Deflection Figures for a GT40 Chassis - Page 2

Ron Earp · 07-30-04, 11:13 AM

Fantastic post, a real gem to have here in the forum database. I'm suprised that more people aren't excited about it. I for one am going to give this a go in the fall sometime. This is the sort of stuff I like, and I'm sure many others do to, some technical information that is well-founded and accessable.

Now, given some of those figures and we know 1 newton meter is 0.737 lb ft, then from a previous post:

"To quote the DRB brochure over 7000 Nm per degree
The roaring forties state that at over 20000 Nm of torque the deflection was 0ne degree."

Then the DRB is around 5159 ft lb and the RF is at 14740. Seems like a big difference so I wonder if there is an error somewhere. What are the figures for the mono replicas? Originals?

JP · 07-30-04, 11:39 AM

Ron,

There does seem to be some big variation. Derek Bell measured his GTD chassis from his Lola T70, it was about 3300 ft/lbs per degree.

Most GT40 chassis are quite similar in design, so its quite interesting to see such big differences in rigidity.

Testing the chassis yourself is a great start, but I'd prefer an independant test to verify any figures.

Regards,

J.P

Hemipanter · 07-30-04, 12:19 PM

Kalun
Concerning Nascar cars, theese ar very special to say the least. I almost question if ther is anything that could be tranfered for use on our cars. Ok, this was a little hard expressed.
The cars are asymetrical, use stagger tires, and a very special shock absorber setting. That may force them in to another chassie build up. All this because of rules. The car has an given ride height, and by twisting the rules, the shocks could be set to take care of this. Which means that they dont really perform its work fully as a shock absorber.

What I am saying is that one is newer sure why a race car is using what it is using. I havent looked into TR numbers of those cars latley, so this is just a note
Kind regards
Goran Malmberg

Ron Earp · 07-30-04, 12:50 PM

[ QUOTE ]
Testing the chassis yourself is a great start, but I'd prefer an independant test to verify any figures.

[/ QUOTE ]

I think testing the chassis ourselves is the best way to get answers, don't you agree? While I like all the manufactuers, I'm an independent scientist and objective, or, are you implying that you don't trust me? [img]/ubbthreads/images/graemlins/shocked.gif[/img] Finding an independent shop to test chassis would be hard, and it would be even harder to have all the different cars there for testing. Any mono figures?

JP · 07-30-04, 01:56 PM

Ron,

What I meant was that it would be nice to know that the figures "we" get, are in fact correct. A bit like having your work marked by a teacher at school, just to be safe!

I'm sure you know what I mean. [img]/ubbthreads/images/graemlins/smirk.gif[/img]

Regards,

J.P

Ron Earp · 07-30-04, 01:59 PM

I think I trust Kalun_D quite a bit with the procedures and engineering.

Lynn · 07-30-04, 03:09 PM

Goran,

NASCAR is on to the trick of shocks that make the cars fit the templates (body height, shape, etc) and then collapse down at speed to make the ride height much lower to add down force, etc. They are compensating-- I am guessing by adding the equivalent weight of the down force at speed-- for the static nature of the template checks now.

Lynn

Kalun D · 07-30-04, 03:46 PM

""Concerning Nascar cars, these are very special to say the least. I almost question if there is anything that could be compared(sic) for use on our cars.""

I agree Goran, Nascar figures may not apply for a GT40, the only thing they have directly in common is the tube frame.

I also am not sure one way or the other if you can have too much rigidity or not.

Hemipanter · 07-30-04, 04:54 PM

I may fill in that a full scale test is quite tricky to get 100% accurate. One easely create bindings that disturbes the measurment.
I would say 90-95% accuracy, if one have some background or understanding of these thing, by using private made garage
tools.

Goran Malmberg

Ron Earp · 07-30-04, 05:00 PM

You can defnitely have too much rigidity for motorcycles. There are plently of examples, but with frame rigidity started to go sky high in the late 80s and early 90s lap times increased in many cases. Then, the "black art" of strategically altering, cutting, and chopping frames to reduce rigidity started whilst the engineers started to understand how chassis flex helped the rider put power to the pavement and increase cornering velocities. Kevin Cameron from Cycle World has written some nice bits about this since he was an engineer on said cycles and understands it well. Don't know if this applies to cars or not though.

Hemipanter · 07-30-04, 05:17 PM

Lynn
Ohlin shocks explained the valving trix a while ago. But they didnt get to specific about it, for some reason.

Kalun
I have not seen anything about to much rigidity. In fact those figures is raising all the time. There has been debates among racers, claiming so and so, but no real explanation what should make "not so high" better. And I cant figure it out myself either.

Ok, flex may take care of stress loads, but this does not concern chassie control.From our standpoint there will be no problem going as stiff as possible, without beeing to stiff.

Goran Malmberg

Hemipanter · 07-30-04, 05:33 PM

Ron
A bike is not subject to twisting in the same manner as a car that get wedge, diagonal load during cornering. A dragster is using chassie flex, but is not subject to very much twisting force. Other than a tremendous engine TQ.
I have connection with ISR Mc of Sweden as I use their brakes. I will make some inquire about this. I also know a few speedway frame builders.
Goran Malmberg

Trevor Booth · 08-01-04, 05:26 AM

Another simple and accurate method in attached photos.
A note of caution. one degree of twist may permanently set a light chassis. A reliable method to establish applied load (L) Mass of vehicle with fuel oil and occupants(W)
L=W*0.25*wheel track (widest). I use a digital level and measure setup angle front and rear, apply load and measure deflected angle front and rear. Deduct set up angles front and rear and the difference is angle of rotation. To obtain torsional deflection at intermediate points use digital level or dial gauges(under chassis sides).

Trevor Booth · 08-01-04, 05:40 AM

Photo 2. Load is applied to RHS end see saw at 1.5m or 5ft to make calcs easy. Load is applied downwards.

killroy · 08-03-04, 12:21 AM

That looks a lot more complicated than what I have seen.

The method I use is clamp a tube to an immobile object (3,000lb welding table) then add a weight at a specific distance and measure the deflection of the tube. Next, clamp one end of your chassis to the immobile object, support the other end with a tube or angle iron under the center of the chassis. The round tube or angle iron will create a ~knife edge hinge with zero chance for binding. Lastly you mount the tube and weight to the end of the chassis supported by the hinge and measure deflection.

Now math, calculate an angle from the total deflection minus tube deflection, then calculate the torque by the weight and the length of the tube (measured from chassis center line). Since a chassis is like a spring, the torque devided by the deflection angle (in degrees usually, unless you like radians) will give you a stiffness number.

That number is close enough and does not have any deflection from turnbuckles (fake shocks). It's a measure of chassis stiffness, the axles and other suspension parts do not count and should be removed if possible.

Trevor Booth · 08-03-04, 05:11 AM

Not complicated at all,and very accurate. If you look closely the see saw has rollers at its base to allow the chassis to translate as it is rotated. The chassis will adopt its own axis of rotation (radius of gyration centre). Supporting the chassis on a knife edge causes a friction component preventing the chassis from adopting its natural axis of rotation. All measurements are taken on the chassis frame thereby eliminating any jig movement and suspension movement. The "fake shocks" are merely a means to support the chassis and apply the load. Start to finish takes 4 hours and can be done on a bare chassis or a complete car. The methodology was developed by Queensland University of Technology circa 1991 and has proven accurate and reliable.

JP · 08-03-04, 09:46 AM

It looks like either method will work. Coupled with some accurate scales one could really get starting point for shocks springs without having to guesstimate.

killroy · 08-03-04, 07:38 PM

If your chasis is light enough, my method can be done in about four minutes. The biggest problem is setting up the chassis on the stands, but if you are measuring a 70 lb space frame like the chassis that I was measuring, that isn't very hard.

A knife edge hinge is just about the lowest friction hinge that you can make, I would bet that the friction would be a lot less than the bolt through a tube that you are using.

We were actually using a roller stand that is a tube that rolls freely on ball bearings and it held from below by an adjustable stand. They work great for sliding a long tube of 4130 in to a cold saw. Since the chassis rolls on top of the tube and the tube is free to roll in the stand, there is no chance for any binding. A piece of angle iron might have just a tiny bit of bind if the chassis wants to move sideways when it twists, but very minor when you consider the ammount of deflection that is being measured.

By the way, I only said fake shocks so that people would know what I was talking about when I mentioned the turnbuckles.

Steve Toner · 08-03-04, 09:28 PM

Just to put this all in perspective, here's something Carroll Shelby said at the SAAC-29 banquet:

"We kept the wheels on the ground on the old AC chassis, because it had 700 foot pounds of flex and it bent so much that we just put wider tires on it, and we beat Ferrari for the world championship. I can't believe how crude you can get to be successful."