no. looks isn't my only concern... thanks. i'm just simply stating the limitations that i experienced first hand by having an inverted shock/true rear coilover on the Z. the shock adjustment **** on my particular coilovers will hit the rear mounting knuckle if the shock tube is lowered too much within the mounting sleeve.

 

I don't believe that that is what Resolute was referring to. You're referring to the fact that an inclined support will need to exert a greater force on a beam than a perpendicular support for a given beam loading. I am familiar with this effect and agree that it is present when mounting the coilover on the upright.

From his earlier post, Resolute seems to think that there is a moment that would tend to force the upright and wheel into a positive camber alignment, and that this moment will increase the wheel rate. This is simply not the case -- while the moment that resolute is talking about does exist, it does not appreciably increase wheel rate. It's main effect is to load the bushings through which the arms mount to the upright.



You said it yourself above. Changes in spring inclination will change the motion ratio of the suspension. As the midlink arcs through its travel, the inclination of spring relative to midlink changes. Even though the spring can bend as an arc, it's force is still applied as a two-force member + a small bending load to account for the spring arc.

The stock setup is designed such that the point where the suspension would become falling rate occurs AFTER the point where the suspension travel expires. However, on lowered cars, the crossover point becomes attainable as the suspension is in a stock compressed state at uncompressed modified ride height. That is why some coilovers with stock spring locations use rising rate springs -- to combat the falling rate effect.


This is true, and is a good point. However, all modern shocks designed to house a coil-on-damper have been sufficiently sized to be able to deal with this bending load, and it shouldn't be a concern that drives people away from coilover setups.

The midlink is designed to be loaded with the stock spring rates + factor of safety. The midlink is aluminum and is succeptible to fatigue failure when it is loaded to the edges of the envelope of it's design. Fatigue failures are more of a risk when a part is loaded in cyclical bending, rather than cyclical tension and compression.

 

No, the wheel rate is altered by the fact that the coil-over is traveling in an arc, and as I stated, this causes the wheel rate to see instantaneous change. It is a function of the cosine of the angle, as I posted. It is also not a factor in the OEM location, as Kuah posted.

As far as moment on the upper mounting bolts and Cosmos- first, don't knock cosmos it's the best I had available Second, it was, as I also mentioned set-up for an Evo and we did not assume a coaxial mount. Hence the bending moment on the upper mounts. If you think no one would be so stupid as to neglect a coaxial mount if they were inclined to put a spring on the damper, then take a look at the cheap coil-over units that have a springs on them.

The change in camber, toe, etc.. from bushing deflection is elastokinematics, and was not mentioned. The effect of the suspension geometry on transient conditions, such as tire loading from roll and pitch moments, is kinematics.

The anti-squat is defined by the inclination of the upper and lower arms as you stated, but the reaction force is a function of torque, which causes bushing deflection and poses a greater effect when the spring is on the damper mount than on the mid-link. There's your elastokinemtics playing a role.

I don't see how you'll run a softer spring on the upright than the midlink, unless you refer to keeping equal wheel rates between the two positions. The change in spring rate that is seen on the upright (as mentioned above) calls for a higher spring to limit travel. Unless you care to just neglect this effect.

Will

 

It most certainly is a factor in the OEM location, as I posted above. The spring and damper travel in an arc in the OEM location, and the inclination angle between the mid-link and the spring does change in the OEM location. It is the same cosine effect that you're mentioning regarding mounting the spring on the upright.

Hahah alright, I definitely hear you on "it was the best I had available" -- I've definitely had similar moments . However, COSMOS results should definitely not be taken as real analysis for large, complex geometry models.

I thought this was a discussion of a well-built street car or race car, not a car built from eBay parts like coilovers with springs seated directly on the upper mount. Any reputable aftermarket or OE coilover or shock worth using will have a pillow ball upper mount with an upper perch on the shock shaft below this point.


Yes, I'm familiar with both of these...I'm not sure what you're getting at.

Anti-squat is not caused by a "reaction torque." Anti-squat is the longitudinal plane version of geometric load transfer vs. elastic load transfer in the transverse plane. The inclination of the a-arms controls the percentage of squat and dive loads that act through the suspension arms and bushings (geometric) instead of the springs (elastic). Increased anti-squat does cause increased bushing deflection, but spring location as nothing to do with this. Spring location will only effect the bushing at the lower mount of the shock, and on a race car or serious street/weekend track car, this will be a pressed-in spherical bearing anyway so I'm not sure how much of a point elastokinematics is in this scenario.

Yes, I was referring to equal wheel rates. And yes, I am quite familiar with using wheel travel in determining ride rate. I was using an all-things-equal model at stock ride height in my post. Spring rate only needs to be increased if static ride height and travel are lowered.

 

If I need an intergalactic star ship built I'm comming to you guys.

Can anyone translate the above into English?

JET

 

If the discussion is of a race car, then reinforcing the mount and the bushing is a given, but I don't think JET is jumping into that arena, so why neglect it as a negative factor in using a true coil-over rear.

Same can be said for the cheap "e-bay parts" coil-overs. Considering most of the threads about a true rear coil-over are about "slamming" the car, why neglect the oft-missing critical component of a coaxial spring mount?

As far as anti-squat, you've either missed my point or not making yourself clear. The arms only determine the reaction point, the instantaneous center, around which the rear suspension will react. So, we have a moment arm that influences the vertical load for a given horizontal drive force. Hence, I used the term, "reaction force." The horizontal force want's to rotate the upright, hence bushing deflection. On a live axle, you get axle wrap. The moment of this deflection is greater the further from the drive center the bushing is, which is worse for the damper than the midlink. This has all neglected static vertical loads. I never said, or meant to convey, that spring location affects anti-squat. The anti-squat will affect the spring, however (vertical force from acting around instant center, and deflection of the mount if the spring is on the damper), and increase the effective wheel rate as a result.

My point from the beginning has been that the Z runs a high amount of anti-squat to begin with, combined with the higher wheel rate that mounting the spring on the damper creates (and we've already gone over about four times the need for a high spring rate to limit the affect travel has on the wheel rate), you get traction under acceleration.

btw, I see your point about the spring still facing an effective loss of rate from declination, but come on- that's miniscule compared to running a spring on the damper.

Will

btw, nice avatar.

edit: oh, the bit about kinematics- you said you weren't sure how this relates to kinematics, and discussed kinematics as the change in toe, camber, etc.. from changes in tire loading. That's the elastokinemtaics. Geometry affects tire loading from various moments that act about the instant centers created by the geometry. That's kinematics, and what I was clarifying.

 

Yes, I did mention this in my very first post as an earlier thought I had, and also posted how it was not found to be true with ADAMS modeling or with Optimum K. I also said that the software is not perfect, and I still wonder how much it affects the wheel rate if a solid bushing is used. Negligible? It appears so from modeling, and I said as much, but it is worth remembering that it is not even a factor in the OEM location.

Will

 

Alright, fair enough -- we can consider need to reinforce the shock tower as a negative.

I'm going to have to disagree with you here, because I don't necessarily feel the need to pander to the "how low can you go lulz" my350z crowd. Any worthwhile aftermarket kit will have a spherical bearing upper mount.

I think we're suffering from an inability to communicate -- can you sketch a quick free-body diagram of what you're trying to describe? I understand that a rotational load on the upright will deflect the bearings, including the lower shock mount, but I'm not sure that you're correct in there being a rotational load on the upright caused by acceleration and anti-squat. With anti-dive, there is a rotational load on the upright caused by the brake reaction torque of the upright, but this does not exist with anti-squat.

Are you talking about a moment trying to rotate the upright around the longitudinal instant center of the rear suspension (the instant center of the side view swing arm)?

As far as I have read, there is no resultant rotational torque on the rear uprights during acceleration with anti- geometry.


If the basis of your argument against spring-on-damper is that rotation of the upright will "lock" the bushing in the lower mount, then come on -- if the rear suspension arms are inclined, that lower mount has already been designed with off-axis and out-of-plane deflection considerations.

I disagree -- the effect will be on the same order of magnitude as the cosine losses of mounting the spring on the damper!


Thanks!

Ah, I think you misundersood me. I was saying the effect on alignment setting by moving the suspension through it's arc (camber gain, toe change/bump steer, dynamic caster) is called kinematics (same thing you're saying). I wasn't saying that pneumatic tire effects were called kinematics.

 

Are you using ADAMS Car (vehicle sim) or just regular ADAMS (analysis of links & pivots)?

Try WinGEO by William Mitchell. It's really good stuff, even the demo. It considers force-based roll centers in the new version too.

 

Thank you Will. What is not taking into account is the presumption that a a true rear coilover has no place. Of course it does, and of course it can be done with probably good results. However, what has been said from the beginning, both in this, and all the other discussions we've had about it over the years, is that there are many other changes that must be done in order to correctly implement such a setup. No one in this discussion owns a full out race car, and no one touting the $1000 'true' coilovers have made any such changes either. So like anything else you end up with a bunch of "I use it and it's great" posts, vs "I use it and here is what I changed in order to be able to do so properly, and here is what I have found vs my previous setup". In other words, no one on these boards doing it, is doing it correctly.

Not only that, but if you follow the camber curve (and the resultant toe curve too) of a true rear coilover setup vs the traditional spring separate from damper, you might see a bit more what Kuah and Will are talking about. That arm is designed to be loaded. This is exactly why the JUN car (which uses a modded rear cross member BTW), and the 380RSC (also using a modded rear cross member), are both using adjustable lateral links (ie camber arms) for both of the positions. These arms stay parrallel to one another during bump and rebound (ends up looking exactly like a Subaru in the rear actually). You'll also notice that both of these setups are using additional reinforcements from the cross member to the hub area to ensure that there is no unwanted deflection during the stroke events of the shock. I am sure this also helps with potential cracking of the rear subframe that can occur as well.

So yes, the true rear coilover can be done, and will have certain benefits vs the conventional setup, but there are many precautions that need to be taken in order to implement it successfully, and those changes are far beyond what anyone here has done. Same could be said for any other substantial geometry change you could do. If anyone was so inclined, you could do inboard shocks ala Carrera GT, but I don't think anyone would debate the various changes needed in order to successfully implement it - same is said from any design that differs so substantially from stock.

Here is a nice rear shot of the 380RS-C - you can see how substantially changed it is from a std. unit

 

in the shot of the 380RS-C did they install two differential rear bushings? I noticed a hole in the sub frame in the same reverse offset position as the stock bushing is when I pressed in my SPL solid mount this weekend.

and does anyone else (I saw Cusco had a bunch of arms) make a replacement mid link like the Jun and Nismo cars have? for those of us running a true rear coilover? would be good to save some weight

 

LOL, r34 you're killing me. That's the whole point of the this thread. It's not about what CAN be done to the rear suspension. This thread is about the pros and cons of mounting the spring on the rear damper. The burden is to show how mounting a spring on the damper is better than OEM. I think, and have posted a few times now, that the only benefit is the lack of motion ratio and lower unsprung weight. Neither of which provides more benefit than the negatives created from needing proper mount reinforcement, a stronger bushing, and less anti-squat to allow a higher spring rate which is needed to minimize the negative effects of traveling in an arc. All of which CAN be compensated, but are not inherently a given when you mount the spring on the damper, and (more importantly), are not even needed if you just keep the spring in the OEM location.


You're telling me. I think the same things have been said at least three times, in three different ways.

A picture is worth a thousand words and would make this a lot better discussion, you're right. But, I don't care to take the time. The fact is, with 20k worth of software licenses on this station, I just now realized that not a single one is for illustration. We have a license for AutoDesk on one of our DOQ stations, but I'm not going to hijack it for an internet thread. Although, since you thought doing FEA in Cosmos was funny, I am tempted to see if you get a kick out of vector force diagrams of the rear suspension done in Paint. Besides, it looks like you got what I'm saying, or at least trying to convey:

We're talking the same things, just different aspects. I was responding to your question about anti-squat being a "reactive force". I'm not sure where the disconnect is in our discussion of this, but you obviously have to realize that the amount of vertical force created by the anti-squat geometry is a result of the amount of acceleration (horizontal drive force). Obviously, the anti-squat does "squat" (bad pun) in static. That's all I'm trying to clear up here. If you think that the wheel rate, and the vertical force acting against the spring, is somehow not affected by this drive force, then... well I'm not sure what to say if that's the case.

Not with an independent suspension, there is not. What I am referring to, is that when accelerating, the upright travels through it's motion as designated by the anti-squat geometry (looking at the upright from the side of the car). The upright moves with the center rotating about the instant center created by the anti-squat, correct? The position of the damper mount and the midlink mount are not equa-distant from the upright center. The resultant vector of the midlink mount moving through this arc is vertical. The resultant vector of the damper mount is inclined towards the front of the vehicle.


Now it sounds like we're on the same page, but I disagree with your assessment. The mount was designed with consideration to the upright's travel, but not while carrying the sprung mass of the vehicle. That is the issue! There is a moment created on that bushing from the severe amount of anti-squat the Z has, and it is designed to have enough deflection for the stock OEM design , but what happens when it is loaded from a spring? Now if we consider you machined a solid Al bushing for durability, then you remove the most of the bushing's ability to deflect, and so you are still back to running a high spring rate to minimize the binding the anti-squat will deliver with long travel. The high spring rate combined with high anti-squat (vertical forces increasing effective wheel rate under acceleration) means you have traction issues coming out of corners, esp tight slow corners where there is no areo benefits. Hence why Nismo T2 springs and GA Cup teams all run between 500 and 600 lb springs in the OEM locations. Pretty low wheel rates, huh? But needed with the anti-squat. If you mount a spring on the damper, then you have issues with the bushing created by the amount of arc the anti-squat geometry moves the upright through, and you'll want to run a higher wheel rate to limit travel, which loads the tire too much for decent traction. Hence, why I've said repeatedly that to use a spring on the upright, I would recommend removing or decreasing the amount of anti-squat. But the point in regards to this thread- It's one more thing that doesn't need to be considered if you keep the spring on the OEM midlink.


This is the only thing I think we really disagree on, and I don't get what you're saying to be honest. The spring in the OEM location sees essentially vertical travel, even with the midlink traveling in an arc. I can see what you're saying, that the force applied by the spring will be affected by the angle of travel, but I think by a miniscule degree. By contrast, the spring on the damper sees an increasing angle as it goes through it's travel, so the instantaneous force applied to the wheel from the spring will decrease as the spring moves through it's arc. Spring rate multiplied by the cosine of the angle, results in a decreasing rate as the spring gains camber. This is considerable compared o the OEM location, and the major crux of the spring on damper design. From what I gather though, you are saying that the same effect, to the same degree, occurs in the OEM location as on the damper? Is that what you're saying, and if so, care to elaborate?

From all that has been discussed, I still don't see a benefit to the spring on a damper. For a desired wheel rate and frequency, we could run an appropriate spring in the OEM location and call it a day. Or we could run an appropriate spring rate on the damper and thus be forced to make changes to the upright bushing, mount, and geometry before we realize the benefits of doing so. It CAN be done, no one is arguing that. But no one in this forum is running a race car. What is being asked is whether it is better to use the OEM design, which can be assumed has all the FEA that a manufacturer has available to do for designing the system, or to use a true-coil over in the rear with the stock bushings, mounts, and geometry.

Will

Btw, regular ADAMS. I don't have any say in what programs are purchased, but your suggestion sounds nice. I'll look into it and pass it along for kicks and who knows what might happen. Thanks for coming out of the lurking pool for this thread. It's been fun.

 

Not sure on the rear mount. I've never had any need or desire to inspect the stock one with any detail unfortunately, so can't help ya there. Looks like some sort of mustache brace with a bushing on either side.

No one makes the parts used on those cars, they were done in house by their respective builders. Wouldn't be hard to make, but as mentioned, that arm is the least of the issues that need addressing when running the true rear coilover

 

very true.

I was also wondering if anyone makes a replacement bushing for the lower damper mount on the spindle. if not, would it be a solid bushing or a telescoping bushing that would need to be made. hmm. oh well.

 

I'm not exactly sure what you're getting at. What do you mean a true rear coilover has no place? How can a true rear coilover have no place if it's being campainged successfully on both the JUN and Nismo cars?

It sounds to me like you're saying that because nobody is doing it right on this forum, it's not a valid setup or has no place on our cars? How does that make any sense? That's just catering to people who don't want to take the extra steps that me and will have been saying are necessary to campaign a true coilover rear (reinforced upper strut tower, captured spherical bearing in the lower mount, using a coilover system with a pillow ball upper mount, etc.)



The camber and toe curves are not effected by spring placement. These curves are determined by the geometry of the upper arm, lower arm, radius arm, and mid-link. Spring packaging can change the way the bushings deflect on the upright, but on newer cars with fresh bushings, this is a negligible effect.

The stock mid-link is aluminum and was designed for the static and dynamic loads with a ~400 lb/in spring. Stiffer springs innately transfer more energy into the mid-link and upper mount in reaction to road loads. When metals are loaded outside of their designed loads repeatedly, the metal can develop small internal cracks and voids which propagate and ultimately lead to a loss of stiffness and ultimately, fatigue failure -- this is a valid concern in racing and is especially the case with aluminum components.

There aren't "many", there are a couple easily done ones. Most importantly, spherical bearings in the upper and lower mounts. The most difficult task is reinforcement of the upper shock tower area, which can be done by tying the shock towers into a roll cage. Any owner that seriously tracks his or her car should consider a cage for safety in the case of multiple rollovers.

 

I want to look into getting the spherical bearings in my lower mount for saftey.

could I find an off the shelf bearing that would fit the location? how would someone go about doing this?

 

You need to re-read what I wrote

What I said, if you took the time to read it, is that it could probably be made to work quite well...if the correct measures are taken. Such steps are not anything that anyone here with a true rear coilover has done. If you are building a LeMans class car and want to re-engineer the wheel to use such a setup, be my guest. Seems to me that doing all that work to get a set of $1000to work (which is what basically everyone here with a true rear coilover setup is running) the way they should, is an excercise in futility. Doing something different that works is great - that's called innnovation. Doing something different without understanding the net changes you are making, or worse, without even taking the potential downsides into consideration while simultaneously thinking that you're making something better, is called being naive.

 

I remember when I used to heat my springs

 

Changes in spring inclination does not necessarily change the motion ratio. That clearly depends on the suspension geometry. The motion ratio is the ratio relating the vertical wheel travel to the compression of the spring. If you model the spring, you should see that the vertical compression of the spring at the midlink (as taken at the midpoint of the spring) is linearly related to vertical wheel travel.

The inclination of the midlink does not change that. The bowing/bending of spring is a different effect from changing the mounting angle of the spring. Some bending/flexing of the spring always occurs, even in a coilover application, and this effect is always considered to have minimal impact on the linearity of the spring. I disagree with you that this effect is on the same order of magnitude as the change of inclination angle of the coilover. In addition, it is always possible to add something similar to the Hyperco hydraulic spring perch to eliminate the bending force applied to the spring.

The static loads are not affected by the spring rates, though the dynamic loads will be. But the stock shock mounting points are not designed for a spring in the first place, so if stiff springs generated so much more load, I would think it will be a lot worse to run them on a coilover, than on the midlink which was designed by the manufacturer to handle the sprung loads in the first place, with some safety margin built in. And given the motion ratio, the road loads will be reduced at the midlink as compared to at the coilover.

Also the stock spindle/upright is aluminum, by virtue of this argument, we are just as likely to see complete failures of the spindle. All in all I don't think this is a fair argument against the use of a midlink...

 

Great thread . . . but you guys are making my head hurt so early on Monday morning!