Spaceframe Comparison

[Stew Stau]

I've been looking at a lot of photos for ideas for a semi-monocoque scratch build chassis. I have seen many claims of "full spaceframe" and "fully triangulated spaceframe" but have yet to see either.

Now I'm aware that it's next to impossible to produce a full conventional spaceframe that will fit under a GT40 bodyshell and still allow humans to enter and exit. However, of the many partial spaceframes out there, there seems to be a lot where triangulation is missed in key areas.

Here are a few of my observations:

1. Untriangulated sills - no structure to stabilise top edges of the open "hole" where you sit. Maybe other areas are worse so this is the least of the problems on some chassis'.

2. No or incomplete triangulation in the engine bay/rear suspension frame structure.

3. No or incomplete triangulation of the front suspension frame structure.

4. Little attempt to stabilise the centre tub to front frame bulkhead.

5. No triangulation in the centre tube to engine bay bulkhead.

6. Omission of load path structure from suspension points, even when such an omission is unnecessary.

On top of these observations, it doesn't seem like these points are consistent from one manufacturer to another. It seems that some have good triangulation on the front end but neglect the rear and vice versa.

I would be curious to know what the torsional stiffness of the various chassis are and wondered if anyone had any real life experience with the front/rear end stiffness, or lack thereof, in these spaceframes. The one's I've been looking at and the weaknesses I can see are:

1. MDA - Good rear end but not sure about front. Centre to front/rear bulkheads?
2. Tornado - Rear looks OK but front may be overly complicated for good load path control.
3. GTD (newest) - front and rear lacks trianguation. Centre to front/rear bulkheads?
4. Southern GT - better than GTD but engine bay/rear suspension load paths could be better transferred to tub. Front OK but still some suspension pick up issues. Centre to front/rear rear bulkheads?
5. Roaring Forties - little triangulation in the front end, looks very weak in torsion. Rear needs closer look but ??? Bulkheads as the rest.

Obviously panelling will stiffen many areas, but out of the 5, MDAs looks most promising out of the box with Southern GTs rear end being fairly simple to stiffen. Both have fairly nice pattern rear gearbox/suspension frame with MDAs being pretty close to the original pattern.

RFs front end could be stiffened fairly easily as the basic shape is simple enough but the rear could be more difficult, especially around the area that the lower trailing arm conflicts with the frame. The Tornado frame is more complex at the front, but probably better out of the box for stiffness.

I looked at the GTD for reference and it's got plenty of weak points, but some cars have done good service on the street for years. Probably a bit difficult to predict on a rough track though and suspension tuning could be challenging as the frame would be fighting you the whole time.

Anyone got and modified any of the above and why? Anyone racing them and found a particular weakness?

I know this has been covered in previous posts but they are often overly emotive and not particularly technically profficient. Too many IMHO and not enough "under X load, Y suspension point deflects Z amount."

I'm an aeronautical engineer and would appreciate people's technical input or actual test/race experience.

Better yet, would anyone with money to burn like to build an adjustable suspension rig and test them? I didn't win the lottery last night so I'm stuck.

Stew

 

[Clayton]

Hi Stew
Have you had a look at Iain's RF, http://www.gt40s.com/forum/gt40-build-logs/22164-rf-117-a.html not many places left without triangles :stunned:

 

[Dave Bilyk]

Stew,
interesting comments,
can't get a new one any more, but for sake of completeness, how about DAX?

Edit; I should say that the drawing is a work in progress, there is some lateral bracing missing from the outriggers above the fuel tanks, and maybe some I havent completed at the base of the windsreen, but it all the main chassis bracing is included.

regards
Dave

 

[Stew Stau]

Yes, I really like his RF. I don't really consider it an RF though as the chassis is effectively a one-off built around the RF base. Lovely machine though, I want one for my garage.

I didn't include the DAX as I hadn't even thought of looking at it. Probably should have though. Cheers.

Stew

 

[SouthernGT40]

Hi Stew

I would be interested to know what you were actually going to use a car for ie just serious track use, road use, the odd track day etc in view of how stiff it sounds you want the chassis to be.

Regards Martin

 

[Stew Stau]

Martin,

To be honest, I'm a bit of a geek as well as a CARnnoisseur and I like to set up my cars and drive them hard. If I could afford it, I'd be doing what Iain is with his RF 117 down under.

As it is, I built my Striker because it was the 7-esque car that won the most 750 Club championships in the '90s and I wanted the best chassis I could afford. I used it for track days as well as road work, but loved the fact that I could adjust the suspension for a particular bit of road and appreciate the performance change. I could have got a Tiger, Westfield or a Caterham but the fact was that Sylva dominated Kit Car racing so that's what I went for.

After only 500 miles, I threw away all the rubber bushings in the front suspension and rear panhard rod and replaced them with the hardest poly bushes I could find. I didn't like the feel of the suspension flexing on turn in and when I was pushing it, I found that the little movement I felt as the rubber bushes compressed on the rear panhard rod made me nervous. After the change, the amount of info the car fed to me on the twisty bits went up 10 times and my lap times reduced purely because I was now more confident that the car was doing what I wanted it to.

I guess I believe that as all GT40 spaceframes are of roughly the same size, weight and function, I want one with the tubes all doing their job properly and not just adding weight or flexing like a noodle under bending loads.

That's partly why I'm looking at a semi-monocoque. Fabricating a scratch built mono is possible with enough time and effort, but a hybrid is likely to be the faster way to the same performance.

I don't do track days often enough (especially after the trouble my wife gave me when I threw my bike down the track 3 years ago), but when I do, I don't like to compromise. I would rather drive a fully rose jointed race car on the street than a soft and compromised road car on the track. I want to be able to feel when the tyres are starting to let go or go off and not wonder if it's the suspension or chassis flexing.

I guess I could have simply said that it doesn't matter what I do with the car, but how I want it to feel and respond that counts. But as I've typed all that other stuff, I'll leave it.

If Iain decides he doesn't want 117 any more and donates it to my care, I'll probably get what I am looking for, plus a bit. I'll take off the rear wing, and probably modify the front splitter so it looks more authentic and then apply some late '60s paint scheme and drive it to lunch on Sunday.

And when my wife complains about the noise, hard ride and difficult entry, I'll take it as a compliment, smile and explain that if she wants a compromise, to follow me in her Jag.

Stew

 

[Stew Stau]

Martin,

Sorry, I forgot to compliment you on your car. Looks like it's coming along well. I look forward to seeing it on the road.

I'm moving to Benson (south Oxfordshire) in June so won't be a million miles away from Southern GT so might pop down for a visit.

Stew

 

[Martin]

Hi Stew,
I have a later MDA and I liked the look of the way the chassis was put together. One of the main reasons for choosing the car actually. You won't be able to get a new MDA chassis now as MDA are not trading anymore but if you want one I know someone who has one for sale.

Martin

 

[SouthernGT40]

Stew
A visit to see Mick Sollis would be well worth the effort.

Martin.
ps I am flat out on my build and would like to think it will be drivable August or September.

 

[Martin]

I agree with Martin, take a visit to Southern GT. You won't be dissapointed.

Martin. Can't wait to see the finished article. It'll be worth the wait.

Martin

 

[Russ Noble]

Stew, you are not going to get definitive, comparable chassis deflection figures. They just don't exist. Neither are you going to get an uncompromised production chassis.

You've done the reading and the looking now you've got to decide which chassis most closely approaches your ideals. It is subjective, it is not quantifyable, it is in your humble opinion. And, if you must, modify it to suit. Or build from scratch as many guys do because they are not happy with the factory offerings.

WRT RF and DRB they have to meet tough Aussie regs and testing for torsional requirements so no doubt there are some consistent data comparisons for those. RCR I think have met these with their mono also, but as far as I am aware no other GT40 manufacturers have. IMHO (there we go again!!) RF or DRB would be a good base for you to modify to suit your requirements.

As I guess would any chassis, it's a matter of comparing the total cost of a particular chassis plus the cost of the mods that you would want to do to it. Somtimes you are better to just start with a blank canvas and a few lengths of tube and make it the way you want it.

 

[Mike Pass]

Hi Stew,
As you say none of the current frame cars are fully triangulated but there is nothing to stop you turning rectangles into triangles and adding in the odd strut and brace here and there. There also seems to be a big difference in torsional stiffness when a roll cage is added even though the full effect can't be realised with the door tops cutting into the roof area.
As you like triangles (as I do) - here is a pic of a Gilbern chassis I made a few years ago.
Cheers
Mike

 

[Derek]

My question is this: If shear panels of appropriate grade and thickness are bonded and riveted to the chassis, does this triangulation become redundant and weight adding? My plan is to bond and rivet aluminum panels (top and bottom) to the square tubing with rigid foam epoxied in the void space. The vertical reinforcements would be 18" on center. It would not be difficult to add the diagonals, I am just wondering if it is worth it? Goals are pretty common, maximum rigidity with minimum weight.

 

[Martin]

Hi Stew,

Apparantly the MDA Chassis could still be manufactured if you wanted one as the fabrication company still has all the tooling.

Martin

 

[David Sturch]

Stew,
I think any one of Dave and Mikes chassis would do. Mike's chassis look's like it's built like a tank. Once panned his chassis would be stiff and also heavy. I think if you were run a big block(428) in it you would have no problems. You can see my Chevron Chassis PDF on my build. 3/4 " crew tube 3/4 " round .065 wall in most places. It is still shy a few braces. The finished product should look like this.
Dave

 

[Old Roaring Forties]

Hi Stew,

for what it's worth some info on the std RF road chassis (without ally panels - contrary to one competitor's lies to prospective clients trying to talk the RF chassis down, they are there to reduce air and water ingress and seperate vehicle sections, not to add to stiffness)

It achieves a little over 12kN / deg in torsion at the wheelbase through the suspension pick up points which is the correct way to measure it. That is without the roll cage. We have not measured a current chassis with a cage - before you ask !

Soon there will be some additional triangulations in key areas but that figure is already twice what a modfied V8 is required to achieve in Australia.

The RF being mild steel RHS we do not expect that torsional rigidity figure to deteriorate as quickly over time as some other materials might - a lot also depends on design and load paths.

Regards Paul

 

[Iain]

Hi Stew,
I'm not sure if you are conversant with "focal or focus " points in chassis design. It may well have been around for some time but over the last couple of years they have started to appear in our V8 super cars. A point mid way between the pick up points, (front and rear) suspension points is identified as the focus point and triangulated back to the chassis, for a very modest addition of weight you can multiply the rigidity of the chassis in both areas. On my car it is up under the dash board at the front .The biggest challenge is then successfully connecting the front end of the car with the back end ! As Mike Pass pointed out the usual benefits drived from a full roll cage are diluted by the door/entry set up, Russ Noble has to my knowledge been the only member brave enough to attack this properly and I admire his flexibility. We opted for the "intrusion bars" which will have helped tie the two ends together. We have also used the paneling, where appropriate to provide additional rigidity, bonded with wildly expensive Aircraft glue and double rivets on the diagonal, we have also double skinned the chassis where possible as mentioned by Derek.

Between the chassis work and a solid mount engine and gearbox I beleive we have acheived the position where the chassis flex is no longer an issue in the car, we have made some progress with the suspension set up in the car but I now understand why it takes 250 people to put a Ferrari on the grid and still they cant win.....shame

Iain

 

[Stew Stau]

Paul,

I take it you meant kN/m per deg. Unless my French to English conversion is rubbish, that's 8830 ft.lbs per deg. I must admit that I find that figure astounding.

Unless it's a mis-quote or mis-calculation, thats a very impressive stiffness from a partial spaceframe and I must admit that I'm surprised. A Winston Cup car in the 12000 ft.lbs per deg area is considered pretty good. Not that they're F1 or anything, but pretty stiff non-the-less. What's an Aussie V8 Supercar chassis like?

I take it the additional triangulation you'll be adding is in the front end.

Iain,

I'm familiar with the concept. Although my Sylva doesn't have such additional triangluation, I know of several similar cars that use the same idea to transfer suspension loadings into a torque loading on the chassis, particularly on the front end.

I must admit that I found that bit of triangulation interesting. If it went to a triangulated bulkhead the loads could be transferred into a mono or spaceframe tub structure running down the sides of the car, but on a spaceframe GT40, that's quite a difficult engineering problem.

I've been sketching out another solution, but there are as many ways of skinning the cat as there are engineers working on the problem I'm sure.

D_LMP,

Panelling can take significant shear loads, that's how a monocoque works. However, the larger the area they span, the harder it is to stabilise them, hence the stiffening ribs etc in many chassis panels. A lot of bonding is flexible, negating the benefit. Rivetting can work, but a non-flexible bond is probably most effective for dissimilar metals.

However, seam, puddle or spot welded steel on steel is the easiest to predict structurally, however, you can end up panelling to the point where it's no longer a spaceframe and then you're into mono and semi-mono construction.

Dave,

"Built like a tank" is fine for tanks, but not for sports cars. The problem with many chassis designs is that they're over-engineered from a pure strength assessment, but lack in stiffness. What is required is to remove all redundant structure so you end up with a chassis that only transfers the loads required and the rest is made up of air (light).

A properly designed race chassis would likely suffer fatigue failures over a long period of use, however as it is designed to win and then be replaced by the next evolution, that isn't a problem.

For a performance car that will last, a compromise on weight must be made to ensure that the structural loading in normal use is well within the fatigue threshold for the material used and the load cycle expected during it's life.

All,

Thanks for the responses. IMHOs from an engineering viewpoint are useful. It's the beer fueled "I heard it down at the pub" opinions that I was referring to originally.

Stew

 

[Old Roaring Forties]

Hi Stew,

yes it is impressive and it's a torque so kN.m/deg really (I left out the meter). The members are fairly substantial and load direction is quite favourable when analysed carefully. My role before RF was to test the cars (and still is) for compliance to Australian Design Rules and one of the tests we did as a supplier to RF1 was torsion and beaming - although not strictly required it's good peace of mind to know an ICV is not going to twist or bend too much !. It is also confirmed within 300kN.m/deg (experimental error, chassis/chassis variation ?) by an independant engineer here who did the sign off for RF1 prior to me.

Current V8SC are over 45kNm / deg but as recently (?) as 1992 the good ones were only around 12 to 14. They made big jumps when introducing bars through the floor section around 2002 ish era.
Many sports road cars are well in excessive of 12kN.m/deg now too albeit mainly for crash and NVH requirements.

As for Nascar I guess each category and class need to have chassis which are suitable for use with the tyre construction they are using and work within the class rules. Most formula at the top end like these use compliance within the chassis to work with the tyres and the rest of the suspension as a system. How often do you see cars in the non professional classes with either missplaced, over or under weighted members and or cages tied into abstract points and way over stiffened in areas not attached to nodes and pick up points ?

In a recent FEA analysis we identified areas front and rear that could do with improvement rather easily without adding excessive additional mass. We are not focussed on this because, as you suggest the chassis is well stiff enough for it's purpose. If anything we will look at weight reduction in the future and belive we can do this without compromising the current figure.

T&B is a massive subject that we don't consider to be a weakness in the car at present.

Regards Paul

Paul,

I take it you meant kN/m per deg. Unless my French to English conversion is rubbish, that's 8830 ft.lbs per deg. I must admit that I find that figure astounding.

Unless it's a mis-quote or mis-calculation, thats a very impressive stiffness from a partial spaceframe and I must admit that I'm surprised. A Winston Cup car in the 12000 ft.lbs per deg area is considered pretty good. Not that they're F1 or anything, but pretty stiff non-the-less. What's an Aussie V8 Supercar chassis like?

I take it the additional triangulation you'll be adding is in the front end.


Stew

 

[Old Roaring Forties]

btw we have never pretended to take any credit for what Iain and Mick have achieved with RF117 but we are obviously pleased they chose an RF as a starting point !

Ross N does an outstanding job in the RF he has modified and developed himself with help from Will C but the chassis itself remains largely unchanged.

I think it is a good starting point and will only get better.

Sorry to take the focus off the general thread but I thought I could clarify and contribute to the RF specific points.

Regards Paul