How Stiff is Stiff Enough??

TEAM78 · 2 Nov 2007, 17:35 (Ref:2058067)

Just doing some research on chassis stiffness and sadly there isnt much out there,
My questions are:
how should a torsional stiffness test be done?
and what sort of value should see?

There appears to be two different ways of doing the test, both dont simulate a cornering load. Also regarding how stiff a chassis should be the values appear to vary hugely upon the vehicle and its purpose, for example a Nascar is 30,000Nm/deg
Super tourer and DTM appear to be around 40,000nm/deg
F1 car I dont know and A Formula Student car 1500-4000Nm/deg,
My work is on a saloon car but would be interested to hear some more numbers on this and approachs peoples use for testing and improving chassis stiffness,
As we all know theres a trade off between stiffness and weight and it boils down finding the least worst solution.
I wouldbe interested in anyones experiences in this and what they found out

terence · 2 Nov 2007, 19:31 (Ref:2058135)

Back in "94" I was involved in the building of John Clelands Cavalier,the brief was to get the car as light and stiff as possible over the previous years effort. The end result was the Torsional Stiffness came out at 32.600nm,the shell weight came down by 17%,unfortunatly I dont rember the overall weightbut we did use lead flashing on the passenger floor to bring it back a bit!!.End result=Cleland BTCC Champ in 95.

We had a 20 ton flat bed that the shell was bolted down onto.Various measuring devices were placed around it,we then secured a very stong 3 metre bar to first the front of the shell and four of us hung off of it!We then did the same to the rear.All measurments were recorded and fed into the computer which gave us the info.

phoenix · 3 Nov 2007, 18:41 (Ref:2058742)

A Ferrari F50 has a torsional stiffness: 25,500 ft·lbf/34,570 N·m per degree.

phoenix · 3 Nov 2007, 18:52 (Ref:2058746)

Lotus Elise (7420 Lb-ft/degree) and Corvette (9720 Lb-ft/degree)

Goran Malmberg · 3 Nov 2007, 20:15 (Ref:2058804)

The webmaster of this site is not to happy for the members to name their home sites, but concerning twisting tecnique a few image is sort of nesseary.
Anyway, fare down on my site there i two images showing a twiting test of my new Nordic Supercar Corvette build. Sorry to say that the final tet numbers are not yet known since the engine is not installed. Numbeer will come in a few weeks.
How stiff a chassis do we need?
Stiffness i dependent on chassis weight and suspenion wheelrate in N per mm (Hz number). Well, what really matters is shock absorber setting as the shocks is what is very much dependent on a chassis that i not flexing. If the wheel travel is longer and the schocks and spring are softer, the demand on chassis stiffness becomes less. We could say that the wheel travel should be seen is in proportion to "chassis flex travel". Therfore stiffness numbers will rearly be seen as an average recomendation. An average 2500 ib street sportscar sprunged to 1,5-2Hz will do fine with say 10000Nm/dgr. While a 2000 ibs 5Hz heavy groundforce car will need 40000Nm/dgr. A weekend track sports car could manage fine using 15-20000 Nm/dgr.
Hope I dont upset anyone using simplified numbers here in order to give some breif idea about something hard to specify.
Regard
Goran Malmberg

TEAM78 · 4 Nov 2007, 19:56 (Ref:2059526)

Some interesting numbers there.

thanks guys
So far the way I see a test should be done is completely fix one corner then other side on same axle give that one degree of freedom laterally, then on the other axle only restrain one corner vertically and the last corner is completely free to apply the vertical load. and of course replace the spring/damper with a solid piece of steel work.
really interesting stuff on the Caverlier terence,
I remember that year 95, awesome car and I liked what they did on the front turrent with the tube coming from the base of the A piller straight up onto the turret.
As for gussets in a cage there benefit appears to be very small if there are quite a few bars present anyway. Im useing a speedsheet to figure out how stiff the sheel needs to be but there again appears to be 3 different techniques for this as well which all provide different answers as they consider things in various levels of detail.
most roll cage companies appeared to be very low tech in their knowledge and approach but a few seem well clued up.

moonman · 6 Nov 2007, 20:19 (Ref:2061480)

How about chassis stiffness in formula ford chassis. Can less torsional stiffness generate more grip or do they build them to be as stiff as possible?

johnny yuma · 12 Nov 2007, 02:14 (Ref:2065340)

Good question given the FF's agile nature.However as an early post said unless the chassis is stiff enough ,the damper and antiroll-bar effectiveness will be unknown at different parts of different tracks where dips,bumps,road camber changes etc are giving anything that can bend a workout.

The Model T Ford was quite twisty in the chassis and this helped its "go anywhere"abilities---but I don't think that translates onto the racetrack !!!

JamesH · 13 Nov 2007, 17:05 (Ref:2066678)

Chassis stiffness test are legal requirement for home built cars in Oz. Locosts are popular there, and have to be modified over original plans to get the stiffness to the requirements.

Worth finding an Aussie who's got through the test for information on what it involves.

James

retro_msport · 23 Nov 2007, 08:46 (Ref:2074014)

we messed about once .. and all un-scientific .. but we took a Mk2 escort and loosend all its 6 point cage bolts ... it was sitting one 4 chassis stands with 3mm and 1 mm shims ontop of those .. the stands where at the ends of the chassis legs, we then removed shims, we took 5mm out of the rear before we got a gap with someone holding the opposite corner down. We then leveled it again and bolted the cage back up, again its a only a 6 point cage (no roof or pillar attachments) and tried again 3mm before we had air .. I then checked a car with a multi-point weld in, again with a Mk2 Escort and on removing a 1mm shim, we had a 1mm gap. Now thats without any bars and men hanging from them.

Also when starting the build of the Escorts i place the shell on 6 level points .. very rarely does the car sit on six points, they after all are 30 years old, its not uncommon to see 12-15mm gaps. The soon settle once i've cut the bits i dont want out of them though

Is it true that the standard Mk1 Ford Focus was a class leader in torsionally stiffness .. i belive the figure was 40,000 Nm ..I may be wrong, but havnt body shells come a long way since the Mk1/2 Escorts

johnny yuma · 25 Nov 2007, 07:06 (Ref:2075042)

Interesting ,when I drove and sat in a Focus on a long trip I was impressed with it's stability and lack of handling vices.Didn't know it had such a stiff chassis.What was less impressive was sitting in the back seats-a road hum coming right into your ears seemingly from behind the seat.Maybe too stiff,or got the wrong grade of rubber bush in the rear suspension ?

Years ago a Mk 1 Cortina gave a better ride/handling package for all occupants than any Japanese sourced 4. How things have changed-maybe !

THR · 30 Nov 2007, 23:38 (Ref:2079074)

The stiffer the chassis the more u can control it with the dampers and springs!
simple.
Stiffer is always better, in all things, i had no compliants yet