Am I missing something here? Cymtric has posted a thoughtfull analysis of the chassis increacing stiffness by 200% up to 2500 ftlbf per degree translated into SI thats about 3400Nm per degree.

But that's rubbish compared to current touring cars with around 25,000 Nm.

Admittedly they are tintops, and heavier and with a higher Cof G thus needing to be stiffer but even so surely the locost chassis could be made at least 5 times stiffer than it is?

Anyone got access to software that would calculate the effect of a roll cage or hoop? Better still anyone tried fitting a roll cage or roll hoop and measured the stiffness?

It would probably be better to compare a locost with a "normal" car to get an idea of the torsional stiffness. A touring car is a highly specialised vehicle, and the roll cage in them undoubtedly provides a great deal of the increased strength given the complexity of them and the fact that if memory serves correctly there is only space for the driver the rest filled with roll cage spars.

ifi can remember correctly i saw an analysis on the locost chassis which raised the torsional stiffness to around 9500 ftlbs per degree just by adding a roll cage. thats nearly 12000 Nm isn't it?

not bad for a tiny car, you could get up to that figure, 25000 me thinks, but is it worth it?

as pre cimtrix cars seem to work ok, is this not all a bit pedantic?

thats gotta be less than or approx 4% of the siffness of a touring car, and it still works!

atb

steve

Cymtriks info doesn't take into account skinning the car with ally/stainless steel panels. no doubt you could put more tube into the chassis to make it stiffer but whether you'd ever reach the stiffness of a touring car or notice any difference in handling compared to the extra weight of more steel in the chassis is highly debatable.

Ned.

[Edited on 1/9/04 by ned]

Liam where did you get the info on torsional stiffness with a roll cage fitted? Thats exactly what I'm looking for - not that I don't believe you its just that I'd like to check out the figures and see how I can apply them to my chassis.

Whilst the chassis as is may well be OK it should be a lot better, Ok my example of a touring car was extreme but not as much as you may think. The old BL 1800 land crab was well up there so it must be possible.

It is vitally important to get as stiff a chassis as possible as a flexible chassis will mask and even negate changes made in spring and roll bar rates. Just because something works dont mean it's the best after all.

With a flexible chassis, which this appears to be, when you change something you might or might not get the opposite result from what you expect.

For example my old TR7V8 is stiffer than this and its an old design - so why bother with a locost if this is the best than can be done?

a locost will still handle better than a tr7v8.

Ned.

try here: http://locost7.info/mirror/chassis.php

lots of info on chassis stiffness from other countries where they have to run finite element analysis etc etc

[Edited on 1/9/04 by ned]

quote:

---

Originally posted by stephen_gusterson
as pre cimtrix cars seem to work ok, is this not all a bit pedantic?

---

Very true- its a fifty year old design thats been proven ........

having said that, if we didn't debate things at the risk of being pedantic, then we wouldnt have much to talk about!

What you tend to find with FE analysis on components especially chassis' is that the result totally depend on two factors, both of which do not change the design of the chassis.

The way the chassis is restrained and loaded in the software (choice by the user) makes a significant difference to the results. Many discussions have been had about the best way to restrain the chassis and load it, the discussions usually end up with no real conclusion being drawn.

Then when you compare it to the actual chassis you have as Ned says to compare like with like, which is hard because of the welded joints, aluminium skins etc. Also can you accurately represent the same load/restraint conditions on the actual chassis?

What FE is actually more useful for in my opinion is the comparison of one chassis design to another, using the same load/restraint conditions. The actual numbers it throws up can vary for the above reasons, but, its useful to know what the scale of improvements of possible modifications might be.

For RGB next year we are contemplating a space frame chassis with a difference, and we would hope to stiffen the chassis by 4 to 5 times at the same time maintaining or reducing the weight. If we do it and it works well at the end of next season i'll tell you all about it.

Kind regards,

Darren

Wesley Linton has written a thesis analysing the Luego chassis - this is posted on the net somewhere and has been posted in previous threads on this forum.

He measured an actual Luego chassis, then modelled that chassis for his FEA. Then he made changes to the model.

Went from 1352Nm/deg for a mass of 120.1Kg and an efficiency of 88g/Nm/deg, to 6448Nm/deg for a mass of 148.3Kg and an efficeincy of 23g/Nm/deg.

Adding a rollcage gave 9152Nm/deg for a mass of 170.0kg and an efficiency of 18.57g/Nm/deg.

Aussie builders have to perform a physical torsion test - perhaps some of them could post their results with the "Aussie mods".

quote:

---

Originally posted by v8kid
"Am I missing something here?"

Hmmmm!



"Whilst the chassis as is may well be OK it should be a lot better.."

Why? It's worked perfectly well for hundreds of builders up til now!



"The old BL 1800 land crab was well up there so it must be possible."

Helllooooooo! Thats another tin top. You cant compare them.



"For example my old TR7V8 is stiffer than this and its an old design - so why bother with a locost if this is the best than can be done?"

Don't. Go and fix up your Triumph. As you say, its stronger.

---

The tortional stiffness of the chassis has to be proportional to the vehicle requirements.

The reason that there are hundreds of cars out there designed to the book without any problems is because its fit for purpose. i.e. a bit of yeah haaa when its sunny.

If you intend to race however, its more of an issue. the structure becomes way more important be cause of the effects it can have on suspention / wheel geometry.

However, I believe that if you have made everything as stiff as a BTCC motor you have over engineered your car and incured a weight and cost penalty for your trouble.

There are any number of factors effecting how stiff your chassis needs to be. Vehicle weight, CoG, Lateral G capability, brake, metalastic bushes in the suspension, etc. The list goes on.

1st decide what you will be using the car for, then work on the stiffness. Also try to bear in mind that most (if not all) of the analysis carried out was undertaken on a bare chassis frame. You would be very suprised at what addition strength the body panels add.

Silex: your "building:" message has pricked my interest, do tell more

Well my friends miss-fortune was to my benifit. Last winter he wrapped his Jap import twin turbo Nissan Skyline round a rather large tree. Not wishing to cripple his insurence he agreed to sell me the motor for 500 quid, so thats basically the donor car.

The engine and gear box and brakes will be the main items I will refit. I have stipped this of all the fancy electrical junk as much as is possible.

I have only just started to lay out the chassis in steel its mostly still in the cad package being tweeked.

The chassis is being made from 48mm diameter tube - wall thickness 1.5mm. I know that this may sound a little flimsy, but the steel is a wee bit stronger than you average tube. Standard tube is normally S275 steel, mine is S500 which is almost twice as strong for the same amount of steel. It also makes for a light weight but very strong structure - good tortional stiffness.

The car itself will be full bodied, there is no way on earth I can shoe horn the engine in a 7 style body.


Hope thats peeked yur interest a little further.

Thanks Ned, Bryan those links are exxactly what I was looking for theory and test results to back it up.

Sorry if I upset the Luddites..

Nothing beats a TR7V8...in a straight line

quote:

---

Originally posted by v8kid
Nothing beats a TR7V8...in a straight line

---

Yes, touring cars are very stiff. They are also very big and weigh a lot compared to the Locost. There is also the issue of diminishing returns. The now defunct kit car mag "Car Builder" published an article on chassis design in which testing of actual chassis at STATUS was referenced. The guideline for stiffness was to match the stiffness in ftlbs per degree to the car weight in lbs for a road car. Double this stiffness for a sportscar. These are minimum stiffness values below which the suspension will have to overcome significant flex of the chassis in addition to the contours of the road.

My improvements to the Locost chassis are not intended to prove other builders or designers wrong or to give definitive numbers. They are intended to give a set of simple and realistic modifications which an amatuer builder can easily make to the book chassis backed up by FE analysis.

I wrote a comparison of my results and Wesleys on this site a few weeks ago. A lot depends on which questions you ask and how you use the answers.

I don't think that my mods are poor. Consider that 2700ftlbs per degree is a lot higher than most spaceframes manage. The book chassis has half this figure. GD state a figure of 2750 for their spaceframe backbone chassis Cobra replica. Ultima Coupes apparently have 3300.

I completely agree that my suggestions are theoretical until tested on a finished chassis. Alan has a valid point here but this does not mean that my analysis is wrong. In fact my analysis of other chassis agrees fairly well with other peoples analysis and with actual test data. This implies that my Locost mods are likely to be a good guideline.

Check out Wesleys thesis and my analysis write up on the lowcost7.info site.

So thats where the 6kn/degree figure comes from!

Assuming the car is being driven energetically it's very likely that 60% of the weight will be transferred to the outside front wheel - for a beefy 1000kg V8 thats about 5.8kN or approx 1 degree of twist in the chassis.

It makes a bit of a nonsence of setting up the suspension accurately and yet the locost seems to go quite well.

As I said at the beginning I seem to be missing something here but I'm blessed if I can see what it is.

Can anyone reccommend user friendly cheap FE software to play with?

quote:

---

Originally posted by v8kid
So thats where the 6kn/degree figure comes from!

Assuming the car is being driven energetically it's very likely that 60% of the weight will be transferred to the outside front wheel - for a beefy 1000kg V8 thats about 5.8kN or approx 1 degree of twist in the chassis.

It makes a bit of a nonsence of setting up the suspension accurately and yet the locost seems to go quite well.

As I said at the beginning I seem to be missing something here but I'm blessed if I can see what it is.

---

I stripped the right bottom ball joint on my car while cornering left so I think there are a lot more things to think about than the loading on the chassis when you look at stresses. The loading was on the front wheel but the chassis and wishbone held together so something else let go. I think you create some problems by overthinking others.

quote:

---

Originally posted by Peteff
I stripped the right bottom ball joint on my car while cornering left

---