Chassis Torsional Stiffness

[Scott Calabro]

Hello All,

If this is not the correct forum I apologize.

Just wondering why replica mfg's. don't advertise their chassis torsional stiffness. Seems to me it would be a great selling point. I have searched and found (I could be wrong) that only Bailey Edwards releases this figure openly on its website. They post 11,500nm per. deg. Considering that the 1969 T70 IIIB coupe monocoque came in at 5000 ft. lb. per deg. or 6779nm per deg., Bailey Edwards looks pretty good.

(Lola figures taken from John Starkey's book, "Lola T70" , from a 1969 IIIB Lola spec. sheet, ref. pg. 93)

I was expecting higher numbers on the order of 10000 ft. lbs. per degree or 13560nm

Please don't take this the wrong way ANYONE, but could it be that in GENERAL GT40 replica chassis are not very stiff, ie. the lack of posted figures. Or if they are, what are the numbers ?

Respectfully,

Scott

 

[Ron Earp]

Just wondering why replica mfg's. don't advertise their chassis torsional stiffness.

Because there is no absolute method of making the measurement. If you look around on the web you'll find race car chassis design papers indicating torsional stiffness for that particular frame/measurement mode, but they aren't comparable to other measurement modes done differently.

There have been numbers thrown around on the forum from CAV, Roaring Forties, DRB, ERA, and others. But the methods used to generate the numbers were not the same, and thus not comparable. One method would seem to generate impossibly high numbers better than current state of the art road race cars, while another method on a similar chassis design would produce figures 1/3 as much.

Nonetheless the topic generally never disappoints with respect to getting forum members in a row.

Ron

 

[Scott Calabro]

Ron,

Thats interesting, but very strange. Jack up the chassis on three points, anchor the aft two points and one fwd one, place a dial indicator under the fourth point and use a lever/weights for generating twist. Then do the math, Childs play !
I guess what I'm saying is it either takes 10000 ft. lbs of force to twist a chassis 1 deg. or it does'nt. You would think there would be a standard to separate the wheat from the chaff.
I must be missing something. If I produced a stiff frame for sale I would certainly bragg about it.

Thanks for the input,

Scott

 

[Fran Hall RCR]

Scott,
So why dont you???

 

[Scott Calabro]

Fran,
You know where I'm at. I certainly can't start a bussiness.
Lets not drift....

Regards,
Scott

 

[Ron Earp]

Thats interesting, but very strange. Jack up the chassis on three points, anchor the aft two points and one fwd one, place a dial indicator under the fourth point and use a lever/weights for generating twist. Then do the math, Childs play !

Don't let me get between you fellows!

I think you are right in that it would seem that there is "one way" to do that test. But none of the manufactuers did the test the same way to my knowledge. If I recall correctly one anchored the two aft points together like clamping down on both after points at the same time with a huge press/block assembly. One anchored all three corners independently. The latter would be different from the former with test results I think. And one I remember reading in email or post used an engine and tranny in the frame, while others didn't, another used a cage in frame.

At any rate, they've not been analyzed with the same method to my understanding (could be wrong here, just going on memory from scanning the forum since 2001).

Ron

 

[Scott Calabro]

Thank You Ron,
That sheds some more light on the subject. But still if a chassis is stiff by the numbers, a mfg. I would hope they would post it and the method used to arrive at those numbers. I know that when I buy a kit someday (chassis or complete kit) I will have to have that info before purchase.
Regards,
Scott

 

[Fred W Ballinger]

I'm with Scott on this one. Its not rocket science - hold the rear of the chassis firm across the shock mounts, twist the other end with a lever across the chassis mounts and one front corner simply supported and measure the deflection caused by a know load -

My effort went as follows

-

Chassis:
The mock up chassis has now become a prototype chassis (see photo "sf front"), and although several of the tube joints are not properly made, some tube sizes are still to be decided, and all the joints are only partially welded, I conducted a torsion test of the chassis (see photo "torsion test"). This was with the aid of a 5 metre long 6" pole and a 90kg friend.
I tested both before and after the addition of the pressed up side pods in 1.6 mm alloy, (see photo "side pod") and was very pleased to record the following:
Torsional rigidity without side boxes = 2700 ftlb/Deg
Torsional rigidity with side boxes = 5700 ftlb/Deg
The chassis mass is 96 kg.

I am using the adaptor plate as a rear bulkhead, and I was surprised to find though, when testing the torsion stiffness both with and without it, that it made very little different to the pure torsion stiffness. I used 2 diagional cross braces tacked in to represent it, the part in the photo's is a wooden dummy.

For comparison I list several other chassis figures, obtained from various sources on the internet, particularly contributors to LocostBuilders

Ultima - 3300 ftlb/Deg (coupe) and 2500 ftlb/Deg (spyder)mass 135 kg,
Lotus 23 - 1500 ftlbs/Deg, 45 kg
Lotus 7 replica, "locost book" spec - 1200 ftlbs/Deg, 82 kg
Lotus 7 replica, uprated "cymtriks" spec - 2540 ftlbs/Deg, 78 kg
Lotus Elise - 7350 ftlbs/Deg,
Lamborghini Countach - 1900 fp/degree.
Ferrari 360 spider - 6250 fp/degree.
Lotus Elan: 5,000 - lb-ft/deg
Porsche 959 - 9,500 lbs-ft/deg
Lotus Esprit SE Turbo - 4,300 lbs-ft/deg
GTD Lola T70 replica - 3300 ft/lbs per degree.

you do have to know if the test is done on :

- a bare chassis
- a panelled chassis
- with engine and trans fitted
- on a complete car

to properly compare published figures.

I would suggest that most engineers, if you talked about the torsional stiffness of a chassis, would expect the test to be done on the chassis only.

If you said the torsional stifness of the car, that would be the complete car - different thing.


Cheers

Fred W B

 

[Gregg]

Fred, after reviewing your list of manufacturers and results, do you believe a torsional stiffness rating of 32,000 nm/degree from a replica to be accurate?

 

[Scott Calabro]

Fred WB,

AWESOME POST !
Its amazing how the stressed skin enables almost double the load carrying capability before yielding. Skin both sides and I bet you get to that magic 10000 ft. lb. number.
Show us more !


Gregg,

Holy s&%t !!! Thats 23,000 ft. lbs. !!!! Thats built like the Brooklyn Bridge !!
Whose chassis is that ?

Excuse me 23600 ft. lbs. (edit)

Regards,
Scott

 

[Gregg]

Scott, that is the point Ron and others have been trying to make. If Porsche, Lamborghini, Ferrari etc torsional stiffness numbers are 1900 - 9500 lbs with all there research and development, would you consider a replica manufacturers numbers of 32,000 nm/degree to be accurate?

 

[Scott Calabro]

No... unless it was demonstrated to me. I think its possible though, look at what Fred did on his own, he's well over 1/2 the way to the stiffness of a 959.

IMHO if the data isn't published and demonstrated somethings up.
It seems mfg.'s are too secretive about this subject.
Not picking on anybody, just if you have a stiff frame, bragg about it !

Regards,
Scott

 

[Bill D]

Scott-
Why do you feel manufacturers are being potentially dishonest by not publicizing stifness figures?

 

[Jeroen]

Hello,

On the site of cav.co.za you can find a stiffness report of a chassis:
http://www.cav.co.za/pdf/Torsional_Stiffness_Test_Results.pdf

They have a result of 32154Nm/1° or 23715ft-lbs/1°.

Regards
Jeroen

 

[Scott Calabro]

Hey Bill,
Sorry for the lapse... I was working on my wife's Eurovan, anyway I don't think that its pre planned dishonesty or an attempt to defraud in any way. IMHO it is a big deal to selecting a kit /chassis.

Jeroen,
THANKS!! Thats exactly what I was looking for.

Regards to you both,
Scott

 

[Jack Dunn]

Great work, Fred. I wonder how the stiffness would improve if your entire chassis was skinned. Or by how much the bonding of AL honeycomb panels to tube frame ala Saleen S7 would improve stiffness.

Regards,
Jack

 

[Fran Hall RCR]

deleted

 

[Russ Noble]

Hey Scott!

You're improving!! :lol: You can now stir up a hornets nest WITHOUT getting political!:bash:

Seriously though, I think a high torsional stiffness whilst desirable (and I have gone to great lengths to try and achieve it in my chassis) from a practical point of view is largely irrelevant. Good for bragging rights though!

The engine bay area is one that is difficult to stiffen up. I have done what I can with mine by solid mounting the engine using the trans adaptor plate and another plate on the front of the motor. Certainly to look at it in isolation as a frame only figure means nothing. The other main area of weakness is the cockpit and this can be stiffened considerably by bracing that externally.

If you are going to consider stiffness then the only figure that really matters from a practical viewpoint is the stiffness of the entire complete finished car. A car is designed and works as a complete entity. It is pointless to consider only the bare chassis except maybe for the academic purpose of analysis. Or because some misguided regulatory body has deemed that you should.

If you are considering stiffness as it affects suspension angles then you also have to account for deflection of pivot point bushes etc. ie soft rubber v nylathon v rose joints.

There have been some chassis out there that in my opinion have been woefully lacking from a design and percieved stiffness point of view but the owners of these cars are often very happy with the way they handle and drive etc. Obviously if you are getting undesirable features like scuttle shake or doors flying open there is a problem but otherwise it is largely academic.

All the above IMHO

Cheers

 

[Fran Hall RCR]

Russ,
your post echos my deleted post....

 

[Bill D]

I believe that all the chassis by all the manufacturers are sufficient to meet the needs of 95% of all the builders.