Arizona ZCar Rear Suspension - Design, Function, Flaw(?)

[JMortensen]

Race Car Vehicle Dynamics by Milliken and Milliken. I think I typed out the section that talks about H arms in the "yet another control arm" thread.

 

While we're at it though, your control arm design doesn't make a lot of sense to me. It looks like you've got a longitudinal pivot on the rear of the arm, but not the front. If the front joint is built onto the control arm itself as it appears to be in the drawing, then you have a pivoting, adjustable length "extra" link that you can never adjust and which can never pivot, assuming the rear part of the arm is welded together. You also have no front pivot for the arm. Would you run a stock bushing in front?

 

If the extra link wasn't attached to the control arm but to the chassis, then the ball joint would swing straight up and down, but the extra link wouldn't. It would have to twist in relation to the rest of the arm. So as the control arm moves in your design, the rear part of the arm with the bj moves up and down, and then the link pivots on a different axis. This would cause toe changes (bumpsteer). What you need is pivots at the points of the one piece triangle, and the front and rear to be inline if you don't want bumpsteer. 

[logr]

It was a quick thought but after reading some of the threads mentioned on here, I see it was thought of a long time ago. I agree the inner pivots have to be on the same plane.

[G-E]

What are your thoughts on something like this:

 

 

Ignore the over-simplification of the pivots, it's the layout I'm focusing on...

[JMortensen]

What are your thoughts on something like this:

 

s30-split.jpg

 

Ignore the over-simplification of the pivots, it's the layout I'm focusing on...

My take: That front toe link, if jointed, now has a bumpsteer effect. The shorter the toe link the worse it's going to be. If not jointed, you have an H arm that isn't all one piece, so essentially a weaker H arm. Also the rear pivot on the control arm is bolted to those sheet metal uprights, and I'd be wary of disconnecting the front and rear inner pivots. 

 

RCVD has a "reversed A-arm and Trailing Link" in their book. I am pretty sure this design is used on a lot of Subarus, and is very similar to what you've drawn, except the trailing link is essentially straight longitudinally. If you were to do this on a Z, you'd have the trailing link pretty far inboard, which is not ideal. Anyway here's what they say about all of these variations:

 

 

 

Reversed A-arm and Trailing Link Strut

 

As the name implies, the kinematics involve an A-arm with the point of the A at the inboard attachment to teh chassis and the two legs attaching to the knuckle. A trailing link runs longitudinally from the chassis to the A arm or the knuckle. The toe is controlled by the A arm and its single attachment point to the chassis. This results in a toe curve very similar to the semi-trailing arm suspension where it will toe-in with both bump and droop travel. This design has good camber stiffness and fore-aft stiffness with a minimum number of chassis attachments spread over a wide base. The ability to obtain desired values for roll camber, roll center height and anti properties are quite free within the general limitations of basic strut suspensions as discussed earlier. Another negative feature is that if rubber bushings are utilized at the trailing link pivots, toe-out will occur whenever a rearward load is applied to the tire such as braking or hitting a bump. the side view instant center will move with the wheel travel because of the trailing link, thus changing the anti features with wheel travel.

 

A-arm and Toe Link Strut

 

There are two potential versions of this rear suspension, one being exactly like a typical front suspension but with a toe link to a chassis pivoted ground. In terms of general geometry, this type of design is very flexible with good control over the side view geometry, excellent toe control capability, and only the standard strut limitations in the front view. The rate of change of the side view geometry is very controllable. 

 

The second version of this type has the toe link attaching back to the control arm rather than the chassis. This design eliminates one attachment point at the chassis. It is possible to obtain favorable toe control with this system and all other geometry requirements can be met similar to the first basic version. One of the reasons this suspension is used is that the ride steer curve is less sensitive to build variations than with the standard toe link attached to the chassis.

 

H-arm Strut

 

An H-shaped arm having two bushings at each end will kinematically perform the jobs of three links similar to the A-arm and toe link. There are limitations, however, when links are combined into structural members. In this case the axis of the inner bushing pivots must per perpendicular to the axis of motion of the strut at all times or bending of the strut will occur. With this limitation comes some side view geometry limits that cause long side view swing arms and very low values of anti-lift. It is unavoidable with this design.  Roll steer is obtained by having the two bushings at the knuckle end out of plane. In other words if you think of the inner two bushings and one of the outer bushings representing three points that determine a plane, the fourth bushing intentionally does not lie in that plane. As the control arm pivots about the two inner bushings the two outer bushings that started out at different height, and therefore as they rise in their arcs the amount that they move inboard is different. By controlling this difference in inboard motion the amount of toe change with wheel travel can be controlled.

 

One negative factor of this design is that any flex or distortion in the arm or distortion of the bushings due to braking loads or bump loads causes side loading of the strut. With the strut side load comes friction or resistance to axial motion. The suspension cannot isolate and perform all of its other functions when friction is present. This suspension has been thought of as a lower cost version of the A-arm and toe link. It almost is, but not quite because of this potential for causing friction.

[rsicard]

Not certain whether Race Car Vehicle Dynamics by Milliken and Milliken is totally applicable to the Datsun 240Z rear suspension.  Some subtle details overlooked.

[JMortensen]

Would you like to clue us in?

[G-E]

My take: That front toe link, if jointed, now has a bumpsteer effect. The shorter the toe link the worse it's going to be. If not jointed, you have an H arm that isn't all one piece, so essentially a weaker H arm. Also the rear pivot on the control arm is bolted to those sheet metal uprights, and I'd be wary of disconnecting the front and rear inner pivots. 

 

RCVD has a "reversed A-arm and Trailing Link" in their book. I am pretty sure this design is used on a lot of Subarus, and is very similar to what you've drawn, except the trailing link is essentially straight longitudinally. If you were to do this on a Z, you'd have the trailing link pretty far inboard, which is not ideal. Anyway here's what they say about all of these variations:

 

This isn't a bad thing like you think it is, the toe curve would be fully tunable by using simple spacers: anything from toe-in on compression to toe-in on droop

 

Secondly, while I didn't draw a connecting piece, I envision a whole replacement bar that bolts up to the factory mount, which would have balljoint bosses for the toe rod and reverse a-arm to pivot from.

 

If the balljoints were facing down from above the new plate, you'd have increased roll resistance; if the balljoints were under, you'd have the opposite, and the bar could be symmetrical to allow being flipped, or even have several potential boss locations for even more toe or wheel position adjustment

 

The toe rod would connect to the arm with a rodend, perhaps studded rodend to make life easier to insert spacers...

 

Makes sense?

[rsicard]

Would you like to clue us in?

Bump Toe in/out, accel and decel, strut extension/compression.  The more parts used in the design, the greater number of points of failure.

[JMortensen]

This isn't a bad thing like you think it is, the toe curve would be fully tunable by using simple spacers: anything from toe-in on compression to toe-in on droop

 

Secondly, while I didn't draw a connecting piece, I envision a whole replacement bar that bolts up to the factory mount, which would have balljoint bosses for the toe rod and reverse a-arm to pivot from.

 

If the balljoints were facing down from above the new plate, you'd have increased roll resistance; if the balljoints were under, you'd have the opposite, and the bar could be symmetrical to allow being flipped, or even have several potential boss locations for even more toe or wheel position adjustment

 

The toe rod would connect to the arm with a rodend, perhaps studded rodend to make life easier to insert spacers...

 

Makes sense?

I think you'll find that you end up with a lot of bumpsteer with a toe link that short. Even if you can choose whether it's toe in or toe out, you don't want a lot of bumpsteer. You might mock it up real quick to see what it does in practice.

 

The new end for the strut sounds like it might enable you to twist a rod end 90 degrees and then install it in double shear with the bolt going through vertically, which would be preferable to a single shear ball joint IMO. You could space out your mounts and then add spacers if roll center adjustability was the goal.

 

If your ball joint was on the bottom, that would raise your RC and increase roll resistance.

 

 

Bump Toe in/out, accel and decel, strut extension/compression.  The more parts used in the design, the greater number of points of failure.

Help me out here. This relates to which design? I'm not saying that RCVD tells you how to build Z suspension, but I can't yet see where the info there isn't applicable to general suspension designs that you might put under a Z.

[rsicard]

Was referring to a lower control arm design that has one rigid leg and the other is a two piece leg with standard inboard and added mid rotatable joints.  Too much flex, changing toe in/out with accel/decel and strut extension/compression, with the mid rotatable joint.

[G-E]

 

I think you'll find that you end up with a lot of bumpsteer with a toe link that short. Even if you can choose whether it's toe in or toe out, you don't want a lot of bumpsteer. You might mock it up real quick to see what it does in practice.

 

The new end for the strut sounds like it might enable you to twist a rod end 90 degrees and then install it in double shear with the bolt going through vertically, which would be preferable to a single shear ball joint IMO. You could space out your mounts and then add spacers if roll center adjustability was the goal.

 

If your ball joint was on the bottom, that would raise your RC and increase roll resistance.

 

Actually you have the RC mixed up, when I say balljoint down, that means the center of the ball is below the bar, hence a low pivot and decreased roll stiffness...  when I say balljoint up, I mean upside down, so the ball center is above the bar. Of course I'd use a balljoint that's retained by a side bolt or pin, not the press-in kind, so it would be up to the end user to chose how much they want.

 

But like I said, I wouldn't worry about bump toe, because you could align it to be zero too, you'd have full control of aligning it's axis to the a-arm's axis, and since we know the strut won't let the a-arm twist up or down, that precludes any complex motions too.

[JMortensen]

If I'm understanding you correctly, you're confused about how roll center works. If the car is raised up a lot as an example, you have a steep downward angle from the inner pivot to the ball joint and a very high roll center. If the car is really low, you might have the control arms pointing up from the inner to the ball joint. This would have an underground roll center. "Bumpsteer spacers" raise the roll center by lowering the outer pivot. 

 

See this link for more: http://www.e30m3project.com/e30m3performance/myths/Weight_Transfer/weight_transfer2.htm

 

Here's one instructive image: 

 

You cannot possibly align that toe link to have zero bumpsteer. Unpossible.

[G-E]

If I'm understanding you correctly, you're confused about how roll center works. If the car is raised up a lot as an example, you have a steep downward angle from the inner pivot to the ball joint and a very high roll center. If the car is really low, you might have the control arms pointing up from the inner to the ball joint. This would have an underground roll center. "Bumpsteer spacers" raise the roll center by lowering the outer pivot. 

 

See this link for more: http://www.e30m3project.com/e30m3performance/myths/Weight_Transfer/weight_transfer2.htm

 

Here's one instructive image: 

 

You cannot possibly align that toe link to have zero bumpsteer. Unpossible.

 

No, I'm pretty clear...  why don't you tell me what the difference between lowering the outer pivot or raising the inner pivot is?

[JMortensen]

The imaginary line is drawn through the outer and inner joints, so if you moved them to the same angle and height, there is no practical difference whether the inner is lower or the outer is higher or both. The instantaneous center would move to the same place. If you moved the outer lower without changing the inner, the starting height of the line would be lower and the instant center would be further away than if you just moved the inner.

 

I see the problem though. I thought you were talking about a bar which would bolt to the strut, but you were in fact talking about the inner pivot. You are correct. Moving the inner up would raise the RC. 

 

[G-E]

Now that we have THAT sorted out, any other thoughts?

[tube80z]

What are your thoughts on something like this:

 

s30-split.jpg

 

Now that we have THAT sorted out, any other thoughts?

 

From your picture and the following description it sounds like you want flexibility in being able to have a non-zero toe curve and the ability to tune it by using spacers on the toe-control link.  The only downside I see is if you stick to something similar to the reverse a-arm design in your diagram you give up being able to also adjust anti-squat and lift.  This assumes you keep the strut and don't add an upper a-arm.  Is this going to be a strut or an a-arm?

 

The other thing to check is if this design is not as stiff as using a reverse a-arm.  To check that on the car you use some method of pulling the wheels together with the vertical position locked.  Then repeat by pushing.  Then switch to pulling/pushing the same side wheel pair.  This is a poor mans K&C rig and you can see camber and toe changes from force at the contact patch.  I think you could do something similar in a CAD package to at least see what the forces were in the members.  If no difference that's great.  If you end up putting a lot more load on a single member or joint then you might have problems.

 

Happy to see a more technical thread in this section.

 

Cary

[G-E]

From your picture and the following description it sounds like you want flexibility in being able to have a non-zero toe curve and the ability to tune it by using spacers on the toe-control link.  The only downside I see is if you stick to something similar to the reverse a-arm design in your diagram you give up being able to also adjust anti-squat and lift.  This assumes you keep the strut and don't add an upper a-arm.  Is this going to be a strut or an a-arm?

 

The other thing to check is if this design is not as stiff as using a reverse a-arm.  To check that on the car you use some method of pulling the wheels together with the vertical position locked.  Then repeat by pushing.  Then switch to pulling/pushing the same side wheel pair.  This is a poor mans K&C rig and you can see camber and toe changes from force at the contact patch.  I think you could do something similar in a CAD package to at least see what the forces were in the members.  If no difference that's great.  If you end up putting a lot more load on a single member or joint then you might have problems.

 

Happy to see a more technical thread in this section.

 

Cary

 

Well the ability to control the toe curve, whether you want it to change in a specific direction or not change at all, I think you'll agree there's use cases and driving styles for each possible setup, and I think the fine adjustment just adds another tool to the arsenal.

 

Both inner pivots are balljoints in my diagram, and the middle pivot is a rodend or uniball as well, all oriented horizontally, they would all act as a solid member in the horizontal plane. The only additional force added would be rotational to the inner pivot bar, just like the azc kit with the tabs hanging off the bar, but unlike the azc kit, I would go with a solid mounted inner bar.

 

The strut isn't really relevant to this aspect of the discussion, though I'd obviously aim toward using my s13 style coilover mounting system, that would allow the camber adjustment, but still operate as a chapman strut.

[zack_280]

To answer the original question...You have to put a set screw in the AZC rear LCA.  Maybe some setups will be less prone to movement, but mine has pretty tight clearances. At an autocross my outside rear wheel/tire was pushed into my inner fender due to movement of the suspension.  I tried to jack the car up and drop it quickly, but it wouldn't move back to the original loaded position.  I had to add a set screw. 

 

I discussed this with CobraMatt and he said he installed a set screw from day 1 to keep from having alignment issues.

Edited August 28, 2014 by zack_280

[NewZed]

I wonder if this guy knows about the set screw - http://www.importtuner.com/features/1408_mti_racing_1971_datsun_240z/viewall.html

 

Might be important at speed.  Car specs. at the end of the article.

 

 

Can't replace the lost week's discussion (a shame), but still interesting to follow where the design is being used.

[G-E]

The only parts lost were drilling and through-bolting the AZC bar and mount to lock it in place...

 

And of course JM's mistaken belief that my design in progress has some unnatural toe change

 

I am still trying to figure out a few key points I can improve on the cars at the same time, since my design doesn't include track width changes as it stands, I have yet to come to a conclusion if this is important, assuming the reason to narrow everything is eliminated...

 

Also I could split the balljoint block from the V, by adding captive shims, things could be lengthened fairly easily, but then what's the minimum length I should account for? Are people interested in widening the rear track width if strut clearance is improved equally for higher offsets?