Gravity Racing - Flat Floors, Diffusers & Downforce

[jimcroisdale]

Hi all,

Several months ago I posted a thread about gravity racing (essentially just grown-up soapbox car racing) and some questions about the current 102mph record holding car, and how it could be bettered.

viewtopic.php?f=14&t=26724

Through that thread i was fortunate enough to meet an aerodynamic engineer who has been key in moving the project forwards, but between us, there are still a few grey areas regarding the aerodynamic implications of having a flat floor.

A flat floor is ideal for us for a few reasons:

1. Low CofG. The car will hopefully be doing 100mph at the bottom of the hill, where it has to go round a sweeping right hand curve. With a proposed 600mm track width, we feel keeping it low is essential for grip, and certainly for driver confidence just with the way it will look.

2. A flat floor will make the body shell much easier to make, as we will essentially only need to make the "top half" of a traditional streamlined shape.

Early drawings and CFD analysis show good improvements over the car we're trying to beat, but the issue of the flat floor still plays on my mind. The KEY aim is low drag for top speed, with the ability then to take a 100mph bend. We figure this means a car with neutral aero, with a slight tendency toward downforce in the lowest drag way possible.

Given a "Cockroach with Flat Bottom" shape, what are the implications of the flat floor? Would it give any downforce without a diffuser? Would, say 30mm ride height be significantly different to a 100mm ride height? What difference would it make?

Regards,

Jim

[e36jon]

Greetings Jim

Apologies if this all old news. Have you seen this Human Powered Vehicle (below). It's the current record holder at 89.59 mph, on the flat! (I remember when 60mph seemed like a big deal.) I'm showing it to illustrate what minimum drag looks like.



In the early eighties / late seventies all of the HPV designs were flat bottom, with a focus on minimum frontal area. They tried to seal off the underbody and wheels in search of low drag. Turns out that the friction / drag from being so close to the ground negated any benefits. The air at ground level is stationary, and the air at the bottom of your vehicle is moving at whatever speed you are going, and in between is shear and turbulence.

I have a great book by Hucho, "Aerodynamics of Road Vehicles" and one of the first things shown is a study of various simplified forms in the wind tunnel, and the minimum drag models all had curved underbodies (in frontal view) and rising underbody towards the rear (in profile).

All this to say minimizing the flatness of the floor will actually produce less drag (In my opinion.). Any downforce you can achieve will come at too high a drag penalty and not significantly improve your handling.

Or I could be completely wrong. Anyway, at least another opinion...

Jon

[variante]

Given a "Cockroach with Flat Bottom" shape, what are the implications of the flat floor? Would it give any downforce without a diffuser? Would, say 30mm ride height be significantly different to a 100mm ride height? What difference would it make?

Yes, even with a prefectly horizonal floor and no diffuser, it will produce some downforce.

For example, some high pressure air that accumulates at the tip of your nosecone will spill under the floor while decompressing. Judgind by the shape of your nosecone, there's a chance that airflow will spiral into little vortices. Also, low pressure at the sides of the bodyword might suck some air out of the floor, like a diffuser would do. The end result being some downforce.

As e36jon pointed out, there is more friction closer to the ground due to viscous effects. That doesn't only produce drag, but increases pressure too (producing lift).

But let's not forget about Ground Effect, which is an incredibly efficient way of producing downforce. That effect heavily depends upon ride height. 100mm height might give you about a lot less downforce than 30mm.

But, as mentioned before, you don't want to get "too close" to the ground due to the prevalence of viscous effects.
"Too close" depends on how performing your undertray is. The less performing, the higher you want it to be. In other words, there's a sweet point (in terms of ride height) between Ground Effect and Viscous Effects, and that sweet point will be closer to the ground if your undertray is potentially able to produce a lot of downforce.

All in all, I think it would be good for your car (if you actually need that downforce) to have a flat floor as long as you can find the sweet spot. You could increase downforce or manage its centre of pressure by raking your car by some fractions of a degree. I would also tend to use a diffuser because that's the only way to produce a sensible amount of downforce for your driver to feel more confident around the corners; flat floor in ground effect + diffuser is one of the most efficient ways of producing downforce, and you would't have to worry too much about the mentioned viscous effects even at 30mm.

[Tim.Wright]

Why do you need downforce? Will you be grip limited at the long sweeping curve?

I think its a big call to introduce downforce and drag elements to improve performance on one single corner.

What is the corner radius?

[e36jon]

Greetings again

I missed your original link and the info you gave at the beginning of the thread. Apologies for not reading more carefully.

So, I am going to stand by my guns on the "closer to the ground isn't better" point of view. This link will take you to a page with lots of info, including this graph below. https://www.researchgate.net/figure/Lif ... _228616843



I did a Google search for "ground effect drag for road vehicles" and found a lot of info, including a number of scanned pages from the Hucho book I mentioned previously. I also looked around at the state of the art in gravity racers, and hyper-mileage cars. The only place I can find the kind of shape you're proposing is in the gravity racing world...

I would also plant the seed that minimal frontal area isn't the same as low drag. It's just part of the equation. Have you ever seen a lenticular shaped wheel for bicycles (Google 'lenticular wheels')? They have much more frontal area but are so aerodynamic that they're banned everywhere except for the triathlon crowd. You may actually come out ahead on drag by introducing plan view curvature to the sides of your design even if it increases frontal area. Keeping the flow attached really matters for drag reduction...


EDIT: I forgot I had this book that couldn't be more relevant for what you're doing, "The Leading Edge, Aerodynamic Design of Ultra Streamlined Land Vehicles" by Goro Tamai. Edit of edit: I see that Flyin' Frog suggested this text a year ago in the original thread... It's great to have something so focused as figuring out what matters or not from general texts can be tough. I also have "Competition Car Aerodynamics" and it wouldn't help you much unless you really knew what to ignore... My last Aero text is "Low Speed Wind Tunnel Testing" by Barlow Rae and Pope. It's not an easy read...

[jimcroisdale]

Why do you need downforce? Will you be grip limited at the long sweeping curve?

I think its a big call to introduce downforce and drag elements to improve performance on one single corner.

What is the corner radius?

We don't need downforce per se, but we need the design to tend towards downforce if anything, and certainly eliminate the possibility of lift being created for reasons we can't forsee. It is likely that the car will only ever be tested at speeds anywhere near the 100mph mark at the event itself.

Our priorities are low drag and also simplicity of build, which a flat floor helps with - we would only have to manufacture "half a body shell", and as amateurs this makes a significant difference to the likelihood of success.

My thought was always "how do the flat floor and road surface react"? If we need a bigger gap to decrease drag, that means raising the CofG. Obviously we dont want to do this too much - just enough.

[jjn9128]

We don't need downforce per se, but we need the design to tend towards downforce if anything, and certainly eliminate the possibility of lift being created for reasons we can't forsee. It is likely that the car will only ever be tested at speeds anywhere near the 100mph mark at the event itself.

Our priorities are low drag and also simplicity of build, which a flat floor helps with - we would only have to manufacture "half a body shell", and as amateurs this makes a significant difference to the likelihood of success.

My thought was always "how do the flat floor and road surface react"? If we need a bigger gap to decrease drag, that means raising the CofG. Obviously we dont want to do this too much - just enough.

There will be a lot of "it depends on..." in any answer to this.

From your description I'm imagining a half tear drop shaped car. Not sure where the wheels fit in (3 wheels??), but I would say the ideal, aerodynamically, would be to seal the car to the ground to prevent any flow ingress under the floor - though whether you could achieve this without appreciably increasing mechanical drag is a big question.

The shape of your leading edge will make a difference - if it is really sharp you will likely get a massive low pressure spike with local flow separation. The ride height will affect the magnitude of this spike and any low pressure underneath, which will give downforce but also drag.

The purpose of a diffuser is to gently return pressure to atmospheric so, depending on what the top half is doing and what the pressure behind the car is, a well designed diffuser should reduce drag. Obviously you have to balance ride height with COG for the best compromise.

It really needs testing with your geometry to see what's optimal though.

[jimcroisdale]

See pics below. Do these give you a better flavour of what we're looking at?

[jjn9128]

See pics below. Do these give you a better flavour of what we're looking at?

Yes. I would look at adding a radius to the bottom edge and maybe slightly raising the nose on the front edge to blend into the floor. Not sure how you're planning to manufacture - maybe that adds a piece and a split line where it's really unwanted.

If you have CFD - do you have numbers for Cx and Cz??

[e36jon]

Thanks to jimcroisdale for the imagery. With apologies to him, please see my modifications below.



I like the way that the shape he came up with moves the highest point as far back as possible. Maintaining high(er) pressure on the body helps with laminar / attached flow. Trying to do the same with the bottom will net similar results. It will get 'weird' between the wheels as the cross sectional area will change dramatically, so figuring out how to manage this will be a challenge.

The indicated floor contour (in 'racing-red') could be flat normal to this pic (into the page), and the diffuser is a straight line, so flat-flat. The diffuser angle shouldn't exceed 7 (maybe 10) degrees relative to the ground-plane to avoid separation. Add in the suggested corner radii and boom, you're world champ...

If you look at the white world-record HPV, and the red one in the original post, they are both designed to maintain laminar flow for as long as possible. In that vein, it may be 'better' to go with a pop-up head area, like the red bike, and then optimize the front end as much as possible. In both cases that means no joints in that critical area. I bring this up because getting the windshield-to-body joint right, as shown in the images, will be tough. If it's particularly not-right it will mess up flow all the way back.

Having fun yet? I am! Thanks for letting me (us) play along...

Jon

[roon]

Just put wheels on one of these. Way easier.



/expertise

[e36jon]

The roof-rack is awesome! There are some that would actually be a decent starting point...

And, there's precedent for repurposing things to go fast: The 1952 SoCal Belly Tank Lakester...

[jimcroisdale]

There are a couple of cars in the UK that use roof boxes! None with the wheels inboard though.... And have you seen the price of them?

[9point81]

I’d be one of those roofbox racers mentioned above!

Not the most aero set-up I grant you, but it was only built for a bit of fun! However if I do build another, it would be more along the lines of this;

[e36jon]

That "Thule Racer" is going to make me smile all day. So awesome.

That second vehicle has some serious sponsors, so I suspect I just don't fully comprehend what I am seeing there. Seems like there is WAY too much free space inside, which equals too much frontal area. That white speed-record bike I posted earlier is packed inside. I was also advocating for getting things up away from the ground, but again, this seems extreme. Maybe there's a mud-bog element coming up just out of frame?

Here's the world-record holding human powered vehicle without the fairing:



Edit: Note the covers over both wheels. There's a benefit to limiting the volume of air that the rotating wheel interacts with. Don't remember if that's covered in "The Leading Edge" or not.