretty sci-fi, so how would it work?

No fire-breathing rocket engines for sure. Small cold gas thrusters underneath the car would blast out compressed air stored in a composite overwrapped pressure vessel (COPV).

This is the same basic tech that SpaceX use to adjust the attitude of their returning Falcon 9 first stages (those puffs of gas you see below). But instead of pre-stored gas, the SpaceX Roadster would presumably use a pump to gradually refill the COPV. The SpaceX Roadster would still be a pure EV!

Now, to find out if all this is really feasible, we first need to figure out the amount of air required by the thrusters to keep the Roadster hovering and compare that with the capacity of a suitable COPV.

To get a Tesla to hover is going to require upward force (thrust): approximately the thrust required to just overcome the force of gravity (bit more going up, bit less going down). That force is given by the familiar equation:

F = Ma

In this case, to just support the car against gravity, that becomes:

F = MTesla g


F is the force required in Newtons

MTesla is the Mass of the Tesla Roadster

g is the acceleration due to Earth’s gravity: 9.8 ms-2

Now to calculate the mass flow through those cold gas thrusters, we can use the specific impulse equation:

F = g Isp m


F - the force calculated above

Isp - the specific impulse for cold gas thrusters

m - the mass flow rate through the thrusters

Substituting one equation into the other, we get

MTesla g = g Isp m

Simplifying and rearranging, the mass flow, m, is then just:

m = MTesla / Isp

Wikipedia quotes the maximum specific impulse for a nitrogen thruster as 76s. Air is mostly nitrogen, so let’s just assume a round 70s (a figure SpaceX quoted for its nitrogen thrusters in a Starship presentation).

Plugging these numbers back into our equation, we get:

m = 2000 Kg / 70s = 28 Kg/s

Those SpaceX cold gas thrusters will need to emit very roughly 28 Kg of compressed air every second to keep our Roadster hovering.

Now let’s figure out how much air the COPV might hold. Looking at the dimensions of a Roadster and given the fact that the COPV will replace the rear seats, I reckon a cylindrical COPV with hemispherical ends of very roughly 2.0m x 1.0m should fit. That would conveniently have a volume of a bit more than one cubic metre.

To find out how much air the COPV could hold, we need to know its maximum pressure. My first version of this article estimated a max COPV pressure, based on old NASA tests, of around 2000 PSI. However, Elon has said recently (Feb 2021) ‘around 10,000 PSI’ – SpaceX has been doing lots of development work on COPVs.

From graphs I found online, air at a pressure of 10,000 PSI has a density of at least 500 Kg per cubic metre (depending on temperature). This sounds a lot, but is around a quarter of the mass of the whole car, so seems feasible.

If the COPV can hold about a 500 Kg of air then a Tesla Roadster might hover for about fifteen or twenty seconds at 28 Kg/s mass flow through the thrusters, as calculated above, depending on the size and max pressure of the COPV.

Some modest acceleration and deceleration for a drift in the hover wouldn’t take much extra thrust. A nice safe ~10-20 m/s (~20 – 40 mph) would allow hops of a few hundred metres.

Control would likely have to be fully automatic across the whole ‘flight’ envelope, given all the obvious risks (hitting things, running out of gas, losing control etc) of manual control.

This would use the car’s cameras (existing Teslas have eight already) to map the flight area for suitability and would need to be very conservative. Tesla’s Full Self Driving technology already creates detailed 3D maps on the fly that would work well for this.

Think a twenty second ‘flight’ doesn’t sound much? Try visualising it! In fact, it’s probably sufficient for those short hops I conjectured. Whether issues such as flying debris and noise would make this impractical, I can’t say. I also can’t say whether it might ever have functional value beyond some serious not-on-public-roads fun!

No legacy automaker could do this. A flying Roadster would be the ultimate pure-EV halo car and an incredible marketing coup.

In conclusion: yes, SpaceX CGT tech really could make a Tesla Roadster fly, allowing 15-20 seconds of hover time and a slow drift across a car park.