Chassis rigidity

[FASTDAD]

Hi Guys,
well I’m about to embark on this project to upgrade and improve this chassis. It’s a Carbir DS3-99 chassis no. 003. I’m going to check the torsional stiffness to make sure that it’s a great foundation on which to work with chassis setups, don’t need a flexaflyer!

My questions are about chassis stiffness. What are some good numbers to have when checking this, be it in degree or inches using dial indicators? I’ll be replacing the dampers with a pair of strut rods on the front with the upper & lower control arms & hubs installed. I’ll be coming off the hub wheel mount surfaces with a lever (maybe 10ft long) to add weight.

On this type of chassis (I’ve done Late models in the past) where is the optimal placement of dial indicators? Under the chassis at front bulkhead, top on the rocker arm pivot etc?

The back (roll bar back)is a whole different situation. Here there is a defendant signs of some movement, both on the upper valve cover mount and the angled brace running behind the headers where it’s bolted to the engine block. Since I know for sure about issues here, should I tackle this first to before making any decisions about the front?
I’ll be bolting up the engine, oil tank & gearbox before taking any measurements.
Thanks for input, I’ll appreciate knowledge
Todd

[S Lathrop]

I have done torsion tests on many of the cars I have built and a few that were built by other manufacturers.

The first car I tested was a Zink Z10 and it came out at 1200 ft. lbs. per degree. The best tube frame car I have tested was about 8000.

I do the tests on fully assembled cars. That way I have a number that I can use in calculating setup numbers. I replace the shocks with solid struts. At the rear I have stands that bolt to the hubs and support the car on the ground At the front of the car I use a short 3" piece of 1/8 inch angle iron and I place the angle iron under the front bulkhead. I use an old engine block as a stand under the front bulkhead with the angle iron as the pivot.

At the front of the car I have 2 structures that bolt to the front hubs and support a 10 ft beam above the chassis. The beam bolted to the struts above the front hubs and it extends to one side to give me a lever to twist the front of the chassis, through the front suspension.

To measure what is going on when I load the beam, I have a lengths of square tubing that are clamped to the far side of the frame and have a dial indicator set to measure the movement of the tubes as the frame is twisted. I place the tubes at the front axle center line, the dash hoop, roll bar bulkhead and the rear axle center line. I can calculate the deflection of the chassis at each bulkhead relative to the rear axle centerline which is the stationary point of the chassis.

I also have a dial indicator available to measure movement between chassis members as I hunt for problems.

Just as an aside, I have modeled my frames in SolidWorks. The torsions numbers for the SolidWorks simulation and the actual car are withing 10% of being the same.

One use for the torsion number is to calculate the spring rate of the torsion bars, especially rear relative to front. The softer the chassis, the lower the spring rates of the sway bars.

.

[FASTDAD]

Hi Steve, thanks alot for the info. It gives me a place to start with using good information. As this progresses I'll keep things posted as to what I find out and procedures I use. Any other info, ideas, how not to do it, is appreciated. I use seat of the pants engineering since I'm not very good at math, so I build projects to see if it works! Anyhow, thanks again, have a Merry Christmas, onwards and up from here!
Todd

[J Leonard]

I think there was info posted on this a few years ago with procedure and photos. Buy S Lathrop maybe?

[Jon Jeffords]

I think there was info posted on this a few years ago with procedure and photos. Buy S Lathrop maybe?

I recall this as well.

[2BWise]

I do the tests on fully assembled cars. That way I have a number that I can use in calculating setup numbers. I replace the shocks with solid struts.

I agree with this point here. The frame stiffness alone doesn't give the full picture. I do my tests fully assembled as well and measure the stiffness between the hubs. That's where the tires experience the loads. If there is a stiffness issue it doesn't give an indication to exactly where that lies but I have used it to track down engine mount failures.

[FASTDAD]

That's what I'll do also. I'll put all the major components back on the chassis and check that way. I'll need to make a few additions to the jig to give adequate support to the floor to get good measurements. I looked for Steve's thread about this but came up empty handed. Very nice that he reposted it. It's a great starting point. Thanks for input.
Thanks
Todd

[Rick Brannon]

I think there was info posted on this a few years ago with procedure and photos. Buy S Lathrop maybe?

Yes there was this’s from my archives.

[DaveW]

https://www.apexspeed.com/forums/sho...ss-Measurement

[Rick Brannon]

https://www.apexspeed.com/forums/sho...ss-Measurement

I read and digest all of your and Lathrop/Pare posts lol.

[FASTDAD]

Thanks for the link Dave! I agree with Rick about reading the posts he highlights. Thanks for all the info. Now it's out into the shop time to put all this into action.
Have a great Christmas guys!
Thanks
Todd
(ps. need to brush up on my math! Coming up with the numbers is the easy part for me, it's what to do with them after that)

[S Lathrop]

As you increase the chassis stiffness, suspension friction will be more critical.

You can measure suspension friction quite easily. Push down on the chassis to see how low the chassis will stay. Take a ride height measurement. Next lift the chassis to see how high it will stay. Take a second ride height measurement.. If you take the difference in ride height and multiply that difference times the spring rate of the suspension at the wheel, you will have a number that is how much you have to change the load on the corner to move the suspension.

I have seen numbers as high as 40 pounds. That is not a big issue if the car is not very stiff in torsion but will make the car very difficult to drive it the as the chassis gets stiffer. I look for numbers less than 20 pounds per inch for chassis stiction. It is tough to get there.

[DaveW]

As you increase the chassis stiffness, suspension friction will be more critical.

You can measure suspension friction quite easily. Push down on the chassis to see how low the chassis will stay. Take a ride height measurement. Next lift the chassis to see how high it will stay. Take a second ride height measurement.. If you take the difference in ride height and multiply that difference times the spring rate of the suspension at the wheel, you will have a number that is how much you have to change the load on the corner to move the suspension.

I have seen numbers as high as 40 pounds. That is not a big issue if the car is not very stiff in torsion but will make the car very difficult to drive it the as the chassis gets stiffer. I look for numbers less than 20 pounds per inch for chassis stiction. It is tough to get there.

The friction is, of course, a force (lb, etc.), not a rate (lb/in), but Steve is correct that it's tough to get that low in friction with normal rod-end, damper, etc. friction. But if you pay very careful attention to all friction points, it's possible to get below 10-lb of total friction from up-to-down in that measurement. It does make a difference in grip and tire wear.

I measure it at the damper and use the spring rate x damper travel up-to-down to calculate the friction force there. Then, if you prefer, you can use the motion ratio to see what it is at the footprint. The object is to minimize that number wherever you measure it.

[BLS]

I read and digest all of your and Lathrop/Pare posts lol.

No kidding. An education right here in the comments.

[DaveW]

The friction is, of course, a force (lb, etc.), not a rate (lb/in), but Steve is correct that it's tough to get that low in friction with normal rod-end, damper, etc. friction. But if you pay very careful attention to all friction points, it's possible to get below 10-lb of total friction from up-to-down in that measurement. It does make a difference in grip and tire wear.

I measure it at the damper and use the spring rate x damper travel up-to-down to calculate the friction force there. Then, if you prefer, you can use the motion ratio to see what it is at the footprint. The object is to minimize that number wherever you measure it.

As a case study, when we first went to pushrod front suspension on Citation FC's, there was a serious understeer issue. The car just skated in the front and didn't want to turn. It wound up being the friction issue discussed above. Once that was resolved, the car had great front grip.

[Lotus7]

When trying to decide on what is an acceptable ‘target’ chassis torsional stiffness, remember that all that matters is the “spring rate” at the tire contact patch. Hoosier suggests that their tires are around 1000 lb/in, so if your chassis is 40,000 or 50,000 it makes little difference (look up ‘springs in series’). Others will argue …. YMMV :-)

[FASTDAD]

So do I measure that with the dampers in place and pressurized or make a holder for the springs and don't use the dampers? It seems to me that the pressure in the damper could have a large effect on that dimension, or no damper & spring and strut at all?
Thanks
Todd

[Lotus7]

Yes, the dampers are part of the functional "wheel rate" while on track, but if you put chassis torsion test loads through them they will bottom out and probably be damaged. Steve Lathrop defined a great method for testing chassis torsion in post #2.

[2BWise]

So do I measure that with the dampers in place and pressurized or make a holder for the springs and don't use the dampers? It seems to me that the pressure in the damper could have a large effect on that dimension, or no damper & spring and strut at all?
Thanks
Todd

You need to manufacture dummy shocks. Solid bar with rod ends on each end that fit in place or the spring/damper.

There is an SAE paper out of Cornell from around the year 2000 that covers chassis torsional stiffness versus suspension stiffness. The rule of thumb has always been something like 10x stiffness, but the Cornell paper does a great job of putting some data behind whether that is enough or overkill.

[FASTDAD]

I can do this. 2BWise; "You need to manufacture dummy shocks. Solid bar with rod ends on each end that fit in place or the spring/damper."
Thanks
Todd

[DanW]

I can do this. 2BWise; "You need to manufacture dummy shocks. Solid bar with rod ends on each end that fit in place or the spring/damper."
Thanks
Todd

Not only that, you can use the dummy shocks to set your ride height, castor, and general alignment settings without having the inconsistency of "stiction" of the springs and dampers at each corner.