Gravity Racing - Improving Aero to Beat the Best

[jimcroisdale]

Hi all,

My racing sport of choice is gravity racing. Each year, there's a race in Canada that basically features the mother of all hills - with an 18% grade for the final drop. It's mental. This year this car did 101mph, a new world record.







Basically, i'm trying to work out how this car can be beaten. It's got about as small a frontal area as possible for a human to fit into, is the max length allowed, and is always ballasted up to the max weight allowance. The tyres are essentially a tried and tested narrow "kart" style tyre, known to be safe at such speeds. As you can imagine, stability is key.

So with all that in mind, as far as I can see (CFD modelling of the body aside, which to my knowledge they didnt do) the only area that could perhaps be improved on is the underbody.

So that leads me to my question: On a car with no engine and no requirement for any real downforce, what effect does the flat underbody and low ride height that this car and others like it have on drag? Is this an area for substantial improvement, or have they got it pretty much right?

Regards,

Jim

leedsgravityracing.wordpress.com

[Greg Locock]

I suggest you look at the history of aerodynamics on solar cars and human powered vehicles. The modern ones seem a lot more sophisticated than a simple minimum frontal area approach. One easy way to think about it is that in the absence of lift or downforce you are mostly concerned with maintaining laminar flow as far back as possible, to minimise skin friction. That is, minimum frontal area is not optimal. Laminar flow is important since it has only 10% of the drag per unit area of skin as turbulent flow. A really good laminar flow design might manage to keep the front 60% of the body laminar, whereas a traditional shape might only manage 20%. Wool tufts are your friend.

So I'd expect the optimum to be more like a stretched egg shape (in 3d)

http://www.metaefficient.com/wp-content ... e-bike.jpg

but with wheelcovers.

You can get a long way with hand analysis but CFD was a big step forward, but before then we built 1/5 scale models to make sure that the vehicle was trimmed for zero lift and used shapes based on laminar flow wing profiles. A somewhat impractical full size bodyshell, smoothed to the umpteenth degree, came out with a /measured/ Cd of 0.07 on a frontal area basis once we trimmed it to zero lift.

At a rough guess you could be looking at a 20-50% improvement in drag.

[hollus]

Would this be of any help?
https://www.f1technical.net/forum/viewt ... =14&t=6677

[NoDivergence]

Not sure why the tip is so sharp, you are going very subsonic here. Also, seems like a lot of frontal area compared to something like this that doesn't need a "cockpit." That's also quite a draggy design, even within those structural constraints. The blending of the "cockpit" to body is very poorly done, look at how aircraft speed racers do it. The rear end bodywork does have a concave structure (Stratford pressure recovery) but seems like it could be extended further.

This is what I'm thinking more of, an excellent natural laminar flow design with no additional cockpit and very good control of the ground/body airflow interaction. https://www.solidworks.com/sw/successes ... lo&id=6682
Extend it out (increases stability and Reynolds number) and add weight for this discipline.

[Greg Locock]

NoDivergence, yes that looks more like it. Eyeball aero can be misleading, but you'll get the laminar to turbulent transition as soon as you stop accelerating the air, in the photos in the first post where the nose joins the body.

Also it looks as though you may be getting some downforce from the nose, that's probably a bad idea.

[jimcroisdale]

Thanks for the replies guys.

A quick rough in Tinkercad:





Is that one route to go down?

Jim

[Greg Locock]

Nice, but i think a real aero guy will be able to make an assessment of the drag contribution from the spats and 'wings' and interference

Also note that the higher velocities under the car will generate drag inducing downforce

[bill shoe]

I'll throw additional advice on the pile:
Human powered vehicles (bicycles with enclosed bodywork) have learned that most of their total drag occurs from the little bits of tire that are left unfaired in the last few mm near the road surface. This is probably because, as someone previously mentioned, these little snippets of exposed tire cause turbulent flow instead of the laminar flow that prevails over the remainder of the body. HPV racers typically enter their bike, then someone on the outside holds them upright, while a third person adds racer tape around the lower tire extremity. This would probably minimize the lower tire drag for you as well.

[Scootin159]

Are four wheels mandatory? Seems going with a single rear wheel would be a quick and easy way to cut both drag and rolling resistance.

[Jolle]

The shell eco-marathon always gives great inspiration for very low drag vehicles.

what you'll see there with all winners, are smooth transitions between surfaces

[graham.reeds]

Any linkage to the ruleset?

[flynfrog]

I think Greg and Bill have sent you down a pretty good path. I also recommend this book.

https://www.amazon.com/Leading-Edge-Eng ... 0PWAQE9FCM


you dont beat them by gaining 25% you beat them by finding the missing 1% they missed everywhere.

A few other notes wind tunnel time is not that pricey any more we have a tunnel local to me they does quite a few land speed cars.

http://www.darkotech.com/


Some other hints
http://www.jackcostella.com/#portfolio

[Tommy Cookers]

[quote="bill shoe"]
..........these little snippets of exposed tire cause turbulent flow instead of the laminar flow that prevails over the remainder of the body.......[/quote]

how do we have laminar flow prevailing over the remainder of the body ?
surely we still have boundary layer thickening, hence becoming turbulent, part way along the body
at best just starting somewhat further back and somewhat less thick than we might have had in the past
we are at the Reynolds numbers of typical light aircraft wings
there are no such wings with 8' or 12' or 15' chord at c. 100 mph that have entirely laminar flow

[Greg Locock]

A typical cockroach shaped solar car maintains laminar flow to about 30% of it length. With careful design a pure shape with no joint lines etc can manage 60%.

Detecting the transition from laminar to turbulent is most easily done with a stethoscope in the windtunnel, the difference in noise is easy to hear. Separation is the other thing to watch for since that increases the recovery drag ,which would be zero if there were no separation, and that's easy to see with wool tufts.

[bill shoe]

how do we have laminar flow prevailing over the remainder of the body ?

I assumed laminar flow over some of the body, but I was rather guessing. I should have just said lower-drag on main part of body and relatively high drag around unfaired wheels.