Something else I haven't thought about is the amount of spacers on the wheels possibly changing the criteria. I run almost 40 mm of spacers in the rear. This may change the information, but for the time being, we will at least have some relative data, and that will be useful.

Erik

Quick

Is Erik's and Chris' race E-Prod?

If so, it's scheduled for 4 am Monday morning.


Now, back to your regularly scheduled 914 component testing...

This is great and even though he hasn’t yet done the experiment I have been learning a lot from the discussion. But it feels like people are making this mole hill into the mountain.

I don’t get the impression that it is important to know every force, in every direction, to the .001? If the goal of the experiment is to know all the forces and how each surface of trailing arm reacts then ok. But if the purpose of the test is to get a practical way to strengthen the trailing arm against twisting force then I say he is right on.

Personally I am anxious to see what he finds.

Am I correct in assuming that 1 degree of deflection is no big deal for a street car?

As for race cars with big slicks, what kind of deflection are you shooting for? Is 1 degree 2 times too much? 10 times?

I love this stuff, I can't wait to see the proposed solutions and the results of testing them.

OK
I did some more experiments this morning.
The first thing I did was to rotate the trailing arm vertical so I could easily apply a load to induce lateral deflection. I was all alone so I couldn't make accurate measurements but the deflection is similar to the twisting deflection under the same loads. If anything it was slightly less. Personally I don't think lateral deflection is a problem as I am pretty sure the force component in that direction is much smaller.
I thought about my setup for a long time before actually assembling anything. Even though it appears to make significant compromises compared to what happens in the real world I am pretty sure it would be tough to make a major qualitative improvement over what I have. However making quantitative improvements wouldn't be too hard. I don't know if the load I am using is correct, and a digital level isn't as accurate as a dial indicator. OTOH, that is the same digital level I use to set up the suspension on the car and I am really only after comparative data, not detailed design analysis.
Since I was working alone I decided to standardize my forces and make it possible to read the level while under load. That meant I replaced my body weight on the lever arm with a stack of steel pieces adding up to about 220 lbs. The only shortcoming to this method is I can't observe the deflection as it occurs. I can only read the level to see the final effect. Good enough for now.
So I repeated yesterday's test with my standardized load and measured a deflection of 1.1 degrees.
Next I replaced the trailing arm with a boxed trailing arm and the deflection was 0.8 degrees under load.
Finally tested my idea for a reinforced trailing arm. Using just on piece of sheet steel that weighed 4 ounces and measured approximately 3x4 inches I was able to make an equivalent improvement to just 0.8 degrees of deflection on the original trailing arm. That compares to about 2 lbs of sheet metal used for a boxed arm, and a corresponding smaller amount of welding which means less distortion of the trailing arm.

You will never load the suspension enough to reach one degree of deflection in a stock based street car. These values only apply to race cars on sticky tires.

I would like to cut the deflection by 50%.
That is precisely what I'm shooting for.

Great project Chris!

I'm really interested in the results.

I'd really like to know where the most flex occurs, and in what axis. Does the arm bend more toward the transaxle? Or does it twist more? Where is the weakest link? The bushings? The mounts (inbaord and outborad)? The arm itself?

Keep the good info coming!

-Josh2

QUOTE (Thorshammer @ Dec 17 2005, 01:20 PM)

Something else I haven't thought about is the amount of spacers on the wheels possibly changing the criteria. I run almost 40 mm of spacers in the rear. This may change the information, but for the time being, we will at least have some relative data, and that will be useful. Erik

The amount of wheel spacers is immaterial. All that matters is the location of the contact patch. Not only that, within the range we are working on our cars small changes to the geometry will have correspondingly small effects on the forces applied.

Well, on a race car with hard bushings and a reinforced outer mounting bracket 95% of the flex occurs in the box section of the trailing arm. I am fairly certain that most of the flex is twist and only a minor component is lateral.
The stock outer bracket, unreinforced, flexes quite a bit. Stock rubber bushings also flex quite a bit.

Here's a pic of the trailing arm in my bandsaw. The kerf of the saw is about 60 thousandths of an inch.

and here's a pic of the rewelded trailing arm. The metal piece I inserted is about .075" thick.
This had the same effect as adding the boxing kit but it still isn't enough for me. I plan to add one or two more things to get the flex down to 50% from the stock arm.

Chris,

If we move the contact patch outboard we will increase the twisting force becuase we change the fulcrum. I am pretty sure this is correct.

As for the discussion earlier, why we need to know specifics is, when setting up the static geometry and knowing the amount of chassis roll, as well as knowing the degree of deflection of the trailing arm, we can understand completely what is happening, and after the arm is strengthened and the camber and toe curves are known, we can better determine what the static camber and toe should be set at to get the best performance.

Other factors such as Roll center and instant centers will need to be taken into account as well, several people are working on this as well. I think Brett, Chris, Kevin and I are attempting to ensure this has been thoroughly investigated and we have soild data and can prove the methodology. We have to ensure the stiffness of the chassis and components prior to altering other parameters or the feedback from the chassis will not be 100% accurate. We know D Finch alters the trailing arms, and Kevin has a stunning idea that may bear fruit, but we will have to test to make sure. If his season starts this year like last, he will need studded tires and a plow on the front of that thing.

More later....

Erik Madsen

PS: EP is me and Kevins race at 4:00am which is only 1 am for you west coast guys.

Chris thats EXACTLY how I imagined reinforcing the trailing arm. I thought of a way to get reinforcement inside the trailing arm without cutting the arm in half, I may try it on a damaged arm I have laying around.

Ok.. Which interior surface did you reinforce?

Did you do the reinforcing to the arm you first tested or to a different arm?

This is a piece of metal that is welded between the two halves.

Tim,

Chris and I have discussed two different methods, one is what Chris has done, next we want to try cutting slots at and angle so that the trailing arm can be triangulated by three peices. This should increase the torsional rigidity to what we are looking to do. Then we have another to reduce the lateral bending moment. We will see what is needed.

Erik

My thoughts were to cut slots also, and insert reinforcement and weld it in place. I though of doing it in 2 or 3 places.


Also the distance of the conact patch from the cl of the training are does matter. If you have the wheel spaced out very far, you increase the fulcrum that is acting on the trailing arm. It may not be a significant amount, but I dont think it should be neglected. I am going to try and crunch some numbers later and see what I come up with.

OK.

The amount of wheel spacers you use is largely due to the specific wheels you use. That's why I said it doesn't matter how thick they are. My wheels are offset differently so I don't need much wheel spacer.
However, EP allows you to have wider rear track which does affect the loads. I just don't think the distance is great enough to alter the methodology of what we are doing.

While I gree that cutting slots is the easy way to accomplish this I want to be able to alter the geometry at the same time. What I need next is the best location to add more stiffeners.

QUOTE (ChicagoChris @ Dec 17 2005, 05:35 PM)

Did you do the reinforcing to the arm you first tested or to a different arm?

same arm as in the first test

Would it be more advantageous to weld a plate non-perpendicular to the centerline of the boxed section? i.e. 45 degrees off in two axis?

I thought you were busy.

Really looks great...

Have you given any thought to carbon fiber re-inforcement???

Not sure of a good structural way to bond it to steel, but this concept is being field tested on old bridges that need additional structural repairs...just a thougt


Rich

QUOTE (McMark @ Dec 17 2005, 03:26 PM)

Would it be more advantageous to weld a plate non-perpendicular to the centerline of the boxed section? i.e. 45 degrees off in two axis?

I was thinking the same thing Mark.

OK, you go to all the trouble to fix the flexible trailing arm. That is good. But you don't solve the problem of the moving roll center caused by your toe and alignment changes. This is another reason to build a different trailing arm.

I am on this now and hope to have some data by Monday.

Chris, thank you very much for doing this testing, and for sharing the results with us!! It is fascinating reading.

--DD

Yes, its interesting but... I like Muellers idea for tools to measure with instead of that level.

This is one of the best threads I've seen here! Thanks for the effort and for sharing the results Chris!

Here is the way we reinforce trailing arms. It isn't an original idea and most of you have probably seen it before, but here's a pic just in case. Its a 'tubular" method...

The tube goes from the bearing carrier diagonally through both sides of the arm. Triagulates and stiffens the walls of the arm in one shot.

Never have tested one, maybe we should send one out to you so it can be compared to others on the same rig. It would be nice to know if we are accomplishing anything....

Here's a shot that shows where the tube comes through the inboard side.

Outboard side...

That arm has an option for increased shock travel? Am I interpreting that correctly?

QUOTE (McMark @ Dec 17 2005, 09:04 PM)

That arm has an option for increased shock travel? Am I interpreting that correctly?

good eye....

stock length shock on a lowered car - resulting in useful travel vs bottoming it out.....

Oooo! shock travel= cool!

QUOTE (Aaron Cox @ Dec 17 2005, 08:07 PM)

QUOTE (McMark @ Dec 17 2005, 09:04 PM) That arm has an option for increased shock travel?  Am I interpreting that correctly? good eye.... stock length shock on a lowered car - resulting in useful travel vs bottoming it out.....

Yup, bottoming out bad...

This gives almost two inches lower and keeps the stock travel.

If anyone wants a set, let me know...

Why not just buy the shock with the right length?

QUOTE (Brett W @ Dec 18 2005, 02:03 AM)

Why not just buy the shock with the right length?

or raise the top perch.

QUOTE (Mugs914 @ Dec 17 2005, 10:30 PM)

Never have tested one, maybe we should send one out to you so it can be compared to others on the same rig. It would be nice to know if we are accomplishing anything....

I'd be glad to do it!

good stuff here guys!!!

in the custom swing arm dept., here are a couple of shots of an arm on a twin turbo tube frame 914 i saw a pca club race at robeling road last year.

right side,

left side

right side

QUOTE (TimT @ Dec 16 2005, 07:41 PM)

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 ?

Right on Tim! totaly agree more accurate test results.

Make a rig on the arm to support the level.

Woah I replied to Tim's post before seing 5 pages here. A little too quick there.

That is Andy McNeal, I believe. That car is super bomb bad ass. Twin turbo 3.4 and fully custom tube chassis. I don't have a scanner so I can't scan teh 15-20 some odd pictures of the car.

QUOTE (Brett W @ Dec 18 2005, 07:43 AM)

That is Andy McNeal, I believe. That car is super bomb bad ass. Twin turbo 3.4 and fully custom tube chassis. I don't have a scanner so I can't scan teh 15-20 some odd pictures of the car.

i don't know who's it is. it was being ran/maintained by a shop in N.C. IIRC it qualified with the fastest time of all there that day, but never raced.

strange idea to see where the twisting starts:

cover the trailing arm in a non-flexible coating such as plaster of paris or sheetrock mud....when dried, it should crack as soon as the twisting starts

I'm thinking that instead of covering the entire arm all at once, just do 6 inch sections or so at a time.....

so, is carbon fiber an option here..

Can it be effectively bonded to steel...

Rich

QUOTE (r_towle @ Dec 19 2005, 10:08 PM)

so, is carbon fiber an option here.. Can it be effectively bonded to steel... Rich

I suppose you could make trailing arms out of CF that would be quite stiff.
Dunno about trying to accomplish stiffening by bonding CF to a stock arm. A lot quicker to do a little welding.

you could just pour hot liquid sugar over it, that stuff when it cools will crack alot under movement.

Here's a diagram of what I intend to try next. Move the slice forward (to the right) about 3", and add a diagonal plate inside as well.
BTW, some further testing revealed that the pivot tube does flex outboard of where the box section attaches. I will investigate this in more detail as well.

Chris, this looks great...though I would not trust my welding on that one....

It does make me think though...

In the longitudinal, where the heater pipe penetrates the inner long, could this type of bracing assist in stiffening that area..

I know in a pure race car, that pipe can be deleted...

But in a street car with heat, could you put plates all around that pipe and connect the inner and outer long in that area to assist in stiffening the car?

Keep in mind...im going in soon...as soon as my kids car is off my beloved jack stands...

Rich

Just waiting for more! Good reading....