Custom Control Arms

[Loren]

Didn't want to clutter up the Mirage "build thread" with this, as it's just ideas at this point.

I've lowered the car 2". I need to take some measurements and do some calculations, but general wisdom says that I've probably lowered the roll center more than the CG, which is bad. (more distance between roll center and CG = more leverage for your body roll to work with)

This happens because the angle of the lower control arms changed. They started out level. Now, because the body of the car is lower, the inner pivot of the control arm is lower. The solution to the problem is to bring that control arm back to level. The common fix is either a trick control arm that repositions the ball joint, or just a "tall ball joint" that effectively moves the end of the control arm down.

Of course, this is a Mitsubishi Mirage. Nobody makes trick control arms or non-standard ball joints. (in fact, the ball joint is theoretically not replaceable separate from the control arm)

Assuming my measurement and calculation indicates that I should fix this problem, I'm thinking about modding the control arms. I know it sounds scary. I used to freak out about it, too. But, I've looked closer at factory steel control arms, the fact that they are welded from the factory... and those welds don't always look that great! Plus, an awful lot of folks are driving around on control arms that are at least marginally compromised by rust. Bottom line: They are an overbuilt part. If they are modded in an even more overbuilt manner, no problem.

And, I've already played with modding a set of Miata control arms. Worked out great. My fear is gone. (mind you, I won't be welding this with my low-precision stick welder)

If I'm going to be hacking up control arms, might as well make the ball joint replaceable. And if I'm going to do that, I want it to be a dirt cheap, common, and durable replacement. There are ball joint "sleeves" available for both standard press-in style or thread-in style ball joints. Could get a set of those and just weld them in. Or, the ol' S10 uses the same suspension as the GM J-body (that's 70's-80's Camaro, Monte Carlo, etc) and is very popular in the hot-rodding world. Stock S10 ball joints can be had for less than $15 each, and would be plenty strong enough for a Mirage. Bonus: The S10 ball joints simply bolt in with four bolts. Would be really easy to fab a flat plate onto the control arm to attach it to. (but, the thread-in Mopar ball joints are cool, too... and equally inexpensive)

Before I get too carried away... I need to know the dimensions of the stud taper of the Mirage ball joint. I'll still be going into the same spindle, so I need the same size and taper. Since the Mirage doesn't have a replaceable ball joint, the only way to get those dimensions is to take the car apart and measure it. (or get a spare control arm... which... I probably should do, anyway)

So, yeah. Just a little mental exercise tonight thinking about how to solve the roll center problem (if it exists). Seems pretty simple once I've done some more measurement. The end result should be to get the control arms back to level, and maybe extend the control arm a bit. Need to do some math on that, too. Extending the control arm changes the angle of the strut ("steering included angle"), which is one of the angles used to derive the roll center... I need to consider the pros and cons of moving it.

This might all sounds completely nuts... but, check around the modified classes with cars that have better aftermarket support. (like BRZ/FRS) You'll find that if the rules allow it, and they are available, custom lower control arms are not at all uncommon. I just have to make mine.

[Loren]

Okay, so we autocrossed the car 2" lower and with stiffer springs. It FELT really good. No real complaints. But, somebody got a picture of it going through a skidpad turn with the inside rear wheel up in the air... aaaaaaand... the inside front wheel also up in the air. Well, maybe not "up in the air", but there's an air gap there. So, we need to do more suspension tuning... very likely getting into the topic of this thread!

Did a bunch more research on Roll Centers and their relationship to CG and handling and all that. And then I took some measurements on the car. And then I drew some diagrams.

Generalizations:
- When you lower a car (particularly a McStrut suspension), if you get the lower control arm angles pointed UP at the end rather than DOWN or LEVEL, you are lowering the Roll Center much more than the CG.
- The easiest fix for this is to correct the lower control arm angle. Either move the ball joint pivot point down, or the control arm pivot point up. It's usually easier to work with the ball joint end, and this is why "tall ball joints" exist.
- Your front roll center should be with 2-3 inches of the ground if possible, but preferably not under ground. (there are exceptions... and we're talking about a street car, not an F1 car with 2 tons of down force)
- Your rear roll center should be higher than the front, usually something like 6 to 10" higher is recommended.
- Your Roll Moment Arm should be as short as possible, but not zero. (I won't get to zero... but, if I did, in theory there would be no "body roll" and the car would feel numb and weird) The "Moment Arm" is the distance between the Roll Center and the CG. Longer "moment arm" imparts more force into body roll. Shorter moment arm puts more force straight sideways and is more better.

Assumptions:
- I don't know the CG of my car, so I'm trying to guesstimate. I really don't "need" to know the exact number.
- I estimated front CG by taking half of the overall height, averaging it with half of the hood height, and then averaging that with the crankshaft height. Totally not scientific, just a mathematical guess. I'm not even claiming a specific number. Just saying that I think it's somewhere between about 18 and 24". This is good enough for approximation and making comparisons between different setups. We're not going to reach "ideal" with this car, so the precision doesn't matter. We're just trying to get further from "awful", and a little closer to correct.
- I don't know my rear roll center height. I can't change it. And it just needs to be higher than the front, which it is. Beam axle suspension should put that rear roll center somewhere around 10-12". Good enough.

Measurements: (this is with the car lowered 2" as it sits today)
These measurements are to the nearest 1/4". Again, I'm just looking to compare different setups, and the setups are pretty radically different. Precision isn't critical for this.

Overall Height: 57"
Hood Height: 39"
Crankshaft Height: 13"
Strut Bearing Height: 28.5"
Strut Bearing Spacing: 43.75"
Tire Contact Patch Centers: 57.5"
Inner Control Arm Pivot Height: 5.5"
Inner Control Arm Pivot Spacing: 27"
Ball Joint Pivot Height: 6.25"
Ball Joint Pivot Spacing: 54"

Drawings:
This was totally worth the time it took to do it. It proves all of the things that "they" say about lowering a car. And it makes it pretty easy to understand what's going on.

Did this one first. It's directly based on my measurements. I didn't put it on there, but the Moment Arm is at least 18.4".

Mirage_FrontRef_Lowered_2inches.jpg (132.16 KiB) Viewed 13964 times

And then I took that same drawing and raised the car by 2" to get to approximately stock. The Moment Arm here is at least 15.3"

Mirage_FrontRef_Stock.jpg (124.07 KiB) Viewed 13964 times

Extrapolation:
- Lowering the car 2" lowered the Roll Center by 5.1"
- Increased the Moment Arm by about 3"
- Created Control Arm angles that generate a very negative camber curve. (bottom of wheel pulls in under compression, less negative camber)

Next up: I'll start playing "what if" games with longer control arms and different ball joint heights to try to get the roll center where it should be and maybe give it a better camber curve.

[lddavis91]

This smells a lot like "scope creep" to me...

[Loren]

Whatever. If I had a BRZ (or a CORVETTE), I wouldn't have to do any of this math or engineering, I could just open a catalog and order a "correct" control arm or a taller ball joint. Those aftermarket parts aren't available for my car, so I have to create them.

[twistedwankel]

Or even a soap box racer for that matter

[Loren]

Observations from playing with moving things around:

- Making the control arm longer doesn't move the Roll Center very much, but it does move it in the wrong direction. I want the roll center to be higher, closer to the CG. So, regardless of any other benefits that might come with longer control arms (more stable camber curve, more track width, etc), I don't want to do that because it doesn't help my #1 problem.
- Adding wheel spacers (or offset) to increase the track width has a VERY small effect on the Roll Center. Anything is an improvement, but it's not worth doing JUST to raise the Roll Center. Might do it for tire fitment at some point. The downside is increased scrub radius and wheel bearing load.

So, the big factor in getting the Roll Center out of the basement is the ball joint pivot height. Right where we started. But, at least now we KNOW.

I experimented with a 1" lower ball joint, and it was okay. Something like 2.3" Roll Center. Going to 1.5" lower on the ball joint put the Roll Center at 3". I'd like more, but I only have just a little over 2" of clearance from the ball joint to the inside of the wheel. So, I'll stick with 1.5" max to allow for some fudge factor.

3" roll center at the current ride height puts the Roll Moment Arm just a little bit less than stock. I'll take it.

Mirage_FrontRef_ModControlArm.jpg (187.84 KiB) Viewed 13950 times

[Loren]

Gave a few days of thought and research to the notion of creating a spacer to bolt between the stock ball joint and the steering knuckle. Found that it's been done before, and it can work. But, it would be nearly impossible in this instance due to the space available. The ball joint stud takes up nearly 1.75" of vertical space. That means I'd have to move it down over 2" to be able to fit a spacer without doing other mods to the control arm or ball joint. Max space available to move the ball joint down without contacting the wheel is 2", so it can't work.

Modifying the lower control arms to fit a heim joint in place of the ball joint is the solution to my front roll center problem. I'm just not ready to invest in doing it just yet.

[mymomswagon]

thanks for documenting and explaining.

in addition to helping with your strut travel issue, is it fair to say that your lowering is beneficial due the lower cg but the roll center issue gives back some of the benefit?

[Loren]

A term I've read recently that I kind of like is "geometric roll resistance". I've heard people talk about "mechanical grip" in reference to the same thing, but I never liked that term. I mean... all grip is "mechanical", right? Geometric roll resistance makes sense. It is independent of your spring rates and bar rates, solely related to your suspension geometry.

So, all of the math and diagrams above are PART of the picture. (I get dizzy when I try to go further than that, but there is more) You can play with the numbers. Somebody who is better at math than I am could even build a spreadsheet to input all of the variables and spit out the results. (actually, this is pretty basic trig... I could probably do it... maybe I should? Would be easier and more accurate than futzing around in Photoshop... now that I've fully VISUALIZED the process and made sense of it.)

in addition to helping with your strut travel issue, is it fair to say that your lowering is beneficial due the lower cg but the roll center issue gives back some of the benefit?

That seems to be a fair assessment. Still a net benefit. Just not quite as much of a benefit. I guess it could be quantified by calculating the degrees of roll per G applied. Without fixing the Roll Center, that will be (I presume slightly) more than it would be if you fixed the Roll Center.

[jbrannon7]

Here is a program, suspension-analyzer demo, that I found recently after you started this thread. It works full feature for 10 days. I downloaded it 5 days ago and am lucky that it already has all the parameters for a C5 corvette, I just had to take a few measurements from my car.
http://suspension-analyzer.software.informer.com/2.4/

[Loren]

I've seen that. Didn't know there was a demo. Too cheap to buy it.

[jbrannon7]

I've seen that. Didn't know there was a demo. To cheap to buy it.

I am not buying it either, when the 10 days runs out it will be gone. However I am making lots of changes to the model and printing the results. If I decide to use it again it will be a demo on my Wife's computer. Then and old laptop, and so on.

[Loren]

That's the way to do it.

Or... What it is, like $70? Given that you're setting lots of people's car's up, might be worth buying. Pass the cost along to your customers.

[jbrannon7]

That's the way to do it.

Or... What it is, like $70? Given that you're setting lots of people's car's up, might be worth buying. Pass the cost along to your customers.

Unfortunately it is $249.00 to $429.00 depending on features.

[Loren]

Ouch. Maybe I was looking at different software. Or maybe I just didn't look closely enough.

[blacksheep-1]

Loren, you are absolutely on the right track the Fox body mustangs have this same issue, when you lower the car the roll center goes under ground..to combat this a lot of after market suppliers use a much taller ball joint to force a little reality into the situation, they also use an an oblong steering rack mount and tie rod extensions to help the bump steer that comes along with all of this. I will post some links, so maybe it will give you some ideas..

https://www.americanmuscle.com/steering-offset.html

https://www.americanmuscle.com/stx2bajokit.html

https://www.cjponyparts.com/steeda-bump ... MNEALw_wcB

Also Circle Track Magazine just had a great article on how to induce camber during turn in, although, with a strut, you're pretty much screwed in that regard

[Loren]

Thanks, Rob.

Here's a question that I'm having trouble finding an answer to. That question is... how do you quantify the difference that "Roll Center Correction" makes? I mean, with lots of detailed analysis, I guess you could calculate the theoretical "degrees of roll per G" of a suspension (that would include spring/bar rates, of course), then change the roll center and calculate again. That would quantify how much LESS roll there actually is.

But, I'm not finding much info about that. Not even a subjective "I corrected my roll center with 1-inch taller ball joint after lowering the car 2-inches, and the car rolls about half as much as it did". Seems that lots of parts are sold with the intent of correcting roll centers "just because some engineer said so".

I understand the geometry, I don't doubt that it improves things to reduce the roll moment by raising the roll center... but "how much"?

[Loren]

I found an apparently free online suspension analysys tool:
http://vsusp.com

[mymomswagon]

thanks to this thread and Joe, I found the Racing Aspirations website. Separate calculators for double wishbone and macstrut. Android and iphone apps as well. small fee to use
https://www.racingaspirations.com/apps/ ... alculator/

[Loren]

That means I'd have to move it down over 2" to be able to fit a spacer without doing other mods to the control arm or ball joint. Max space available to move the ball joint down without contacting the wheel is 2", so it can't work.

About an hour before I came to this conclusion, I ordered some parts.

I ordered some 3/4" thick round steel pieces with the proper 7-degree tapered hole in them. (they're sold as weld-in shock mounts, I guess some shock mounts use the same design?) And I ordered some tapered ball joint pins. (intended as part of a custom tie-rod height adjustment for bump steer correction)

The pieces with the tapered hole, I intended to use as the bottom of my spacer assembly. Stock ball joint stud should fit right in there.

The pins, I intended to bolt or weld into the top of the spacer assembly.

Those parts arrived today, and they look just like I expected them to... that is... like they should work as intended.

The height of the spacer is still kinda critical if I don't want to have to modify the lower control arm. But, if I ended up making the spacers 2.5", I could still use them. I'd just have to modify the control arm... but, the mod would be much SIMPLER than doing a custom control arm end. The one thing that this forces that I wasn't sure about before is that the upper pin can't be a bolt-in. There's just not enough space for that threaded shaft and nut on the bottom... unless I make the spacer offset... which I could. To keep everything inline and as compact as possible, the pin would have to be welded into a hole in a 1/4-3/16" top plate. I think I'm okay with that, though. (I'd have somebody do a proper TIG weld on it)

The end of the contol arm is stamped steel with the edges being folded down by about 1/2 or 3/4". That's the part that could contact the inside of the wheel. So, depending on how much interference there is (if there ends up being interference), I could either just trim 1/8" off of the bottom of the control arm... or cut most of that folded down material off of the bottom, and recreate that reinforcement on the top side where there's plenty of space.

So... I'm back to "this could work". And I still like this option better than custom ball-joint or custom heim joint setup.

I need to pull a wheel off, pull a ball joint apart, and verify that all of my parts and pieces actually fit like they should. Then I can mock up a spacer and see if it will fit! If it passes a fit test, then I'll get to proper fabrication of some good spacers.