Rear Lower Control Arm Design (LCAs): H-arm VS. A-arm

[TheNick]

Like the attached?

 

Very similar - but with more of the stock underbelly removed and the inner pivots higher in the vertical axis.

 

I'd think that keeping the CA the same length as stock would be pretty advantageous - the ones in the pics are what like 3" shorter than stock? You know why but for those that don't - the longer your CA's are, the less dynamic alignment change and roll center migration.

[tube80z]

Very similar - but with more of the stock underbelly removed and the inner pivots higher in the vertical axis.

 

I'd think that keeping the CA the same length as stock would be pretty advantageous - the ones in the pics are what like 3" shorter than stock?

 

Look again, looks to be about the same as stock to me.

[JMortensen]

Look again, looks to be about the same as stock to me.

It looks like he has them in about the stock height too in the first picture...

[Flexicoker]

The problem with the heim joint in this horizontal position is that it can easily bottom. Basically it will restrict the amount of travel available. I like the idea and especially the double shear mount of the rod end (could weld the clevis to the strut housing too, for that matter) but the implementation is a little tricky because of the range of motion with the rod end.

 

I don't have any arms in front of me to measure, but I think there will be enough misalignment in a normal spherical bearing with some cone washers/misalignment bushings to accommodate the suspension travel on most cars. With bolt-on clevises you can rotate them to make sure you have the most travel in both directions, someone would need to take some measurements to be sure before making any though.

[JMortensen]

I don't have any arms in front of me to measure, but I think there will be enough misalignment in a normal spherical bearing with some cone washers/misalignment bushings to accommodate the suspension travel on most cars. With bolt-on clevises you can rotate them to make sure you have the most travel in both directions, someone would need to take some measurements to be sure before making any though.

I can tell you that it gets REALLY close on my front tie rods with 5/8" rod ends and spacers, and that's with the strut sectioned, etc. With a non-sectioned strut and no droop limiters I'd bet that it would be a problem. For a race car there would definitely be some workarounds that could make it happen without too much effort, whether it be droop limiters or sectioning the strut a little shorter than it would normally be. With a stock strut I am pretty sure it would be a problem. It would have to be tested to be sure, but that's my SWAG.

 

EDIT--Being able to twist the clevis might be enough to make it work, and might be a reason not to weld the clevis to the strut...

[heavy85]

What's the practical benefit of bringing the rear suspension back closer to factory roll center? This is not a theoretical question but a real seat of the pants 'how will this lower my laptimes' quandary.

 

Cameron

[thehelix112]

I think:

 

Less rolling moment (weight transfer?) due to shorter distance between RC and CoG. This will mean you can run lighter sway bars, which will let you get back some of the independent function of the rear wheels. Meaning that perhaps you can get on the gas earlier after hitting a ripple strip exitting a corner?

 

Disclaimer: 2 seconds worth of thought, and I suck.

 

Dave

[JMortensen]

Roll Axis Inclination: http://www.eng-tips.com/viewthread.cfm?qid=91420&page=9

 

The effect on TLLTD they're describing can be showing using the Weight Transfer Worksheet that comes up so frequently.

[Lazeum]

I'm following your thread since the beginning; I've tried to come up with nice charts to picture my thoughts about it.

 

First, I've heard bending in toe link arm. How it is possible since a heim joint cannot transfer moments (it is a ball joint)?

I'll be more than happy to discuss about it (of course, otherwise I won’t post on the board), if you can destroy my theory, please do so We all need to learn.

 

Based on my knowledge/analysis, A-arm does not seem to be as good as H-arm as far as load distribution. As far as Adjustability, it is another story, A-arm are probably better... Let me try to explain.

 

I think we have to go back to the strut/knuckle assy to start with... There are NO LOADS IN Z direction taken by the arm during normal operation, this is how the strut/knuckle is designed.

You could also see the difference between H-arm and A-arm. Load distribution is most probably BETTER on H-arm. You could orient Toe link on A-arm to match performance of the H-arm for 1 case at a time but not all of them.

 

First, description of load/reaction force of the systems.

In bold blue are the free degrees of liberty allowed by each joints.

 

 

I've tried to play as well with some load cases, to analyse the situation deeper...

I’ve put some round arrows sometimes to shows stress impact on arms. I think A-arm as always higher stress than H-arm.

 

The more severe case as far as suspension are braking and wheel bump. This is based on my knowledge, it was my job before...

 

First is Braking case, the worst one by far (you brake wayy stronger than you accelerate)

 

Then they is wheel bump that can be very severe, especially on street cars

 

and cornering as well

 

To continue to the current topic about roll center; if the arm axis (both outer joints and both inner joints) are parallel to the ground, there is no reason to get some Z load into the system, stiffness in Z direction is almost useless, you can clearly see it, none of the solution have “volume” in Z direction.

 

Current production car today uses an off centered bolt and washer to play with parameters such as camber and center roll position. It seems to be the easiest way to do. However, I’m not sure if it would be possible on s30 without body modification.

 

I’m waiting for your comments, now

[Lazeum]

In addition to what I just post, if you look at the suspension of the s30, you can clearly see that the rear mounting point of the control arm is not designed to get load torward X direction. The only piece that avoids the suspension to bend in this particular direction is the conrtrol arm itself. The rear support piece are flat, they would bend under load very easily.

 

So X loads are going through the front mounting point. In braking cases, A-arm might still be even worse...

The braking case I've described is most probably wrong since I put most of the X load of the rear mounting point which is not true.

 

Now, I'm just destroying by myself the theory in public

X load on the rear mounting point should be removed. Another proof is the size of the bushing, front inner bushing is bigger than others.

This is how the suspension should have been described at the first place

 

To go further, it would mean with the A-arm, Y load would be much more severe. The bent cross member (the one that is straight on early s30) might get much more stress in this condition. A beaffier part might be also required to avoid failure in this location.

 

I think we need to be realistic afterwards, many heavily modified cars with 300mm wide tires have been on the road, probably some with A-arm suspension design. Failure doesn't appear a lot, the suspension can definitely take some abuse (as long as car are driven a little since failure will be related to fatigue).

[TheNick]

Look again, looks to be about the same as stock to me.

 

Yup you're right - the 3rd picture in the series is at a deceiving angle and makes it look like the heim mount is at the end of the 2x3 tube that is bolted to the stock locations.

[JMortensen]

Very cool diagrams. I'll give you my thoughts. Not saying that I know I'm right here, but these are my thoughts on your diagrams.

 

To continue to the current topic about roll center; if the arm axis (both outer joints and both inner joints) are parallel to the ground, there is no reason to get some Z load into the system, stiffness in Z direction is almost useless, you can clearly see it, none of the solution have “volume” in Z direction.

If you read back in previous discussions about roll center, there is a term "jacking" that comes up. If the LCA is not parallel to the ground, there will be jacking. If it is lower on the inboard side, the suspension will jack down under cornering loads. If it is higher on the inboard side, the suspension will jack up under cornering loads. A higher roll center will allow more roll resistance with less spring rate, because the suspension will tend not to compress under cornering loads. A lower roll center requires higher spring rates for the opposite reason. Typically on a lowered Z the front roll center is underground and the rear is about neutral or slightly above ground, which still gives that favorable roll axis inclination. With guys installing 25 and 26 inch diameter tires (read: big rims), they tend to lower the cars more and then the rear is also underground, and I have to question whether there is any roll axis inclination anymore in some of these situations, especially where they move the front LCA pivot up but don't do anything to the rear. This would tend to make the car harder to catch in a slide since it has less yaw damping. Maybe that can be overcome with bigger rear tires. I think it's going to depend on the total setup on each car you look at, but certainly I think we can say having a RAI that is lower at the front than the rear tends to produce a more stable easier to drive car.

 

Current production car today uses an off centered bolt and washer to play with parameters such as camber and center roll position. It seems to be the easiest way to do.

G Machine adjustable bushings do this for the rear of the Z, but they have a fairly limited range of motion and they are hard to set exactly the same front and rear because the front bushing is larger in diameter. They definitely don't have a large enough range of motion to get the roll center back above ground on some of the really lowered Z's, like Terry Oxendale's car.

 

However, I’m not sure if it would be possible on s30 without body modification.In addition to what I just post, if you look at the suspension of the s30, you can clearly see that the rear mounting point of the control arm is not designed to get load torward X direction. The only piece that avoids the suspension to bend in this particular direction is the conrtrol arm itself. The rear support piece are flat, they would bend under load very easily.

The front bushing controls the braking load. The rear controls some lateral loading, which I suppose comes at a bit of an angle since it goes through the range of motion available after the front has done it's job. With rubber bushings in there they both probably do some of everything. With monoballs or rod ends on front, I think the rear is more effectively isolated, which means the front does more work in the X plane and the rear does more work in the Y plane.

 

The braking case I've described is most probably wrong since I put most of the X load of the rear mounting point which is not true.

I think you're on the right track here.

 

To go further, it would mean with the A-arm, Y load would be much more severe. The bent cross member (the one that is straight on early s30) might get much more stress in this condition. A beaffier part might be also required to avoid failure in this location.

Not so sure about this. Regardless of which arm type you're looking at, the front bushing handles most of the braking load and the rear handles most of the lateral loading, as in either scenario the front pivot is stouter and the rear is weaker. I guess I'm not seeing how the rear pivot is more stressed in the A arm. I think it is more isolated with monoballs or rod ends in use vs bushings and gets more directional force vectors from the lateral loading, but I don't think it has more loading in the Y axis in either design.

 

I think we need to be realistic afterwards, many heavily modified cars with 300mm wide tires have been on the road, probably some with A-arm suspension design. Failure doesn't appear a lot, the suspension can definitely take some abuse (as long as car are driven a little since failure will be related to fatigue).

Including viperredls1 who broke the AZC H arm, fixed it, and broke it AGAIN with his LS powered autox car on sticky tires.

[Flexicoker]

Lazeum, you're braking scenario is not complete, you also need an X-direction force applied to the spindle.

 

Also, the bending reference was not talking about the toe-link, it was referring to the entire H-arm. If the strut to control arm pivot axis is not parallel to the control arm to body pivot axis, then there is going to be a moment acting on H-arm trying to twist it, and the only things reacting that force will be the torsional stiffness of the arm, and the strut itself, which will cause more friction in the strut. The A-arm design eliminates that problem.

[260DET]

I think:

 

Less rolling moment (weight transfer?) due to shorter distance between RC and CoG. This will mean you can run lighter sway bars, which will let you get back some of the independent function of the rear wheels. Meaning that perhaps you can get on the gas earlier after hitting a ripple strip exitting a corner?

 

Disclaimer: 2 seconds worth of thought, and I suck.

 

Dave

 

No Dave, no suck. Typically in my corner of the world competition S30's are lowered with no consideration for roll centers, except, perhaps to compensate in the front. Then they try to run front ARB's only, which is basically a good idea because as you mention it allows the IRS to work better.

 

But, without going into it deeply, no account is taken of the change in roll centers and inclination so typically the handling is not improved over stock and may even be worse. Which is disguised by the use of semi or slick tryes. Its very frustrating to see.

 

With the Hybrid280ZX very careful attention has been paid to roll centers and inclination together with removal of anti squat and so on. As a result the car handles and drives beautifully and is starting to do very good lap times even with me at the wheel.

 

A lot of such attention was prompted by following various suspension topics on this forum over the yeears

[heavy85]

Less rolling moment (weight transfer?) due to shorter distance between RC and CoG. This will mean you can run lighter sway bars, which will let you get back some of the independent function of the rear wheels. Meaning that perhaps you can get on the gas earlier after hitting a ripple strip exiting a corner?

 

Not only to improve the independent function but this would also help spread the load between both rear tire due to less weight transfer and gain overall forward acceleration. Since getting the power down is one of my biggest issues I need to give some serious thought to the rear suspension. I've just taken the rear for granted focusing on the front end but maybe I'm leaving a lot on the table.

 

Cameron

[TheNick]

If you read back in previous discussions about roll center, there is a term "jacking" that comes up. If the LCA is not parallel to the ground, there will be jacking. If it is lower on the inboard side, the suspension will jack down under cornering loads. If it is higher on the inboard side, the suspension will jack up under cornering loads.

 

Technically - anytime the one wheel instant center is not exactly on the ground - there are jacking forces present.

 

This is why the Kinematic Roll Center calculation has been proven to be more of a "internet engineering masturbation exercise" (I think I stole that from JohnC somewhere ) recently as Bill Mitchell has presented papers at a few recent SAE conventions detailing the benefits of the force based RC calculations (which you can't do with a CAD program or without knowing actual tire loading). His most recent versions of WinGeo has the capabilities to calculate the FRC - you just need to estimate tire loading based on the desired roll gradient, lateral acceleration and CG height.

 

 

I could try to find the papers - but I'm not at my work computer and don't have access to the SAE library without it.

 

 

Continue on with your regularly schedule thread

[TheNick]

 

With the Hybrid280ZX very careful attention has been paid to roll centers and inclination together with removal of anti squat and so on. As a result the car handles and drives beautifully and is starting to do very good lap times even with me at the wheel.

 

 

What did you do to remove anti squat? I realize you are talking about a ZX model but the discussion probably is still worth it.

 

Spacers under inner control arm pivots?

[TheNick]

Not only to improve the independent function but this would also help spread the load between both rear tire due to less weight transfer and gain overall forward acceleration. Since getting the power down is one of my biggest issues I need to give some serious thought to the rear suspension. I've just taken the rear for granted focusing on the front end but maybe I'm leaving a lot on the table.

 

Cameron

 

 

Lateral load transfer is based solely on CG height and track width - the RC location doesn't have anything to do with it. If you are having issues with putting power down then you need to know what the dynamic alignment changes look like under bump. Does the outside rear toe out or in under bump? How much? Whats the anti-squat look like on these cars (I don't know but I have a feeling its a rather large percentage - but I could be wrong - its happened before )

 

Maybe you are running too much static camber? Try stiffer springs and run less static camber? Not as much roll so you won't need as much camber for lateral acceleration and the tire will be more upright when its time to get on the gas for exit.

 

A large rear sway bar will also make it harder to put power down (what type of LSD? Clutch? 1.5 way - 2 way? Helical?) The LSD type will factor into what size of rear bar you need to run. In a higher HP car - I think it would be advantageous to find a way to not run a rear bar at all like 74_5.0L_Z.

[TheNick]

Great thread discussing RC movement relative to CG height

 

http://fsae.com/eve/forums/a/tpc/f/125607348/m/47110403711

[EMWHYR0HEN]

A large rear sway bar will also make it harder to put power down (what type of LSD? Clutch? 1.5 way - 2 way? Helical?) The LSD type will factor into what size of rear bar you need to run. In a higher HP car - I think it would be advantageous to find a way to not run a rear bar at all like 74_5.0L_Z.

 

 

I agree. I prefer heavy rear spring, no rear ARB, and a one way.