Specifically I am looking for rotational stiffness (ie. twist) of the trailing arm due to the tire contact patch resisting sideways sliding of the car. I'm only interested in forces greater than 1G during cornering.
I already know what happens but go ahead and take a guess.
I'll post pictures of my test fixture shortly.

Chris the trailing arm appears to be a pretty stout piece. Im interested in your info, and I could maybe model a trailing arm on a FEA program that I have..

lemme find the FEA output I did for my 911 roll cage

WAG:
1 deg @500 lbs.

I'm assuming you're trying to twist it.

What exactly do you mean by "1G". I'm sure you don't mean a force equal to the weight of the trailing arm. Do you mean something like a force equal to 1/2 the weight of the car with a moment arm equal to the radius of the tire/wheel combo? Are you measuring deflection of a rigidly monted trailing arm, or are we including bushing flex too?

By the way, this is a cool thread. Can't wait to hear about your experiment. Good stuff.


-Ben M.

Chris,
Do you intend to test before and after boxing?

I'm guessing that most of the flex happens in the tube where the pivot shaft sits, not in the box section of the arm.

--DD

My guess is it has flex designed into it at an amount equal to slightly less than the tensel strength of the material from which it was made. IE it will bend before it will snap like an airplane wing

So far JP is the only one really on the right track.
You're only off by a factor of two.
What I did was mount the trailing arm pivot shaft in a support fixture I fabricated. Then I welded the base plate of the fixture to my 800lb welding bench.

Then I attached a 5ft piece of structural tubing to the side of the trailing arm where the bearing retainer bolts on.

by putting a 200lb man's weight at the end of the structural tube I can simulate a 1000lb twisting force on the trailing arm.
By my rough calculations this is somewhere in the ball park of what happens when a slick shod race car goes around a corner at 1G+. Keeping in mind that the contact patch is roughly 1ft away from the axis of the wheel and the slicks allow a car to corner at close to 1.5G.
The digital level is on the structural tube to measure the rotational deflection of the trailing arm in 0.1 degree increments.

I had a friend sit on the end of the tube and saw the level move by one full degree!

I have several more tests planned already.
1) measure a trailing arm with the standard box kit installed.
2) measure a trailing arm with my idea of a better stiffener. *
3) measure the fore/aft deflection of the arm by applying the load 90 degrees to what I did for this test. This will simulate what happens under 1G of braking force.

* My better stiffener will be very simple, lightweight, and will allow me to make a camber change to the trailing arm at the same time.
I just need to make another fixture so I can weld it together with the correct geometry.

Put the level on the trailing arm, not the lever

let all pieces be at rest, zero the level, given the odd profile of the trailing arm.. maybe some foam or balsa wood will be needed to make a suitable platform for the level

the reason I say this is there is deflection/movement in the plate that you bolted to the bearing keeper

In all testing and modeling are good... 1 deg at 1G wonder what they were designed for ?

and yes I see the angle you have welded on the botton of the setup

....go on....

So can you tell with your set-up which area was/is most susceptible to torsional loads? e.i.- where would "beefing it" by most effective?


p.s. - someone mentioned FEA software earlier in this post. I'm still in the process of honing my CAD skills, and found a company that has a "free", limited (can't import CAD files, but has a descent CAD program attached), 300 node version of their software for engineers. I downloaded it a couple of weeks ago, but haven't had time to mess with it too much yet.

Free FEA software

The 300-node version is about 2/3's of the way down the page.

Chris,

This is exactly the kind of testing required to provide a benchmark by which to compare against. My custom trailing arms took so much effort fabricate that by the time I was finished, I was frankly too burned out to continue on with making a test fixture. I'm glad you're doing it!!

Unless you intend on making the stock trailing arm your "standard", you'll otherwise need to qualify your fixture. If you don't, then your data will be flawed (unfortunately). You'll need to first determine how much the test fixture deflects under load. Don't get me wrong, cause it's always seemingly easy to blow holes in someone elses work, but that's not my intention. Continue on, and do share your results; I remain very interested in your work. Are you going to test to fail?

I think if you examine the stock trailing arm design, I think you'll find it quite good (from a design standpoint). The design of course, was dictated by production requirements, but it is still very good. The area I never liked, was the pivot shaft and how it ties into the chassis mounts. Think about the camber and toe adjustment. You must deflect the inner ear in order to move it around. I've seen where some try to re-inforce the inner ear, only to have it crack. In my opinion, the inner ear is just crying for a spherical joint or conformal structure of some kind......sorry, guess I'm getting a little OT.

Andys

If you mount the level in various locations on the trailing arm...gobs of modeling clay would work....then load the lever arm....you'll be able to factor out any mounting plate flex while determining where the actual flex is occurring.

QUOTE

If you mount the level in various locations on the trailing arm...gobs of modeling clay would work....then load the lever arm....you'll be able to factor out any mounting plate flex while determining where the actual flex is occurring.

Yes..

but I wonder how much the chassis yields?

chasing the dragon..

very cool

Put that level at the in board mounting in the direction of the pivot bolt and see what happens.

Chris,
This is what the pro teams call "off season testing".

P

I am confident that wouldn't change the test results Tim.
I have a good sense of structural integrity (for lack of a better term).
What happened was exactly what I predicted. Pretty much ALL of the movement occurred in the box section of the trailing arm. It was very easy to observe this during the test. The movement was not linear along the length of the box section either. Slightly less deflection was observed toward the front where it is wider.

The fixture did not deflect.
The structural tube/plate connection to the trailing arm did not deflect.
The pivot shaft housing did not deflect.
The tube and forging section of the trailing arm did not deflect.
Very minor movement of the outer pickup bracket was observed. This bracket was a reinforced stock unit. I also have stock and my custom brackets I can test.

The loads I imposed were well below the elastic limit of the materials of the trailing arm. It returned to the exact starting position after multiple repeated tests. I had no plans to test to the failure point but that is not a bad idea. Next would be to have two men lean on the bar.

and you told me you were busy...

My research is primarily aimed at Production race cars that receive custom modifications by me. With that in mind I intend to make a fixture which will allow me to measure stock trailing arms for damage, and will allow me to make minor alterations to the relationship between the pivot axis and the axle axis, ie. static camber angle.

When I raise the pickup points I use a 3/16" plate for the inner ear that is much more resistant to cracking than the stock sheet metal, yet allows for changes to the pivot shaft angle. The plate can be bent minutely if desired by striking it with a BFH. The range of adjustment necessary is very narrow so the stock system continues to be suitable for my application. My goal is always sophistication in planning and simplicity in operation and appearance.

QUOTE (r_towle @ Dec 16 2005, 10:34 PM)

and you told me you were busy...

I had exhaust system parts cutting in the bandsaw the entire time I was working on this.

There is no way that box would flex before that mild steel piece of 3/8" flat plate would deflect. Put rhe level on the arm.

QUOTE

I am confident that wouldn't change the test results Tim.

I never said it would change the results...

Im just giving some insight based on what I do.. sometimes I get to go and strain gauge a bridge..

I see a couple between where you have the level, and the c/l of the trailing arm assembly, I also see a source of error there, but this is moot

Your testing process and methods are sound.

You mentioned the trailing arm rotated one degree.

We can learn where to reinforce the trailing arm, instead of welding all that heavy factory reinforcement on

good stuff

Im going to measure a spare trailing arm I have and model it.. It will be neat to see how it compares with your testing

One more goal I have is to possibly be able to tweak traling arms that have been slightly bent from hitting a curb for instance. At some point they become deflected beyond repair but I think some will be restorable.
Erik had one change between 1 - 2 degrees of camber with no toe change after he slid across a curb at Mid-Ohio when he spun going through turn one in qualifying. Turn one is taken at about 90 mph.

QUOTE (Racer Chris @ Dec 16 2005, 06:24 PM)

You're only off by a factor of two.

Hay, it was a WAG, not a SWAG.......
I would have had to charge you for a SWAG

It's half inch plate Rick. The box section of the trailing arm is only sheet metal.
Believe me, I am very careful with testing like this. What appears to be crude is very well thought out and my observations are made thoroughly.


You are correct but the stiffness of my components was chosen to be well beyond what I needed for these tests. The structural tube is 3x3x1/4"w. The gusset was included to spread the applied force to all four bolts at the mating surface. The trailing arm is resting on a single point along the weld bead beneath the round tube section and it can shift freely at that point if it wants to when the load is applied.

Chris, all my observations are made based on what I do for a living.. "test things"

you have it all covered..

having said thay you could take your invesitgation to another level, and find where the trailing arm needs to be reinforced..

if you think the trailing arm need to resist torsion more, there is a real easy way to do that, and that is without welding on that trailing arm reinforcement stuff

OK, agree with your load guesstimate.....if about 90% of the weight is on the outside rear tire, so....
Do a bit of rerigging so you can do the test with bushings in the trailing arms. I suspect that most of the plastic bushings
will cause a goodly amount of deflection.
If you push out 1 deg of neagtive at 1G as it sits, bushings would take another 1/2 deg, me thinks.

Good job Chris. I have meant to do this but I had no interest in keeping the stock trailing arms though. Here is what happens when you exceed the point of elasticity.

dfgdsgs

I spoke to a GT2 racer a Sebring once (actually more than once) and he said he cut a 1 3/4" hole thru the trailing arm and welded in a piece of tubing. The tubing stuck out of the arm about 3/16" on both sides, just enough to get a good weld around it. Said he could tell the differance next time out.

Sounds easy to me.

heheh

Paul thats correct

use tertiary reinforcement, make that big trailing arm box section into smaller individual sections...

my thoughts were other than a tube.. but its the same in the end

Tim,
If I had a strain gauge laying around the shop it would have been on the trailing arm. Then moved to several locations for a real thorough analysis. I have a very good idea of what will work to cut down on the torquing of the trailing arm. I think about 4 ounces of sheet metal is all I need to cut the flex by more than 50%.
One problem with the factory style stiffening is it reinforces in the wrong way, by doubling up on material thickness. I don't want to prevent the arm from bending when it is hit by another car that drives into the side of me. I only want it to limit the flex from cornering loads.
When you see what I do you (not you Tim) will be surprised at the simplicity.

This test was done with my Delrin bushings installed. The only thing holding the trailing arm to the fixture was the ends of the pivot shaft. There was no observable movement at the tube that supports the bushings.

I have no interest in abandoning the stock trailing arms.
That thing is quite a pretzel. How'd you do that?

Wrapping it with sheet metal does help to a degree, but you really need to get inside and tie the sides of the box together, so to speak.

Paul

QUOTE (TimT @ Dec 16 2005, 11:27 PM)

my thoughts were other than a tube

Yep...
I'm planning to kill two birds with one stone. Can't do that by just inserting a tube.

QUOTE

but you really need to get inside and ties the sides of the box together, so to speak

OK, that my $.02 worth, I'm going to bed.

I expect to see a full report on my monitor in the morning.

Paul

great thread.... I love this stuff..

Im out also.. Ill check back on the rebound

QUOTE (URY914 @ Dec 16 2005, 11:38 PM)

I expect to see a full report on my monitor in the morning.

I won't have more until late afternoon. I have to do more tests in the morning, then watch an SCCA race at 1 o'clock.

I can't wait to see if my thoughts about stiffening the arms match with yours. Thanks for all the effort Chris. When are you going to make a similar jig for the chassis?

QUOTE (URY914 @ Dec 16 2005, 09:21 PM)

I spoke to a GT2 racer a Sebring once (actually more than once) and he said he cut a 1 3/4" hole thru the trailing arm and welded in a piece of tubing. The tubing stuck out of the arm about 3/16" on both sides, just enough to get a good weld around it. Said he could tell the differance next time out. Sounds easy to me.

this is really a great thread!


I can't wait to see the test results of a stiffening kit equipted swing arm.
I'm dying to see if the commonly thrown around statement about it being "a waste of time" is actually hogwash.

I have it on good authority from AJRS, that it actually is important to stiffen them up. We have the braces in the chassis to the mounting points and I'm told that after stopping the console/ear flex it because necessary to stop the swing arm deflection.


Paul, regarding the guy with the tube welded into an arm...
ha... we put a tube into our Longitudinal for the same reason.
This is supposed to be a big reinforcement over a stock longitudinal for little or no weight... (not that I'm good enough to feel the difference!)

brant

nice work Chris.....

QUOTE

If I had a strain gauge laying around the shop it would have been on the trailing arm.

come on Chris, you've got to have one of these sets to use to measure deflection

Mike,
I have a dial indicator and mag base but not all those tips. I'm not really a machinist after all.
Fine instruments aren't necessary to measure the trailing arm flex anyway. You can watch the movement and see exactly what is happening.
A stock trailing arm definitely isn't stiff enough for a competition car running on slicks.

I saw one of the 1997 GT1 chassis on a deflection dyno at the Wiessach facility, and they had around 40 dial indicators at different places on the chassis. Forget about any photos, strictly verbotten (sp), and they probably would have shot you on the spot.

Al Swanson has run this part on a computer modeling program where he works (can't tell you that either ), and found that welding the so called stiffeners actually weakend the the part, primarily due to annealing the steel.

The other consideration would be the bushings used. IMHO any stiffening would be mute with anything less than solid bushings, for obvious reasons.

I've always thought the tubing through the trailing arm would be the best way to go.

Now you are seeing the reason why many professionally developed racers are using a different, custom trailing arm. You can bandaid the stock part or you can build a much stronger yet lighter custom part.

The only reason to keep the stock trailing arm is a class rule that says you can't change it.

For a race car I would completely eliminate the bushings. Run a monoball or rod end. Look at Blakewell's car and Sheridan's car.

Here is Blakewell's setup. This is a Preston built chassis. He has a custom rear trailing arm but I don't have a picture of the arm.

I have'nt seen this thread until this morning, but wanted to add some comments.

Chris and I started talking about this because som eof the info we have found for 914's seems at times to be untrue. People have done things for different reasons and many are not known. Such as The outer (stock) pick up point piece is extremely flexible, Chris has a stronger piece, I used it it works well. Why some people use a trailing arm stiffening kit and others do not. Here are the things we want to know:

-How much does a standard trailing arm twist in degrees of deflection when we load it near max load that we might see with slick under racing conditions.
. Only flaw I can see in Chris's methodology is maybe we have under or over estimated the load at the wheels. It also has a lever that would act upon the trailing arm, the Wheel. 11.75inches from the centerline. So we really need to make sure we've thought that through.

-How much lateral deformation when a max cornering load is applied. Meaning doe the arm bend thowards the transmission when we load it and if so how much.

-Where does it bend or Twist

-How does a trailing arm that is stiffened using the typical weld on plates perform to the same tests.

-Where can the arm be properly strengthend while adding minimal weight. Keeping in mind that some bending can be advantageous in gaining car "feel" at the limit. Same with Chassis twist, if its too stiff it looks good on paper but does not give the driver any feel and therefore be difficult to drive at the limit.

These are the problems we (Chris) are trying to solve.


Erik Madsen

QUOTE (Thorshammer @ Dec 17 2005, 08:49 AM)

-How much lateral deformation when a max cornering load is applied. Meaning doe the arm bend thowards the transmission when we load it and if so how much.

All very good stuff!!

Eric's question is very important. If you guys could apply to load at the tire contact patch you'd greatly enhance your investigation. Chris's current fixture measures the largest component and it's a great start.

Chris, if you could modify your fixture to apply the load at the contact patch so you get that additional force (bending) and then you could measure the trailing arm and then measure it again on the vehicle. Then you could determine what percentage of the compliance is associated with the trailing arm and how much compliance is associated with the chassis. That would be sweet.

I ran a crude version of this and decided to start over with a new trailing arm (and for some other reasons, too).

Keep it up, it is great when science is applied to making things work better!!!