Vehicle frame in torsion

[pontiacjack]

Hello all. I usually lurk on the vehicle-driveline forum, but came over here to ask about torsional rigidity (or lack thereof!) of vehicle frames. Here's why: I'm totally rebuilding my '59 GMC pickup and was alarmed to discover (accidentally) that the chassis (no cab or bed yet mounted) has very little torsional rigidity. Over a small range (up to two degrees twist) it measures in the ballpark of 125 lb.ft. per degree (and is fairly linear in this range). This low number is apparently by design, not from degradation- it had almost no rust, has no cracks, and rivets are complete and tight. It's admittedly not a substantial design- open C-channel rails and no X-member. But here's my problem- poking around the Internet for torsional moduli (competition cars and hotrods), I find numbers ranging from 3,400 to 7,500 lb.ft. per degree!! (the low number is from a 1954 article!)

I'll appreciate any comments which will help me "get a grasp" on this huge discrepancy...

 

[BrianE22]

I believe a substantial amount of torsional rigidity comes from the body.

 

[NormPeterson]

From the body.

Parallel open cross sections such as channels and angles are very flexible in torsion. Closed section body features such as the sill structures and 3-D effects such as through the roof structure as where virtually all of the x,xxx ft-lb/deg comes from.

I think some of the aftermarket frames of rectangular tubing construction are of similar x,xxx ft-lb.deg torsional stiffness all by themselves.


Norm

 

[Tmoose]

Boxing the [ can move the decimal point over 1 place.

X members can help, but the holes for driveshafts and exhaust have to be beefed up aplenty.

http://weldingdesign.com/blodgett/no_twisters_swingers_allowed_0501/

 

[TheBlacksmith]

These vehicles had low horsepower and weren't meant to be driven fast on interstates. Frame flexibility probably helped with the rough roads and farm tracks. Even the 60's muscle cars had fairly flexible frames and they stiffened the springs to improve handling. It is only recently that engineers had vastly stiffened the frames and forced the suspension to take all the deflections.

 

[pontiacjack]

Thanks for responding.

I was sort of hoping someone would have some familiarity with actual procedures for measurement. For example: are these numbers typically obtained by twisting the frame over a significantly larger range than the two degrees travel that I used?

Tmoose- If I had discovered this weakness prior to media-blasting and powder-coating the frame, I definitely would have boxed it. The chassis has way too much "clutter" to use an X-member. I did, however, bolt X-shaped members (steel straps in tension) to four rectangles formed by frame rails/crossmembers behind the engine (one 'X' is both top and bottom of frame, other three are top only). If you can believe it- the frame measured only about fifty lb.ft./degree before this addition!

 

[Tmoose]

You may find those bolts retaining the straps loosening regularly.

Similar to what B-Smith said, If it's intact and sound it will serve its intended trucky functions as well as when the vehicle was new, which was perfectly adequate even if not particularly refined.

 

[GregLocock]

FSAE teams write a paper on testing their body for torsion. 2 degrees is probably enough.

I'm surprised the bare chassis is that low, a hand calc for a typical C section in the longitudinals would show you whether that is reasonable.

Although it would be relatively easy to stiffen it up you will be adding stress raisers, which will need to be patched in their own right.

Unless you have some desperate need to improve the handling then I'd leave it alone.




Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[NormPeterson]

I suspect that having to cope with 500 ft-lb or so of engine output is another criterion here.


PJ - if you could tie the top and bottom straps together with a web or some sort of truss arrangement you'd improve the general torsional stiffness substantially. Particularly if you can get some beam depth in your other X-fabrications. I agree with Tmoose with respect to difficulties keeping your pinned connections pinned precisely in place. Bolts with their shanks in closely fitted holes and no threads in the shear plane or the metal would be a better detail structurally for your purpose.


Omer Blodgett/Lincoln Arc Welding Foundation has enough discussion concerning design for torsional loading in "Design of Welded Structures" and "Design of Weldments" for what you're trying to accomplish.


Norm

 

[GregLocock]

Look at the ladder chassis in a 55 Tbird. That is about 10 times as stiff as your figure, and uses the same sort of technology.

I ran an FEA on one and found it to be a very balanced design, that is to say, upgrading any individual member resulted in a lower torsional frequency. It was possible to improve the torsional frequency but large gains relied on better joint efficiencies, and using closed tubes not C sections.

Here's the sort of thing

http://www.jalopyjournal.com/forum/showthread.php?t=271673

The X in that frame is crucial for the torsional stiffness. the two cross members are probably closed sections, and their joint efficiency into the main rail is probably low and important.







Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[dicer]

I am fairly familiar with those Blue chip and task force trucks.
Others mension that the body structure adds stiffness, maybe so with a car but not so much with these trucks. How did you do the test to arrive at that figure?
One of the reasons for wanting the flexable frame is these trucks were designed for use on unpaved roads, around farms etc. if you have a stiff frame and desire to haul a nice load, you run the chance of breaking something, even the newer pickups don't have a very stiff frame. I think its a trade off of vehicle weight vs frame stiffness. To stiffin it up your going to have to add more material, especially if you don't want it cracking or popping rivets or welds. And is one of the reasons a newer pickup weighs more than these older ones, and then of course can haul more too.

 

[NormPeterson]

At least the body on those trucks needs to be sufficiently stiff such that the fixed glass stays in the window frames and the doors don't fly open when you're out driving on that uneven terrain. Contribution to total general torsional stiffness would depend on the body to frame mounting details.


Norm

 

[patprimmer]

My understanding is that the springs are hard to carry a load, the chassis is soft to distribute shock loads, the cab and tray are independently mounted to the chassis so they don't really reinforce much and the tray and body obviously move independently of each other when you watch the gap between them when traveling.

The cab is reasonably rigid compared to the chassis and I expect is mounted to the chassis so as to isolate chassis flex from the cab a bit.

The doors etc stay in place as the cabin is more rigid.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[dicer]

The cabs are mounted to the frame on brackets in front of cab that extend I'm guessing (too lazy to go measure it), about 10 inches from the frame rail. The rear mount points are closer to the frame maybe 3 inches or so. And the for and aft separation is maybe 4 feet. There are rubber cushions (washers that are about 3/4 inch thick that the cab sits on. Its is bolted tightly to them. The pickup body also bolts directly to the frame.

 

[patprimmer]

As the pick up body has no closed top, and a poorly tied in tail gate it should be somewhat lower in torsion that the cab with a fixed roof closing the 6th side of the box and door locks and hinges that are reasonable at tying in the two sides.

The mounts between the cab and chassis will reduce the stiffening effect the cab has on the chassis.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[pontiacjack]

Glad to see a good discussion- that's all I was really looking for- hoping to learn something from my observations. Thanks.