We do zero model making, all CAD now. We used to do a lot of physical prep work but find CAD is so much easier. Vindaloo was built in the time honoured fashion of putting a wheel at each corner, connecting them and then filling in the rest. Classic 'fag packet' drawings were as complex as it got.

90% of the car was drawn in CAD, including a body that was run through a virtual wind tunnel (with very limited success!) It was particularly useful for designing the steering controls and front uprights including disc brake mounts - shame the girls didn't get the setting out right when they brazed it so we had a dragging brake and, shall we say, an "interference" fit with the disc. The 10% we didn't draw were things like mounts for bodywork, brakelight, wiring and seatbelts. And the 200 or so rivets we used when the glue holding our chassis skin on came apart.

In terms of physical models we built a "coffin" to set the controls and cockpit size out - nothing complicated, just a couple of broom handles screwed to three pieces of ply with two pieces of wood forming a seat propped in it. Somehow though it didn't quite work so this year we'll be cutting the front off the chassis and moving the front bulkhead forward by 100mm as the drivers were sitting too high.

I'm going to revisit the virtual wind tunnel this year to try and get some more useful numbers out of it - particularly looking at a flick at the start of the cockpit and the rear chassis shape. At the moment I get lots of reasonably predictable streamlines but numbers that don't make much sense and are quite variable run-to-run.

we go for the make it up as you go aong approach or talk about the opttions and then do something completely different

Have any of you used Cfd yet ? and if so have you found it beneficial at the speeds involved 25 - 30 mph . Models can help to visualise a concept (what are you using as a manequin Brian ?) . Have any of the Greenpower cars used full CFD / CAD / FEA before cutting metal, as talking to teams at last race it seems that most investigation is done after the car is made ?
Sorry to be so curious but some of the design solutions I saw seem at variance to the success or otherwise of the car
Tigris

On our first car - which was a Greenpower kit, we created designs on paper using measurements form profiles along the cars frame length and simply thought logically about air turbulence. From our designs various styrofoam models were built until a reasonable shape was agreed upon - we have not used CFD modelling to scientifically prove our design yet. Even with CFD there are so many factors that you will have difficulty in accounting for that the final design may not perform as it was calculated to do so. Many cars appear to have found a fluid shape and do perform particularly well on the track;. The CFD will be helpful but the compromises you will probably have to make during construction will have a greater effect on the final cars performance.

Phill

Good point Mr T
I've always thought that these computer " simulations " only really help if you already have a known datum ie car to compare your modifications to . In other words building a new design from scratch using only computer tools may not actually give you a good car ( 2010 Virgin f1 car !! ) As i said there are some very quick "barn doors" out there .

As we use plywood for all the major structural members in the cars we design and make( everything apart from rear & front axle assemblies & rollbar) we can easily use card and paper for major construction decisions. Its easy for our Yr 8 & 9s to realise their sketches and designs and gives them some idea of the complexity (& feasability) of their designs. If they can't make in card & paper then they will find it even more difficult to make it for real, let alone begin to be able to design something on computer. We have to think slightly differently when designing given the materials we use. We use 2-D design to model dimesions, geometry and check for interferences once we have some final designs. We try to keep all curves non-compound and look at combining parallel sections and simple cones to get curvy shapes when needed. We also look at forces and load pathways early on and integrate these into the designs as they develop by doing physical tests on the card models. As far as aerodynamics go we don't have CFD facilities here but are hoping to build a scale wind tunnel to test new models in over the next few years. Once the kids get a handle on these basic processes they seem to be much more independent and are able to use CAD for designing at a higher level in later years.

Brian & zebedee were designed in 3d CAD, not in great detail, for example Brian's chassis was made completely different to the original CAD design, materials shape, everything, but cross sections from the CAD were used to make the mould for its body. Zebedee was made quite accurately to the CAD design. After Zeb was made we worked out how to model/simulate in CFD. Next year's car and some big solar panels are being CFD'ed now (or soon), so it has the now fashionable long tail ;^)
Motor, gearing, performance simulations etc. all modelled using XL
If you're using CAD the first and most important thing to model is the driver.

Would have to agree that the driver size & position determine the layout of a major portion of the car. Made the mistake with our first Eco-Marathon car of designing in Sept, racing in July - drivers grew in the intervening months = cockpit too small, new drivers needed= disappointed kids

One, life size chassis mock up out of bamboo and glue to go to the welders

Our (steel) chassis was built within about 3mm of design by using wooden jigs, I'd like to think we could do that every time. Having said that, the first part I always make is the main roll hoop as it's the most difficult, and we got it wrong. I measured it, updated the CAD model and could adjust the front roll hoop so we kept the 2" rule.

In terms of CFD I use it for looking at changes - the only thing we used it for on the car this year was the nose shape. Too low and the air all goes over the car and across the cockpit creating drag, too high and there's drag from forcing it through the small space under the car. We also found it better to have as short a nose as possible so the tail can be as long as possible.

I'd much rather use a proper wind tunnel as you get real values - for instance we investigated taping the bodywork seams and any exposed rivet heads at the final and it knocked at least 10 seconds a lap off. Unless you draw every last rivet and use a small enough mesh size, CFD isn't going to tell you that...

With RR7 the team made 5th scale models out of foam. 3 or 4 separate models were made by different team members and tried in UWE's wind tunnel to find the best. With RR8 several rough designs were produced using 3D CAD (Blender was used) and tested in the, web based, VWT we had developed and used later in RR7's development. As mentioned above a rough model of a 6 foot person was used to work around. The shape that looked best and seemed relatively easy to make was drawn using 2D-CAD to get the shape and position of the main bits and printed onto a large piece of paper. The main tubular frame was bent and cut to fit this. Much of the detailed working bits were "designed" as the car was built to fit this frame with 3D CAD being used for the steering and rear subframes and cardboard/paper templates used to check and cut the material. Pro-Designer and later Pro-Engineer was used as this was used in the school and hence the pupils know how to use it. So it was a mixture of 3D-CAD, 2D-CAD, paper and just making it to fit for us.

we start with the roof box. then simply try to fit everything in, Simply....

First version was too wide and far too short- tricky to fit it all in. Then version 2 was longer but Oh so Narrow. turning circle is Huge...

we are using more models on the new car but mainly CAD

@myton - by huge, how large are you talking?

i have no idea. we lose patience and lift the front to turn the car around.

if interested i can measure the angle of the steering wheels. our first car could turn on a penny then we saw under the 'hood' of ypod and alex educated us on how small the turning angle needed to be. hence the new car barely turns the wheels.

Leaning may turn the car quicker than the steering :)