Jump to content
HybridZ
tholt

Yet another Rear control arm design

Recommended Posts

Like just about everyone else, I have an overwhelming compulsion to design and build my own rear control arms. My application is for an autocross only car, so keep that in mind as you pass judgement. It will never see the open road or >100mph...

That said, this is pretty much a direct copy of a setup I have seen on a GT2 Z and it hasn't fallen to the ground yet after several years of competition. Only the sway bar link has been added. I really like the total on-car adjuastability. One of my main goals is to facilitate as much track width/camber adjustment as possible from the bottom. Simply detach one end of the mid link and all heims can be adjusted. Toe can be controlled with just the rear adjuster.

The inner mounts will probably be a simple bolt-in frame that will locate the inner heims as close to their stock locations as possible. I will update as that get's developed?

Does anyone see any major concerns with this? It seems to be a proven design, but I could be wrong...

rear-arm-design-sm_thumb.jpg

Share this post


Link to post
Share on other sites

You say you're looking to change the track width? I would caution you against running too much exposed thread on the rod ends, according to the rod end manufacturers they want 1.5x the diameter threaded into the tube end, which gives you only 5/16" adjustment on a 5/8" rod end with a standard threaded tube end that is 1.5" long. This is a highly stressed area and while I think I might stretch those recommendations elsewhere I wouldn't here.

 

That said you could simply get longer turnbuckles to make the arms longer. The arm itself doesn't seem like it would be very rigid at all. The angle of the link from the inner rear pivot to the outer front pivot essentially makes it good for nothing. I realize that in the drawing all it is doing is holding the front rod end setup horizontal but it really isn't positioned to do even that job very well. A shorter arm positioned perpendicularly to the clevis would be better. It seems like those clevises would be under huge amounts of stress, since they take all of the load whenever the strut doesn't just want to move straight up and down. All of the accel and decel loads that the arm takes, and also all of the cornering loads are going straight into those clevises. It's asking a lot of those parts. Too much for my tastes.

 

If you were going to the trouble of starting from scratch with the arm and the chassis mounts, seems like the thing to do would be to move the pivots to the center of the car and make the arms as long as possible. This would also get rid of the uprights behind the diff and I think that is a particularly weak part of the original design of the car. You could improve on that quite a bit if you wanted. Also if it were my project I would tie the front and rear rod ends at both the strut end and the chassis end together much more securely than your proposed design. Your design would allow for a lot of side force on the strut shaft.

 

My engineering degree doesn't exist, I took trigonometry 4 times before passing, and I sell doggy doors for a living, so keep in mind who is criticizing you... :D

Share this post


Link to post
Share on other sites

My engineering degree doesn't exist, I took trigonometry 4 times before passing, and I sell doggy doors for a living, so keep in mind who is criticizing you... :D

 

:lmao::lmao:

 

dont worry jon. some peoples brains are not wired for math. im and engineer, i stay away from words.

 

jimbo

Share this post


Link to post
Share on other sites
Like just about everyone else, I have an overwhelming compulsion to design and build my own rear control arms. My application is for an autocross only car, so keep that in mind as you pass judgement. It will never see the open road or >100mph...

 

Does anyone see any major concerns with this? It seems to be a proven design, but I could be wrong...

 

No real major concerns but some feedback. If you make the top triangle all from tubing it will be lighter than using the clevice arrangement. It will also be stiffer from a compliance standpoint. The toe link should be cleviced (or better yet fed into a double shear gusset) that is from this triangle. How these are attached to the strut will make a difference in compliance too. If you can I'd mount the swaybar to the strut to reduce friction on the joints and remove a small bending load.

 

I ran a similar set of arms on one of my cars when I was trying to get parts figured out for the tube car. One issue I ran into is that you have to watch for is the halfshaft hinge points. When I ran longer lower control arms there seemed to be some kind of binding when the car was loaded under power. If you look at the stock arms you'll see that these points are very close. CVs may be the answer but I didn't have them at the time to try.

 

If you want to be able to adjust track you might consider changing the inside to sphericals and using shims to adjust them. If you do all the work in the shop then this is quick and easy at the track.

 

For a Z I'd first look at working the front for better grip before playing with the rear.

Share this post


Link to post
Share on other sites
One issue I ran into is that you have to watch for is the halfshaft hinge points

 

Very good point!! I dont think you can safely change the length of the arms without causing very undesireable geometry problems with the pivot center of the axles,cv's sure-you could get away with more there but thats another thing.

 

I like your thinking though,i am actually trying to come up with an original/effective design for some rear lca's myself.Unfortunately you will probably find yourself at least somewhere close to what has already been done succesfully by the time you have it right.

Share this post


Link to post
Share on other sites

What I don't like is that the outer 'main' rodend which attaches to the blue clevis, and the inboard rodend on the toe-link are going to be in bending (I'm calling the pink/green member the toe-link). If you made the blue guy capture a properly secured spherical bearing, you'd eliminate one potential source of failure. You could also switch the mounting arrangment of the toe link so that you instead have a seperate A-arm consisting of the two links mounting to the blue thing, and a seperate toe-link. Ideally, you should never have threads (rodends) trying to take a bending load. Their effective diameter is significantly less than the minor diameter of the threads due to the stress concentrations that the threads create.

Share this post


Link to post
Share on other sites
What I don't like is that the outer 'main' rodend which attaches to the blue clevis, and the inboard rodend on the toe-link are going to be in bending (I'm calling the pink/green member the toe-link). If you made the blue guy capture a properly secured spherical bearing, you'd eliminate one potential source of failure. You could also switch the mounting arrangment of the toe link so that you instead have a seperate A-arm consisting of the two links mounting to the blue thing, and a seperate toe-link. Ideally, you should never have threads (rodends) trying to take a bending load. Their effective diameter is significantly less than the minor diameter of the threads due to the stress concentrations that the threads create.

That's how Cary ran his control arms if I recall, except he had a full triangle in the front part with 3 rod ends and then a toe link at the rear with 2. The problem I have with that design is that the bottom of the strut isn't held perpendicular to the ground and the rear toe link (or the rear part of Tom's arm) don't prevent the strut from twisting. That twisting imparts a side load on the struts. Struts already have issues with side loads without adding more into the equation. It might work, but it's not optimal in my opinion. How much less than optimal it is is up for debate. Seems like Cary was going pretty fast with his old arms from what people tell me.

Share this post


Link to post
Share on other sites

Which twisting force are you referring too? The force that could be described by sort of a circular arrow when viewed from the side of the car (I'm not sure that made any sense), which would be the reaction force from the brakes, and to a lesser extent any longitudinal force, can be simplified into a longitudinal reaction force where the strut meets the chassis, and a larger longitudinal force in the opposite direction where the strut meets the control arm. So basically, any twist in that direction is going to be handled the same way as a longitudinal force would be by the control arm. I'm willing to bet that torsional rigidity of the stock control arm adds negligible little stiffness to the entire strut assembly.

 

A possibly easier way to describe this is; imagine looking at the drivers side of the car, when you hit the brakes, there is going to be a torque created about the axle in the counter-clockwise direction. Imagine the strut is only connected by a pin joint at the top, and one pin joint at the bottom. There is going to be a force that the chassis applies to the top of the strut thats going to be acting to the right, or rear of the car, and theres going to be a force from the control arm to the bottom of the strut acting towards the left, or front of the car. If you've got an A-arm attached to the front of the strut and a toe link on the rear, the A-arm is going to handle this load completely, the larger the angle between the arms, the stiffer it will be while handling a force in that direction.

Share this post


Link to post
Share on other sites

Ah, a better example...

 

Imagine you've got a swinging pendulum, the top, where it pivots is essentially the top of the strut. Whats the easiest way to stop this pendulum from swinging? Do you apply a torque to stop its rotation, or a a force to the bottom of it to stop its swinging?

Share this post


Link to post
Share on other sites

Forget about the toe link for a second. Imagine you have a triangular arm which attaches to the stock front and rear inner pivots and to the front of the strut housing. There is no strut insert in the strut housing for the purpose of this example. Set the car on the ground, and the strut housing will rotate forwards (would turn in the same direction as the wheels when driving forwards) until the monoball binds. That's just with the weight of car, so think of the force on the strut when you hit a bump in the middle of a corner

 

Now add a rear toe link and it does the same thing. Tom's arm is so weak in the back that it would also do the same thing.

 

The stock arm is pretty weak too, but my modified arm (and Terry Oxandale's and John Hines's) has a large tube which would prevent that twist.

 

Here is a pic:

DSCN1447.jpg

 

I think that twist is significant because the main issue with strut suspension as I understand it is the stiction created by side forces on the strut housing when turning. That lateral force wants to lock the strut so that it can't move. But now you've designed a control arm which by design increases the amount of side force on the strut. In a game where people are putting torrington bearings on the ends of the springs because the spring windup causes unwanted forces on the strut it just seems to me like you would want to avoid loading the strut if at all possible.

 

This is one of the very few things that I've really disagreed with Cary on, but I really think I'm right on this one.

Share this post


Link to post
Share on other sites

You finally made sense to me on you last reply Jon... Good point, but I think the thing that has probably convinced me to ditch the plan is the fact that the forward joint (blue) is only held by 3 rotataing points. In other words, that forward triangle is not a triangle. It's a quadrangle with a short leg running through the blue piece. The only thing that will keep that joint from rotating fore and aft is the friction of the clevis bolts. It might be sufficient to do the job, but when it slips it's going to get real weird real fast.

 

Maybe there is a reason that the car I saw this on is a perennial back marker! :icon7:

 

Time to go back to the drawing board for a more conventional design.

 

Thanks for the reality check guys!

 

Tom

Share this post


Link to post
Share on other sites
... but IMO trying to resist the rotational torque via the LCA is like trying to open a door near the hinge ... not real effective nor a loadcase I think the LCA was designed for.

I think there were some formula cars that were designed with an H lower control arm and an upper link. I've been looking in some books and can't find an example, but I'm sure I remember reading that, and I thought they were F1 cars. I thought I read that in Competition Car Suspension, but I can't find it now. I did find a similar setup in Race Car Vehicle Dynamics, but it's just listed as a type of suspension and the car it's referenced with is a 90s T-bird, not exactly the pinnacle of performance, but I think it proves the concept of controlling the twist with the lower control arm. I think the control arm can do the job just fine, if the arm is properly designed for the job.

 

Back to our cars for a minute, I think this is the key point: If Nissan didn't want the control arm to deal with those twisting loads at all, they could have easily designed a lower A arm and a toe link instead of the lower H arm that the car got. I think the fact that they used an H arm is proof of what the arm is intended to do.

 

The only thing that will keep that joint from rotating fore and aft is the friction of the clevis bolts. It might be sufficient to do the job, but when it slips it's going to get real weird real fast.

Yep, that's one reason why I was so hesitant about the clevis link to the rear part of the arm. That could be a solid tube welded in and then the design would be a lot better in my opinion.

 

Tom, have you looked at John Thomas's arms (post #36): http://forums.hybridz.org/showthread.php?t=106457&page=2

Share this post


Link to post
Share on other sites

I've definitely seen Johns control arms... My only gripe is there is no provision for toe adjustment. I think I am leaning more towards something like the Arizona steel tube arms with it's adjuster but add heims on the inner mounts...

 

Tom

Share this post


Link to post
Share on other sites
I've definitely seen Johns control arms... My only gripe is there is no provision for toe adjustment. I think I am leaning more towards something like the Arizona steel tube arms with it's adjuster but add heims on the inner mounts...

You could do something like John has done coupled with my/blueovalz's/jeromio's "poor man's toe adjuster": http://forums.hybridz.org/showthread.php?t=89111

Share this post


Link to post
Share on other sites

"Yep, that's one reason why I was so hesitant about the clevis link to the rear part of the arm. That could be a solid tube welded in and then the design would be a lot better in my opinion."

 

Yup, Jon's right, won't work until something binds up. That rear piece is like a 4 bar link. It'll flop around.

Share this post


Link to post
Share on other sites

I still disagree, if you have the strut connected at the top, and only the front mount of the LCA attached, it will support the weight of the car just fine, you would even be able to drive around on it if it weren't for the fact that you have nothing to control toe. I really don't think that the stock control arm has much torsional rigidity to speak of. I think the strut will see more stress if the 'twist' is resisted by a moment from the control arm rather than a linear force... I could probably prove that by making a shear and moment diagram. I'm probably not going to change anyones mind this way, so I'll just have to prove everyone wrong when I make my own eventually :) however, as long as you guys are making them stronger than they need to be, I won't complain too much. ;)

 

Just stay away from rod-ends in bending please

Share this post


Link to post
Share on other sites
I still disagree, if you have the strut connected at the top, and only the front mount of the LCA attached, it will support the weight of the car just fine, you would even be able to drive around on it if it weren't for the fact that you have nothing to control toe.

Absolutely no argument from me. The issue is that it could be better.

 

I really don't think that the stock control arm has much torsional rigidity to speak of. I think the strut will see more stress if the 'twist' is resisted by a moment from the control arm rather than a linear force...

I'd agree that the stock arm is pretty weak. But the stock chassis is pretty weak too, as are the stock uprights in the rear suspension. Just because they're weak doesn't mean that they weren't intended to do a job that they're really insufficient for. I'm not sure what you're getting at with the last part though. I ran my suspension through it's travel and watched the top of the strut and shimmed the strut back and forth on the spindle pin until I saw that the strut stayed centered throughout the entire motion of the suspension. With that said, I fail to see how a rigid control arm which centers the strut could make the strut "see more stress". Plus, I don't think the force that will get to the strut will be linear at all with the toe adjuster link. I mean, you hit a bump and there is going to be a spike in side force which will tend to make the strut want to bind right at the point where it should be moving. That's my uneducated view of it anyway. Am I missing something?

Share this post


Link to post
Share on other sites

If I were to build rear control arms again (Been there, done that, don't ever want to contemplate the complexities again) I'd start with the OEM rear arm and build off it, and using the "poor" mans toe setup as Jon suggests. The Rear OEM arm is far more durable than it is given credit.

 

Mike

Share this post


Link to post
Share on other sites
Plus, I don't think the force that will get to the strut will be linear at all with the toe adjuster link. I mean, you hit a bump and there is going to be a spike in side force which will tend to make the strut want to bind right at the point where it should be moving. That's my uneducated view of it anyway. Am I missing something?

 

So how does the front not bind solid when you brake and hit a bump? Lower triangles and toe links are used on a number of WRC cars. I figure if it works for the front there's no reason it can't work in back. But I know we've agreed to disagree on this one.

 

Cary

Share this post


Link to post
Share on other sites

I have had great luck with the control arms that I built. They are light and strong. My control arms, the stock control arms, and the ones sold by AZC are all examples of an H-arm strut (Milliken pg 641). The other option would be an A-arm and toe link strut. Each has their advantage and neither is meant to resist the twisting moment applied to the rear suspension when the brakes are applied (That function is performed by the strut).

 

PICT0005.JPG

 

The H-arm strut has a limitation that is described by Milliken, "...the inner bushing pivots must be perpendicular to the axis of motion of the strut at all times or bending of the strut will occur." What this means is that as the control arm rotates about an axis, and the end that attaches to the strut follows the arc of a circle. This is true for all rigid parts of the control arm. With our control arms (H-arm), the entire control arm is rigid and the bottom of the strut is rigidly captured. The strut tube is (supposedly) perpendicular to axis of the spindle pin. So, if the axis of the spindle pin is parallel to the axis of the inner pivot, then the strut will always be perpendicular to the axis of rotation and no binding will occur.

 

Now lets try a mental experiment. Let us adjust the heim in the front to get some toe-in. We have now turned the axis of the spindle pin so that it is no longer parallel to the axis of the inner pivots. This rotates the strut housing by a small amount and because the strut tube is angled away from the center of the hub, the top of the strut will try to rotate toward the back of the car. Well guess what: The top of the strut is captured by the bearing in your camber plates. You just put your strut in a bind. If you have rubber isolators and bushings you will probably get away with this. The rubber will compress before the strut bends. If you have spherical metal bearings everywhere, you will see evidence of the strut binding (springs hitting threaded collars). Interestingly, those cheesy aluminum/delrin eccentric bushings don't cause the same problem because the axis of the spindle pin and axis of rotation stay parallel.

 

Now for the other option: What are the advantages of the A-arm toe link strut? The A-arm, like the H-arm is constrained to rotate about an axis fixed to the chassis. The A-arm however only has one point forced to follow an arc. That rigid point connects to the strut and forces the attach point to follow the same arc. The link between the strut and A-arm is spherical and as such has two degrees of freedom. The connection is free to rotate about the axis of the spindle pin and to rotate about the axis if the strut housing. We need an extra constraint to control rotation about the axis of the strut housing. We need a toe link to do this. This is the important part: The toe link should only control the rotation of the strut about its axis. The toe link MUST only add one constraint. To accomplish this both ends of the tie link must be free to rotate in all planes, both ends must be heim joints. Allowing the toe link to rotate freely allows the strut to rotate about the rigid end of the control arm and prevents binding. If either end is constrained to stay in plane (clevis connection), then we are back to a H-arm and its limitations. Every control arm that I have seen made for a Z of this type was made wrong.

 

So lets try the same mental exercise with the A-arm toe link strut. We adjust to toe link to add some toe-in. We'll use the toe link in the front and the rigid link in the back. The spindle pin is again rotated so that it is not parallel to the axis of the inner bushings, and the top strut tube tries to moved slight back on an arc, but it is constrained at the top by the camber bearing. What happens? The whole strut rotates around the pivot at the rigid point on the control arm. As the strut compresses, the strut houing tilts further and further forward to keep the axis of the strut housing aligned with the bearings at the rigid end of the control arm and the camber plate. Because the toe link is not constrained, the strut housing is not in a bind and the control arm sees no torque.

 

 

So, after all of this I have come to a couple of conclusions:

1. If you have an H-arm set-up, keep the axis of the spindle pin parallel to axis of the inner pivot bushings.

2. I will only adjust the toe of an H-arm set-up using the inner bushings.

3. I am going to build a set of A-arm toelink control arms so that I and have more adjustment possiblility without binding.

 

Damn it, I used to be satisfied with my control arms. Oh well, back to the drawing board.

Share this post


Link to post
Share on other sites

Jon - here's something else to think about and an experiment that I think would help validate your thought. Disconnect the top of the strut from the car. Now push the top of the strut forward and backwards with a reasonable amount of force. How much does the top of the strut move? If it's more than very very little then your control arms are not doing anything to help with rotational loads into the strut. Reason is that the strut/housing would have to deflect that amount before the control arm would even start to take load. If the top of the strut moves very very little then the LCA may take some load. One could argue that a stiff control arm would put more load into the strut. In reality the unibody will move around relative to the LCA. By making the LCA stiffer or not using a toe-link type of set-up you are taking away that degree of freedom (flexible LCA/soft bushings from the stock set-up) and as the chassis flexes and gets out of perfect alignment then it could bind up the strut - similar to what is said above. Not saying in the real world that's going to happen nor do I think there is anything wrong with your design I'm just giving some things to think about and a little experiment in your free time.

 

Cameron

Share this post


Link to post
Share on other sites

74_5.0, I think you just said more clearly what I was trying to say. My point about the strut 'seeing more stress' was that if you simplified the forces the bottom of the strut was seeing to either a linear force (like an A-arm would provide) or a moment (imagine the control arm being replaced with a torsion spring that was free to move for and aft) that the moment acting on the strut to restrain its movement would cause more stress within the strut. I wasn't trying to say that a stiffer control arm would make the problem worse, I was just trying to show one extreme vs. the other. I also was not trying to imply that an H style arm was wrong, just that I beleive it is overkill, and a better strength to weight ratio could be acheived with an A-arm and toe link setup.

 

I might be slightly biased since I welded all of these, but I beleive this is pretty close to an ideal setup:

 

http://img.photobucket.com/albums/v237/Flexicoker/F06/FSAEWest6-14-2006Alex038.jpg

 

All of the bearings are captured speherical bearings, no rod-ends in bending, and the alignment is done with shims. If there was no toe-link (like the bottom arm on our racecar), the inner bearings could be replaced with rod-ends, and they would see no bending loads, however, since the axial load from the toe-link is not in the same direction as the the rod-end is, it would be wise to keep it a spherical bearing. The outer bearing in an A-arm should never be a rod-end, as it will see bending loads, and I think that all 4 points in an H style control arm will see bending loads.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×