strut tower bar theory

Ok, I've read through this thread and the associated links. This is going to be a long one, and I'm not sure where to begin. For those who hate my long-windedness, go ahead and click to another thread or you'll just get mad.

Weight Transfer
You can make a car that has nearly zero lean, but you will still have weight transfer. It will be the same cumulatively (front + rear on one side) as a car that leans a lot. What I mean by that is that the total weight transfer from one side to the other will be the same. Note that is is also an approximation - dynamic changes in CG height do have an effect. This type of effect needs very detailed analysis, and that's beyond the scope of this little blurb.

What you can change is the rate of weight transfer. That's what you do when you put a bigger bar on either end of the car, or when you adjust your damping settings. At the final steady-state, the weight transfer is the same magnitude, but the rate of transfer was changed.

The second thing you can affect is the location of the weight transfer from front to rear. Putting that big rear bar on the car makes the rear weight transfer happen more quickly, and thus you load the outside rear wheel faster. You have also changed the roll couple of the car - this describes where the weight transfer goes on a car. I'm not an expert on roll couples, but it's what you want to calculate after you have determined the roll stiffness of a vehicle to give you an idea of where the weight transfer is going when you make a change. You want to check it (if you can) to make sure you haven't created an unstable monster. Somewhere I have equations for this, but I'm still on vaca - my files are at home. The calcs involve moments of inertia and/or roll center heights, IIRC. But again this is what you change when you alter roll stiffness of a vehicle.

This is why putting a 40mm rear bar on a Focus (with no other mods) would give you an undrivable car. The weight transfer would happen quickly and predominantly in the rear, saturate the tires, and then the rear would break loose quickly. You've altered the rate and location of the weight transfer. You have *not* changed the magnitude (again, neglecting CG migration), but you have slapped it nearly all on the rear.

Chassis Stiffness / "Strut Tower Braces"

First: ALL braces, be they front strut braces or rear shock tower braces, have some effect. No matter what brace you use or how crappy/good it is, you have changed something.

Second: The question is how much. As an anology, think about a river. I could leave my house and go whiz in the Detroit river. Have I changed the overall level of the river? Yes. Have I changed it a meaningful amount? Well, aside from probably making the water cleaner (hee hee...a little DTW humor there), no. Do the same with a Dixie cup - now you have a huge effect on it.

Now onto some illustrations using various car examples. Let's set aside our Brand X flames for now, as I'm going to use examples that are both Focus and non-Focus.

Example 1
A great example of what you can do with braces is the 1997 Integra Type R (ok, all Integra Type R's). What many people don't know is that you can't make a GS-R into a Type R *exactly* since the body shells are not the same. Here is what Honda did:

1. Double metal thickness in the C-pillars
2. Cross-car brace behind the taillamps
3. Rear suspension tie-bar, visible under the car. It connects the inner pivot points of the rear suspension.
4. Front shock tower brace (no struts here - this is a double wishbone suspension).

There may be another brace in the rear hatch area that I'm forgetting, but it is a hidden one. Regardless...

Why did Honda do this? Well, Hondas are a lot of things, but "known for bank-vault rigidity" isn't one of them. They knew they were going to use spring and bar rates that were more than double those of the GS-R, so any flex the GS-R had would be made a lot worse. Honda wanted to maintain cargo capacity (hence no rear shock tower brace), but also stiffen up the weak points of the car. I'm sure they used a finite element analysis model to find the weak points, and then they applied local stiffeners to the car where they would have the most effect.

Example 2
1988 Toyota MR2's (and most of the Gen 1 MR2's) came with front strut braces. Why? Same reason outlined in that BMW E30 link that is in this thread somewhere - MR2's (and many mid-engine cars) will lift the inside front wheel during max cornering. That BMW E30 article nicely explained what happens during something like that. Toyota apparently felt a brace was necessary to have the handling they wanted, as well as the NVH qualities.

Example 3
Someone here mentioned that they put a brace on an "old Civic." That is a PRIME example of a car that truly needs bracing. The 1984-87 Civics/CRXs (and their sisters, the 1986-1989 Integras) were wet noodles. Major flexible flyers here, and they ran pretty soft (~100 lb/in) spring rates (yes, I know these cars didn't have springs in the front). If you stiffened these cars up at all, you'd bend the body more than you'd move the suspension, it seemed. Geometry moved all over, ride went to hell, etc. Braces were a big help.

Example 4
This is just a general example. Sometimes these braces are there for things other than handling. Some convertibles have terrible cowl shake. Adding a brace (look under the hood of a '95 Mustang convertible) can help reduce that shake, making the ride quality/sound/etc. much nicer for the passengers.

Now let's come back to Focus land, or at least a bit more of reality.

I don't have numbers to prove this, but experience and feel tells me that the Focus is not a flexible flyer. No, it isn't a bank vault (no hatchback really can be), but it is certainly a heck of a lot more stiff than the Civic/CRX I mentioned in Example 3. Also, a stock Focus is running pretty light (~126 lb/in) spring rates that don't tax the structure that badly. Do braces have any effect? Yes. Do they have a big effect on a stock suspension? No. We're talking whiz/Detroit river here. Should you put in braces if you are going to go up to 450 lb/in springs, big bars, and competition-level damping? Very likely, yes. You've moved from the Detroit river to something more like a kiddie pool.

"Why doesn't the Focus use a rear suspension tie bar like the Type R?"

The Type R attaches its suspension to sheet-metal suspension mounts that are part of the body shell. The Focus attaches them to a separate subframe. The subframe itself is acting as the tie bar.

"I want to feel happy about the nice shiny thing under my hood." Someone had said something like that earlier in this thread.

Absolutely! Feel good about it - adding stiffness is usually something good! But don't attribute a huge change in handling to it if you are relatively stock suspension-wise. It's more like the difference between grippy and slippy sta-bar bushings - it's there, but it is very subtle in street applications.

My Point
Sometimes it seems like I never have one, doesn't it? The point is that with a car as solid as a Focus, huge gains won't be had with simple changes at near-stock suspension tunings. Yes, the Focus flexes - it is not perfect, it is not a block of granite, it does flex. But it is a pretty good base to start with compared to some other cars out there. That's about it.

Anyway, I think that's it for now. Hopefully this sheds some light on things, somewhere, somehow. Thanks for putting up with the verbosity.

[ 04-25-2003, 05:24 PM: Message edited by: CTB ]

MXi,
The check for being my personal PR man is in the mail.

Rod, absolutely.

You must know, that you are talking vertical forces. Most people associate strut tower bars with with lateral forces, which is what we're discussing.

Vertical loads are a whole other matter.

But you're absolutely right. The dynamic forces exerted by the shocks would be many many times the maximum force exerted by the spring. And if you bottom out your shocks, it's a whole other matter!

For example, the front bumpstop will exert a force of 2700lbs if compressed 1.5". Think about it.

Anyone know off-hand what the SCCA rules are concerning "Additional" welding w/o adding any parts or relocating anything?

Getting back to the original topic (sorta). I researched roll couples a bit, so now I have more info to complete what I started in my long rambling thread.

While you can calculate a front and rear roll couple (basically the amount of weight transfer), what is most important here is the roll couple distribution. This can be just the ratio front and rear roll couples (expressed as a number like "2"), or it can be expressed as a percent. Example: You calculate that the roll couple distribution is 2. This means that the amount of weight that is transferred to the outside front during cornering is twice that of the amount transferred to the outside rear. Numerically, that means that if 1000 lb is transferred in total, 666.7 lb will be transferred to the outside front, and 333.3 lb will go to the rear. If you express this in percent, then a roll couple of 66.7% would do the same thing (1000 lb total, 666.7 lb goes to the front, 333.3 lb goes to the rear).

As you can see, if a roll couple distribution is greater than 1 or 50%, then more weight goes to the front and the car will understeer. This matches nicely with the concept of why putting a bigger front bar on makes you understeer. Or rather, it explains it. If you have a roll couple distribution less than one or 49.9%, now you have a car that will have terminal oversteer since the back is going to load up and break loose first. Not a good idea, in general.

If you have a roll couple of exactly 1 (50%), then do what one document I read once said to do: Bronze the car, as it is now perfectly neutral and you'll never be able to do that again.

[ 05-01-2003, 04:53 PM: Message edited by: CTB ]