Underfloor and Body design of Greenpower Car (similar to Electrathon)
 

Hello all, new here.
I'm currently designing a Greenpower car and would like to know which type of underfloor and ground clearance would be best for low drag. I don't know whether this is a good place to ask but since it's eco, electric and focused on aero, I'll try here.

Some information on a Greenpower car, it is essentially a extremely small single seater electric car and we enter into races with other student built cars. We have the same batteries and motor and therefore we have to design the most efficient car possibly to go as fast and far as possible in a hour race. Aerodynamics is the biggest factor which is why I want to get it right.
(http://www.greenpower.co.uk)

Fastest cars tend to go at 40mph + which all of my CFD testing is at. The racing is done on large tracks so only minimal steering angle is required.

We are given a set of rules and regulations that we have to comply with and the couple mains one are:

Minimum 500mm track, 4 wheels are required.
Maximum 2800mm length, 1200mm height and width
Only 14" - 20" bicycle wheels are allowed.
Must have a open cockpit opening of 350mm wide and 600mm long.
Ground clearance can't be any less than 30mm.
The bottom of the driver and batteries can't be higher than 100mm off the ground for roll safety.
Finally, it has to fit me (6ft) in the car.

If you are interested, you can have a look at this link which has many of Greenpower cars examples.
http://www.flickr.com/photos/querth/...h/21880485434/

Two most recent winners - See the difference in ground clearance and FA
http://www.facebook.com/GreenpowerRa...type=3&theater

With that out of the way, I have spent hundreds of hours designing a wide variety of bodies and testing it in a CFD software called Autodesk Flow Design. This doesn't however include any moving road simulation which is important for seeing what effect the ground clearance has on airflow underneath and behind the car. I was wondering if any of you guys would have any idea of what would be better in reality.

What would be better, 30mm of ground clearance which reduces the turbulence underneath the car but increases frontal area and makes it more difficult to taper the body at the rear without destroying the underflow,

or

95mm of ground clearance (maximum), which reduces frontal area which is relatively small already but have more turbulence underneath which can cause more drag and air spilling from underneath.

Would be interested to hear what your opinion is.

The link to the gravity racers is not working, let's try this again.

https://www.flickr.com/photos/querth...7637777375534/

Right one?

The biggest difference between those gravity racer examples and ones favored in this forum is the outboard wheels (there) verses a monolithic body enclosing the wheels (here).

EDIT: Let's test that Facebook link too.

https://www.facebook.com/GreenpowerRacing

Correct?

The first link is almost right, just not the right album. I'm sorry about the links, too new here and can't post them just yet. For the time being, just put the two dots in the link.

http://www.flickr.com/photos/querth/...57659968306378

The second link wasn't any good, below are two better ones. These are the currently the fastest on the grid.

http://www.flickr.com/photos/querth/...7659968306378/

http://www.flickr.com/photos/querth/...7659968306378/

Hope this helps

It depends. :) Firstly, this is the sort of question that is of interest here, but...
...maybe you should be schooling us.

Questions:
Are you competing in Goblin, 24 or 24+ class (I hit 6ft when I was 14)?
This standard motor and two batteries—where do they go?
The ~14x24" rule for the cockpit opening how is that measured?
What fabrication skills and tools do you have a available?

Here are three contrarian shots in the dark. First Luigi Colani:
http://ecomodder.com/forum/member-fr...wu5wo1-500.jpg
The body is flat on top and curved on the bottom: free downforce. Minimal body/ground interference drag. This car ran at Bonneville Salt Flats.

Secondly, my own design I would like to run at Bonneville someday:
http://ecomodder.com/forum/member-fr...11-5-38-12.png
The four-wheel layout mimics the Speed Demon. It may or may not be allowed under your sanctioning body's rules.

Thirdly, a screenshot from a Youtube video which I believe represents an early iteration of the Edison2:
http://ecomodder.com/forum/member-fr...9-10-19-49.png

Yeah, well the hours spent is really just messing around with a variety of shapes, but with no definite answers which leaves me with no real idea of what is the ideal ground clearance and plus the software is nowhere near accurate enough to give me useful results, only useful for showing pressure areas and rough idea of the turbulence behind the car.

I'm competing in the F24+, which has the exact same technical regulations as F24 but we have shorter races with the same energy available and therefore its faster.

All of the car run with the 24V electric motor in the rear although nothing in the rules says it can't be front wheel drive but this is a complex and heavy system. The 2 x 12 V car batteries can be placed anywhere in the car as along they are 100mm or less off the ground, some cars have them in front of the feet, behind the helmet or one on each side of the driver. It all depends on the body.

Bearing in mind, the car is much smaller than a standard car, the electric motor is only 100mm in diameter and 200mm long. Some of the smallest and fastest cars only measure about 400mm high, and 570mm wide and 2.8m long and this would be including the wheels.

This is the rule on cockpit opening.

T7.4. The cockpit opening, when viewed from directly above must be able to accommodate a rectangle 600 mm in length and 350 mm wide with no intrusions, e.g. helmet fairings, instrument covers, etc. must not overhang this area. The drivers’ helmet must be positioned at the rearmost point possible of the cockpit, creating a clear space in front of the drivers’ helmet.

In terms of fabrications method and tools, the car will be mostly handmade, will probably make some moulds to form the body which might turn into a monocoque. Im open to anything, its only concept at this stage. The body design ideally should be simple so that the time spent in the manufacturing is reduced greatly. Also a possibility of some parts like jig panels etc being CNC routed locally, but will cost more.

For this kind of competition, we don't need any lift or downforce at all, all focussed on getting the lowest drag possible and I kind of think if there is any downforce or lift being exerted on the car, it's using up energy from trying to moving the car around.

I would love to do that almost three wheeler design, but the rules states that there must be 4 wheels, 500mm minimum wheel track and must be symmetric along the centreline.

I will try post some pictures of some of body designs I have come up with.

This sounds similar to the Electrathon racers that compete locally. Those are limited by battery weight, not number. What are the battery dimensions?

Electrathon racers tend to have stiff, if any, suspension and a thin noisy body that rattles as they go by. I would consider Polymetal. It is stiff and would make for a rigid moncoque, but would favor simple curved surfaces.

The 'almost three-wheeler' design could be built with 4 wheels, even enclosed. The rules sound like they would permit a narrow rear track, like an Isetta. The basis is that the body can be constructed from triangles, diamonds or hexagons. Each edge is mathematically defined to whatever precision you can cut to. If it's flattened into a 'peel pattern' it could be cut out of a flat sheet with the darts cut as overlapping tabs that are riveted or spot welded. Sort of a quasi-compound curve.

I will now wait quietly for your pictures.

Interesting, never heard of Electrathon before, but looking at it, it's quite similar to Greenpower here in the UK. They have very similar design and rules as we have. Only thing is that they can have three wheelers, so much better than four wheels for efficiency, can only wish that we had that rule :( Going to spend some time researching into some of their cars.

We have two batteries that are supplied by Greenpower and they are 36Ah batteries, (195L 130W 165H)

Yes, if the body was quite simple, I could use that method for the car, however I think in terms of body designs I have been working on, I was thinking more on the lines of strip planking which has been used for kayak building i think, will allow me to make consistent smooth curves on the edges to prevent air spilling and causing drag. Also the body needs to be really smooth to keep the boundary layer close to the surface as possible to minimise drag.

Now have 5 posts which means I'm allowed to post pictures, coming soon.

https://i.imgsafe.org/6be3eef.png

This one is a front wheels in, rear wheels out. Main idea was to taper the rear completely so it's a smooth flow. However there is additional drag from the exposed wheel fairings. This has 100mm of ground clearance which is maximum.

https://i.imgsafe.org/6c73c06.png
https://i.imgsafe.org/6d00dad.png

This one has what I call a "channel" underneath the car, as the wheels have to have fairing, I just used the whole side of the car as a fairing which massively reduce the air spilling underneath. The top of the channel is 100mm off the ground but the side parts are only 30mm off the ground. However the question I was using this thread for was how does the ground clearance affect the flow underneath, does it stall it and cause chaos. This design depends on that alot. This has been producing the similar drag levels to the first one despite having more frontal area.

https://i.imgsafe.org/6d78ac8.png

Another one, high off the ground, all wheels in. As the air go through the wheels underneath the car, it tends to expand and cause more turbulence behind the car, but this has the minimal frontal area.

design
 

I would have you consider the 1993 Team New England TNE-II design.It would have won the North American Solar Challenge had it not been for cloudy weather.It was faster than the 'winner' by far.Cd 0.10 and low projected frontal area,plus very stable in cornering.
https://en.wikipedia.org/wiki/Team_New_England
Also,the 2013 Cambridge University CUER solar racer,Cd 0.10,and even lower projected frontal area and surface area.
http://i1271.photobucket.com/albums/...itled12_15.jpg
Laminar boundary layer shapes are going to look great in the wind tunnel and CFD,depending upon Reynolds number,but in a real race environment,it will be impossible to attain a laminar boundary layer flow due to Earth's turbulent boundary in which you'll be racing.

The designs you show are thoughtful. The first one should be very stable, it's used on land-speed record cars. The second one I've seen called a catamaran design. That was the first thing I thought of when it sounded like the question was 'what's the perfect single number for all cases everywhen'. The third is my least favored.

It looks like you've been thinking about that mandated hole in the top, with some drainage channels at the back. In some ways it more important than the bottom, being more tightly regulated. What would you think of an everted NACA duct for the leading edge of the opening? Or just adding a half-circle to the rectangle with a plexigalss windscreen. Angle the sides up past the driver's eye level like a F1 race car?

How do you find the NURBs modeling? Is there a tool path to get you to a cutting pattern? I like polygon modelers because you can pick dimensions off even non-contiguous vertexes/vertices.

Here is something to tie it all together: aerohead's example (just implemented as a three-wheeler) in a polygon modeler (just blown up to motor-home size*). It's completely malleable on the X-, Y- and Z-coordinates, to fit your need.

http://ecomodder.com/forum/member-fr...-w-caption.jpg

*I should update the graphics since the motor home donor vehicle is gone. Think Tesla Model S drivetrain.

Tried a few models of having 1 or 2 naca duct or have the top surface slope down as much as possible. However any of these don't reduce the drag much as they disturb the air earlier, causing more turbulence. The only solution that might help is the last one, which slopes down as much as possible without interfering me getting out of the car in emergencies. This reduces the low pressure in the cockpit opening.

https://i.imgsafe.org/12db569.png
https://i.imgsafe.org/e65e799.png
https://i.imgsafe.org/e6b0d11.png
https://i.imgsafe.org/62907e4.png

I like the second design better than the first as it will most likely be easier to manufacture especially for the mounting for the rear wheels compared to the first one. But I have concerns about the airflow under the car as it might churn up and cause large amount of drag or it might keep its laminar flow??

I don't use NURBS surface modeling for my models, I use Solidworks CAD software to make all of the bodies including the chassis component and so on. I could export the body shape into another software for flattening but to be honest the body is all compound curves to keep the airflow attached. The second body design will consist of two parts, plus a nose cone and am planning to make simple moulds for it.

Second and first from Permalink #9? The catamaran body shouldn't churn the air if it's smoothly finished. It may have problems with increase surface area in crosswinds. Note that aerohead's example and my third one use a 'wingtip' pointed downward to enclose the wheel.

Have you compared raising the cockpit edge toward the rear, rather than the drainage channels on the sides of the headrest. It seems like that would bleed turbulence rather than slicing it off.

Solidworks uses splines, see https://en.wikipedia.org/wiki/Non-un...ional_B-spline. It is a parametric solid modeler as compared to a direct polygon modeler. Why Use NURBS for Surface Modeling...? suggests that Solidworks is needlessly complex for what you are trying to do, compared to say Rhino 3D or Maya. But, use what you have.

cockpit
 

Would the rule book allow something like an elastic Lycra Spandex panel to enclose this cockpit area ,which Velco's onto your helmet leading edge below the visor ?
In an emergency egress situation it would simply shear away at the velcro attachment points as soon as you moved appreciably.
It WOULD be aeroelastic and probably 'shudder' a bit,but would be better than that enormous void which cannot support any kind of airflow.

Freebeard - What do you mean by raising the sides of the cockpit so it bleeds turbulence? Unfortunately, I can't have any bodywork above the eye level except the front wheel fairings and mirrors, its one of the rules. So at the moment I'm only peeking over. I have also included a picture of my previous car with the 1st body and the 2nd body I made for it and as you can see, I can only just see over the top.

Aerohead - Unfortunately, it wouldn't be legal, it has to be a permanent cockpit opening. Would have a fully enclosed driver if I had my choice.

https://i.imgsafe.org/2c5677b.png
https://i.imgsafe.org/ca2a7a8.png

I made two different style bodies for my previous car, one with 30mm of ground clearance, 16" wheels with full fairings around them. That one also had a long nose and a chopped off backend.
I wanted to improve the performance, so I decided to modify the chassis so the ground clearance is 85mm, as well as using smaller 14" wheels. I also reduced the nose to the minimum 200mm and made the back end tapered. Both of the bodies came to the same ballpark of performance and therefore I don't have a real idea of which approach to make for the new bodywork. Your opinions?

I also did more testing of having the channel vs no channel, and it showed a significant difference, about 25% increase in the drag when there's no channel (although it would be less than this in reality) because when the airflow trips over, it has a effect on the size of the wake behind the car. It also reduces the pressure in front of the helmet as you can't see the green area in front of helmet. So will definitely be having this feature.

https://i.imgsafe.org/58d8571.png

Yeah, I have been using Solidworks for a long time and am used to it. Also been using a Solidworks built-in flow software for the recent tests, and has been giving much more useful and consistent results.

So that answers the question about the cockpit opening. The sight-line requirements and CDF analysis point the way. Bringing the fairing forward to cover the back of the helmet looks like a good move if the rules allow it.

What materials did you use on the two previous iterations? The beveled sides on the first one look like they would provide a little more front wheel clearance. Did the blunt nose and longer tail show any improvement?

opinion
 

Just thinking out loud:
*consider a 2-D streamline form to minimize frontal area
http://i1271.photobucket.com/albums/...titled4_26.jpg
*do the 30mm clearance at the front axle
*do the 100mm clearance at the rear
*this offset will allow you to use the entire under body as a drag reducing diffuser (no downforce).
*place the batteries just above the belly where you can to keep the C.G. low for roll stability.
*rake the body with low front/high tail.This will help keep the upper airflow attacking the body all the way to the canopy,where the flow will be compromised by the opening.
*place the canopy as far forward as you can to use whatever is left of the 2,800mm to streamline behind your head,like this Summers Brother's Goldenrod
http://i1271.photobucket.com/albums/...itled23_10.jpg
*try to fully integrate your helmet as a bulbous leading edge to the fairing.
*integrating the roll bar into this rear fairing will help the flow to recover from the open canopy turbulence.
*if you can't get the roll bar inside the fairing,then streamline the tubing as best you can and blend it into the top of the shell to kill hook vortice formation.
*put your 500mm wheel track in the rear
*design the wider front track to agree with the streamline contour
*Fully enclosing the wheels,even though it adds frontal area,completely offsets the interference drag of 'open' wheels.

I used corroflute, which is a very light plastic material and used Polyproplene hot glue to fasten the panels together on the inside and a strip of electrical tape.

The bezels are for reducing the frontal area as much as possible as corroflute is a flat sheet material and can't be formed into compound curves and therefore I used three sections on the side as shown below. By the way all of the wheels are cambered 10 degrees inward.
https://i.imgsafe.org/343e554.jpg

Aerodynamically, yes it was a better body with much better rear aero as it tapers to a point. However other factors such as increased in car weight, and dying batteries meant that it performed about the same as the first version.

Aerohead - Thanks for the ideas, there's a lot of restrictions in the rules that prevents some of the ideas you're mentioning, I will try to rattle through them, a lot of them I have tried to implement into my design.

The 2D side profile of my best body so far is similar to a streamline shape as you can see below, just stretched out. It's slight cambered towards the front to minimise the squashing of the airflow at front of the car.
https://i.imgsafe.org/e9de692.png

This has a 95mm ground clearance at the middle and towards the front and increases up to 150-170mm at the backm, so similar increase to what you're saying. I'm not sure I was clear before but the 100mm ground clear only applies to the batteries base height and the driver's bottom. Everything else can be much higher so there is some flexibility there.

The top surface does keep increasing in height until the canopy, same as you suggested. Also the head fairing is as faired as much as possible.https://i.imgsafe.org/95f36e3.png

The roll bar has to be exposed because there is a line is drawn from the front roll bar to the rear, there must be 50mm of clearance between the line and my helmet. Also recently, they banned any fairing for the top 150mm of the rollbar, the other teams weren't very happy at all about that.

The plan is both the front and rear track to be at the minimum of 500mm plus all wheels have a 8-10 degrees negative camber to reduce the frontal area even more. The frontal area is about 0.225m^2 or less, about 10 times smaller than an average car.

I did some flow simulations on the three designs I posted on here, I did them on the more accurate Solidworks software and it gave me some surprising results. The 3 wheeler even though it has completely tapered rear end, the amount of surface area made it a lot worse, about 40% higher drag force than the second design. So I would agree with you with that completely enclosed wheels are the way to go.

Also if you're interested to know, the third design (the catamaran design with the side flanges cut off) I posted in permalink #9 has about 15% more drag force than the second design. The channel underneath and the side flanges is definitely helping out reducing the drag.

For work with corroflute (I assume like coroplast) that looks really good. The slab sided one was probably less work? Nice surface finish though—is that like coroplast with an aluminum skin?

The teardrop aerohead posted is ideal, then there's those rules. The teardrop has maximum width at 30% of the length. You can truncate the last 10%, a Phantom tail. The rear axle could be at 60% plus or minus, with reduced track the further back it is.

http://ecomodder.com/forum/member-fr...14-1-42-00.png

This shows a flat floor with a half-exponential horn replacing the flat rear truncation. It provides space for the rear wheels.

catamaran
 

I'm not surprised about the catamaran.HONDA tested one for the 1993 World Solar Challenge and it came in Cd 0.15 (if my memory serves me) vs Cd 0.10 for the 'pumpkin seed'/ flattened-torpedo (as Goro Tamai refers to them) shape that they ultimately used in competition.And won.:)

roll bar
 

It's a shame about the exposed roll bar.It will have premature laminar boundary layer separation.
http://i1271.photobucket.com/albums/...d2/06-2834.jpg
Would the rule book allow some sort of dimples,trip wire, or micro-vortex generators on the tubing?
I know it's kind of splitting hairs,but the interference drag from the tubing loop will be disproportionately large,especially since it's in the aft-body area,compromising the head fairing.
During World War-II,the directional hoop antennas on Allied,Luftwaffe,and even Japanese Imperial aircraft were often enclosed within streamlined housings to cut drag and extend flight range.
http://i1271.photobucket.com/albums/...ntitled315.jpg
http://i1271.photobucket.com/albums/...titled45_1.jpg

Freebeard - Thanks, it's just coroplast, no aluminium skin or anything, it's a light material although difficult to shape. The second body was much easier, all done by hand, where as the first one, I had to make a jig to lay up the panels in and glue it all up.

I wasn't sure which ground clearance you were thinking of so I tried both 30mm and 100mm models. The inverted horn shape at the back does help to reduce drag quite a bit. The 100mm model gave me a slightly lower drag than the 30mm model but only by 3% despite having a much larger frontal area. Both model were about 15-17% less drag than catamaran design in permalink #9. Here's some pictures of what I did.
https://i.imgsafe.org/821c4b0.jpg
https://i.imgsafe.org/0a20ab8.jpg

After this I went back to the catamaran design as I wasn't happy with the front section as I felt it was too abrupt to the airflow. I moved beginning of the side flanges much further back and rounded the front nose off as much as possible. I also made top two edges along the body to have a much bigger radius to allow the airflow to flow more easily around the corners. This gave me the lowest drag force so far, about 22% less than the original catamaran body and a drag coefficient of 0.11. I'm going to have do tests on the most of the models at 15% yaw as airflow is never going to hit the car dead on with crosswind and so on.

Aerohead - Unfortunately there's not much I can do on the rollbar, the only thing I can do is, there's been a trend lately to hide the rollbar behind the helmet to minimise the impact of it by having a very small bend radius, like I had on my previous car.

All of the rules of Greenpower can be found here, there's a section on roll bars, banning the use of fairings - http://www.greenpower.co.uk/sites/de...ons%202016.pdf

No dimples on the roll bar...would a very rough paintjob help?

Thanks for that. I've had some predictions verified before, but this is the first time someone has confirmed a suggestion I've made, so this is a big deal to me. :thumbup:

I know that you can't expect absolute Cd values from your software, but just for my bragging rights, could you close off the horn with a flat truncation and see how much difference is made there? TIA

Coroplast, if you hadn't noticed, is popular here. It's light and inexpensive. There is a triple-wall variety. The open-wheel racer I posted in Permalink #4 is based on a 4-sided pyramid, shrunk on one diagonal and stretched on the other. It has had a geodesic subdivision applied, in this case 6v or a frequency of six. At 2v it is very angular, like an F-117. Someday I would like to compare 2 through 6v in a wind tunnel. Anyway...

The point being that it is all flat surfaces, be they triangles, diamonds, hexagons or a mixture. If you take a big sheet of coroplast you could cut reliefs on the inner edges and darts on the perimeter to fold a geodesic shape. Here is a (spherical) example from Google:
http://starship-enterprises.net/file...%20Pattern.jpg
https://www.google.com/search?q=geodesic+peel+pattern

Here is something I prepared for another thread showing an edge treatment for coroplast. Un-edged on the left, wrapped with tape in the center, and double folded and wrapped on the right.

http://ecomodder.com/forum/member-fr...12-1-55-49.png

Less inexpensive but more permament is Polymetal. I've worked with samples of it:

http://ecomodder.com/forum/member-fr...1-100-0866.jpg

Compared with 5/8th inch plywood, it is about as stiff, 1/10th the weight and twice the cost.

Check the rule book for perforated base plate. Applied to your roll bar it would be like teeth of a comb diminishing in size behind the roll bar uprights, but not touching it. I doubt a strip of sandpaper on the front would do much in the turbulence coming off the helmet. But you could wear a feather boa or inflated HANS device to fill up the opening you can't put a tonneau cover over.


Edit: Here's another example, using hexagons:
http://cdn.shopify.com/s/files/1/036...g?v=1397584600
http://thecookinmama.com/category/diy/

They didn't make V-cuts, or just wide cuts, so it's folded the wrong way. With an ellipsoid, the relief cuts on the inner surface will be wider in the pointy ends than in the middle. Due to the dihedral angles.

I had a look at removing the inverted horn, and it had odd effects on the drag figures. The front section has not been changed at all. On the model that has 100mm of ground clearance, flat base, removing the inverted horn increased the drag from 5.4N to 5.9N, increase of almost 10%. The pictures shows that without the airflow that going through the horn and pushing up the wake, it allows the airflow from the top go down and push aside all of the other airflow.
https://i.imgsafe.org/1e9883f.png
https://i.imgsafe.org/27f132b.png
However what's interesting is that for the 30mm ground clearance, the drag values are near identical. I think this is because of the size of the horn and the low ground clearance makes for a rapid expansion of the airflow which has similar drag level to having a solid back end.
https://i.imgsafe.org/27a5fa1.png
It all depends on the size of the horn and the ground clearance but I think it can be used to reduce the drag if done correctly. In terms of the Cd accuracy, in terms of real world predictions, the software wouldn't be any good unless I can get some real world validation, but it's accurate and consistent for when comparing designs and their drag values.

The two rules on rollbar aerodynamics:

Aluminium or steel square or circular section roll bars are to be used and must be strong enough and of sufficient dimensions to perform satisfactorily - Means that I can't use aero profile tubing.

The top 150 mm of the roll bar must not have any fairing or other aerodynamic aid - Not really specific as I could have a fairing behind it that's mounted on the body but not attached or touching the rollbar. Kind of against the spirit of the rules but is a possible loophole so it's possible to have a plate behind it. Plus I'm only having the minimum 150mm exposed to the airflow.

Thanks for the info on the different methods. Although it would be easier to make, any sharp edges will make the air trip over and I need to make sure that the airflow stays laminar as possible for long as possible for the best performance. If I was to make another body out of corroflute, I would make it like my first body and use thin long strips and have multiple chamfers for the corners. Also incorporate the foam/fibreglass nose of the second one.
But I think I will make a fibreglass/carbonfibre bodywork that is smooth and curvy. Just need to make sure the body shape is right .

Thanks for taking that extra step. I bookmarked you post. I understand that varying parameters, more than just ground clearance (straight vs exponential horn, etc.) could optimize the result. And that it's for comparison purposes only. I'll take 10% as the significant number. It's not really my idea, it came from a rectangular duct under a drift racer or something.

I mapped out how I would tackle coroplast FWIW. I'm a fan of formless construction. If you make a wide slot in the flutes and backside it would result in a radius. The ones going laterally are the problem, so you lay out with diamonds. Chain them together and the result is a long strip with non-parallel edges.

That first body you built could be rendered as a triangulated low-poly model. I don't know much about the tool path from the Solidworks model to a cutting pattern.

If your thinking fibreglass/carbonfibre bodywork, look into http://basalt.guru. It's a competitive product. You could for instance layer a fiberglass skin with a basalt mesh reinforcing backing, the thermal expansion is the same.

For that first body, I exported that model into a paper folding software called Pepakura and once I got the pattern which wasn't perfect, had a quite a few broken panels, I had to modify it in Draftsight, and used that to cut the big corroflute sheets.

If I'm doing a body similar to the catamaran design, I think I would use the underside of the body as part of the chassis which is a very wide upside down 'U' shape and use a laminate structure to form the backbone of the chassis. The front 200-300mm will be a foam structure for the crash protection. The whole top and side surfaces of the body will be a single piece of carbon fibre or fibreglass shell with minor reinforcement. All depends on time as always :)

The channel underneath makes it useful for making the chassis stiffer but I have to make sure that having a channel underneath is good thing, the tests tells me that it helps a lot for having really low drag. If I remove the side flanges of it, it tends to increase the drag (10%) by exposing the more of the wheels and as the wheels are static in the virtual wind tunnel, the effect would be greater when the tiny wheels are actually spinning around at 40mph. This is usually minimised by using wheel fairings underneath.
So its a case of whether having 4 individual wheel fairings (smallest FA) or two long sweeping fairings like the catamaran design (slightly bigger FA) or completely lowering the ground clearance down to 30mm and reduce the draggy underflow (biggest FA). What would your choice be?

A flat floor that has a stiffening longitudinal corrugation that begins toward the front, is wide and flat under the driver seat and then begins an exponential flare to the truncation of the aeroform. Cross pieces between the tunnel and the catamaran section for seat mounts.

Maybe a flat, or at least lower, floor between the front wheels with half-catamarans that grow out of that to enclose the rear wheels?

At some point packaging the drivetrain and suspension on that floorpan is all beyond me.

I just noticed this thread in my monthly mail. Bravo, and apologies for lateness/duplication.
While the air under the car is certainly in shear, a popular way to get around the lack of a moving ground plane is to simply mirror the shape and observe the gap between the two "bottoms." An ideal shape in free air, when brought close to its twin, will distort progressively so that the two adjacent surfaces become flat as they touch, leaving no visible seam as they merge into one body of the original shape but double volume.

The exposed wheel bottoms add frontal area, but enough ground clearance for easy street use is good, because it can better handle the inevitable imprecision. Wheel exposures benefit from fairing, but make sure to align it with the local airflow, not just the wheel itself. Use smooth wheels and wheelwells, with minimal gaps at the openings. Detailing matters quite a bit. Don't let air leak from one pressure zone to another through the interior, except through planned, tapered ducts.

For your cockpit, you want to create a stable, attached vortex to produce a virtual roof. Some decent examples can be found in the LeMans sports classes from the 60s, before the lads discovered downforce. Draw a shape that has a fairly flat roof the shape of your cockpit opening, and then just cut a hole. The ideal shape for the cavity is very rounded for easy, fast circulation, blended to a rounded rear edge that won't catch air, and is lowered enough to account for the growth in the boundary layer. Anything you can do to feed high-velocity air to the start of the gap will help reduce turbulence. There isn't much in the literature, but I could maybe scan what I have if there's interest.

Exposed roll bars are very nasty. Just strategic roughness helps if it has to stay round, but even a crude cardboard or foam fairing taped on can hardly miss at halving the drag. Good ones can reduce it by an order of magnitude.

The softer the ride, the less momentum lost on each bump, and the lighter the tires you can use. You can achieve suspension just by controlling the flex in the frame. The most basic improvement would be to have the batteries and/or seat on springboards. With more travel comes more potential troubles about keeping things perfectly aligned and not scrubbing, but it is easy to make big gains. It is pretty easy to design A-arms that replace the hinge with a wide quarter-elliptic spring, and then blend in to the rest of the structure without heavy transitions. I have articles on full suspension from frame flex, and on Coroplast fairing construction linked at The Car Cycle

Re: Selection of materials, from among those popular in the field.
Choose the ones you enjoy working with, and have the tools/facilities for. You can machine aluminum with carbide woodworking equipment. The exotics are wasted if you don't have craftsmanship. Honeycomb panels can be handy and light. For a one-off, I'm partial to using a lot of wood, and with epoxy encapsulation it can be excellent structurally, but the coating must be protected from even small nicks where water can enter, and from UV light. A thin layer of fiberglass gives good mechanical protection, and will tie together many wooden members.

Hey there, veteran F24 and F24+ racer (and national podium winner) here, so I should be able to give some insight. Our team gallery is here if you want to see some more photos, including construction galleries. https://picasaweb.google.com/thepodmovement/

https://lh3.googleusercontent.com/-D...2/IMG_5667.JPG

Our F24+ car. I understand the rules have changed somewhat from when we raced, but it should give some inspiration. The batteries lived behind those removable front wheel panels.

https://lh3.googleusercontent.com/-8...2/IMG_5721.JPG

Our (slightly left field) F24 car, designed as such because I was too tall to fit in a car with the batteries in a somewhat aerodynamic place, so we put them outboard to halve the frontal area vs putting one either side. I ended up outgrowing this car too and getting a pass to move on to F24+ early (I was the reason F24+ cars are allowed to be longer, heh).

The chassis design for both is semi-monocoque. Welded steel frame (old school tables with the chewing gum peeled off) with components bolted to it, but not strong enough to support the car on its own. The rest of the strength is provided by bamboo and plywood, held together with fiberglass. Incredibly stiff, and I actually completely wrote off (accidentally! They cut me up) a honeycomb aluminium and carbon fibre F24+ car in the F24 car (for a while we raced it in both formulae) with barely a scratch.

Note the design cues: an upwards taper on the rear end and high ground clearance: we wanted to get as much of the car way from turbulent air near the ground as possible. We also used cheap and available materials; steel, bamboo, plywood, household filler. In our minds the expense of high end composites, TIG welded aluminium (which we would have had to have outsourced - not our M.O) weren't worth it, we thought we could (and did) beat teams which depended on expensive materials through better design and driving. Seriously, driver training is a huge factor. Did you know that you don't have to brake into the chicane at Goodwood?

Any more specific questions feel free to ask, now that I'm not involved any more I'm happy to share trade secrets with other teams ;)

P.S.
 

Don't write off underbody flow as a mess of large-scale turbulence. While there may be a lot of disturbed air near the ground in a breeze, you can also think of the ground plane as a layer of enforced laminar flow at free-stream speed. Your own laminar surface, if present, will be tripped by the front wheel openings, spreading across at 15 deg. You have to use a slight rake to allow for the progressive increase in thickness of the boundary layer.
Laminar flow is what got the Mustangs to Berlin, but it isn't easy. You need almost-shiny or better surfaces of near-optical precision to specific curves, protected from vibration. This is not a realistic goal for an inexperienced builder, but keep it in mind. It isn't hard to get laminar flow on the 1st 10%; it just keeps getting harder to maintain on the way to 70%.

Thanks Bicycle Bob and Joe for your messages.

This is the side profile of my body so far. The top of the bodywork increases in height until about halfway which is where the cockpit opening starts and I don't have much control after that due to all of the turbulence in the cockpit.

The channel underneath as seen in the picture, starts quite late, just in front of the front wheels and has a smooth transition from the nose to the fairing. The ground clearance at this point is the maximum I can have it at and it does increase in height at the rear, producing a slightly low pressure area underneath, which is at a lower pressure than the airflow on outside. I think this helps to reduce the amount of airflow bleeding from the channel to outside the body and keeps the airflow at the rear relatively neat. Hopes this makes sense.
https://i.imgsafe.org/f40b484.jpg

The cockpit opening and the roll bar is fixed, can't do much about them.

One of the disadvantages of using tiny 305ERT wheels is that they have almost no deflection in the wheel and the smaller circumference makes it have higher rolling resistance. I have thought about using minimal suspension, but the amount of space I have inside the car means that chassis flex done correctly would be better solution.

Joe of Loath - I always have wondered what's it like to drive one of those, I raced closely against one with same design but in carbon fibre at the final of 2013 and it looked good, is it slightly more unstable in the corners or does it not matter too much? Also did you ever use or experiment with derailleur gearing?

I see a huge variety of roll bars - surely you can minimize that drag. I don't see windshields and side windows, but I don't see a rule against them. Even with the increase in frontal area, completing the cockpit with an attached vortex looks like a win to me, if done well.

One nice thing about small wheels is the reduced surface area friction, but I'd choose wheels based on available rubber. I'd also think about adding a belt to any tires I could find, and doubling the pressure. That makes suspension more desirable. You want to keep the roll center very close to ground level to avoid undue scrub. With narrow tires, camber control is less of an issue for traction, but any suspension can help or hinder cornering speeds, depending on how it is done. Solid axles are a good way to get started with suspension, but you MUST have a bend in your tie rod, giving it a bit of flex, or the Axle Tramp can wallop you.

https://upload.wikimedia.org/wikiped...SmithFlyer.jpg
https://en.wikipedia.org/wiki/Smith_Flyer

The traditional way to do a flex frame is as a 2-dimensional plate. Bamboo is a good suggestion.

http://i.imgur.com/OsfSE.jpg

Bike frames are made from bamboo with joints wrapped in hemp tape. You could also use basalt tape. Bamboo plywood in 1/4-1/2-3/4". You could CNC your single A-arm suspension pieces from this.

Deflection radially or laterally? Your cross-section shows a flat transom instead of a horn.

Only way that I can minimise the impact of the roll bar is to reduce the size of it. I can't have any fairing or aerodynamic aids on the top 150mm of the roll bar and this would include the helmet fairing. Therefore I only have 150mm of roll bar exposed above the fairing as I did on my last car. There has been trend in the last few years to make the roll bar smaller and hide them behind the helmet.
https://i.imgsafe.org/5e770f9.png

The racing is done normally on airports and racing circuits and therefore for a small car, the track is large, smooth and doesn't have many tight corners. Almost no teams run suspension, probably because it's not really necessary and can add weight and complexity to the car. Here is a video of me driving a previous car in 2013 final to show what a normal track and race would be. Also shows that only minimal steering range is needed.
https://www.youtube.com/watch?v=yKg6vsMjsU4

Nothing is allowed above the driver's eye level, even windscreens aren't allowed above that level due to this rule.
T7.4. Bodywork, including windscreens, to the front or sides of the driver’s helmet may not be higher than the eye level of every driver. This does not include front wheels or their coverings.

The 305ERT wheels would have less deflection up and down than 16" or 20" wheels and especially with my wheel which has a 28 or 32 spoke count, which is high for a such a small wheel.
Yeah, although the horn design did give a quite low drag value, my modified catamaran design had even lower drag, mostly due to its gradual increase in the ground clearance and is mostly likely my body of choice.

You could still do a lot with clear panels up to the Driver's eye level, and accept that the cockpit opening ends with a rounded obstruction. A gentle lip around the opening might establish better flow. Some F1 cars in the mid 60s met that spec, and used a clear duct- a double windshield - to help throw air over the opening.
A streamlined structure can be added just in front of the roll bar so it can draft at 1/4 the usual drag, and minimal dimensions used. If people want a handle on their roll bars as often seen in pictures, make it longitudinal, or add a finger gap to one spot on the wind splitter ahead of it. Or, maybe it can legally almost fill a gap in a continuous headrest fairing.
I am well aware of the usual wisdom that suspension adds weight and complexity. That's why I demonstrated that it does not have to, except for the considerations in the frame design, always looking for useful flex instead of making everything rigid. If you only need short travel, it gets much easier. It saves the rest of the structure from vibration and shock, even if the driver is comfy enough, and is good practice for the real world. A pan chassis lacks torsional rigidity, but that only becomes important for tuning the oversteer/understeer on a poorly balanced design.
Sharp corners on the body are OK on aircraft, but likely to cause problems in crosswind driving, causing early separation. It is easy to get a smoother curve on Coroplast in 2-D by slitting the inside skin between a few ribs. Blunt the blade tip to make it easy to control and angle the knife to keep the cut edges from jamming.
Bamboo is lovely stuff. It is six times stronger than aircraft aluminum, although it bends six times farther before failing, so it works best in a space frame. Resin and fiber lugs are a good solution to the joining difficulties.

What would you think of a 'buckboard' made from a sheet of bamboo plywood with L-shaped slots so the four corners are articulated A-arms? The width of the arm would depend on the Modulus of Elasticity of the plywood.

Those sound like very short, weak swing arms to me. If you had another set, about one wheel radius above the first, you might have reasonable A-arms. Bamboo or wood, plywood would not have an efficient arrangement of fibers to take the stress. Designs for unreinforced wood spring suspension do use much of the wheelbase for the large springs needed, so I'd be more enthusiastic about a plywood pan chassis that flexes overall, and has reinforced corners hung from solid axles.

How do you add a belt?

You sand off the tread, wind on a belt from cords producing a flat tread area, and then mold on a new tread. This isn't as good as a proper radial carcass, but it is a big step in the right direction.

Maybe somebody could convince Michelin to sell their Solar Racer tires as a regular catalog item.