Higher spring rates w/o rear bar or not?

[yokoseiki]

I think you're right, but for the most part folks here are talking application and not theory. Whatever theories they share is just how they have the logic plotted out in their heads for tuning and probably has no basis on kinematics. For the most part folks are "right" in that the side that is somewhat "stiffer" will offer less grip whether that's in pitch or roll. In addition forks are correct in assuming that removal of a roll device like a sway bar will change the balance and appropriate change in spring rate is probably necessary to bring that balance back.

[albrazzi]

Wondering out loud, if you consider the positive affects of a roll center correcting kit I think the discussion would be different, it just depends on how far you want to take it. In OEM form if you use some restraint in lowering the benefits of hard springs and swaybar on the front speak for themselves, mainly because you are controlling the Camber curve. If you add roll center correction the Camber curve is on your side, so some things change. Harder springs dont require the same rollbar that they did without roll center correction and you are in a different setup mindset.
Like I said just wondering out loud.

[illinipo]

Obviously if your sanctioning body and class allow roll center kits, you should use them. You can lower the car AND get the control arms pointing the right way, and the bonus points are not needing a bunch of spring or bar to get the roll resistance you need

Forgive me for not knowing your understanding of all this as this may be taken as talking down to you, but the point of the roll center kits is to bring the roll center closer to the CG. A lighter spring with a shorter moment arm has the same effect as a stiffer spring with a longer moment arm (distance from RC to CG). This is why you can use ride heights to adjust the balance of the car instead of bar holes.

For DE (this thread, LOL)... not really worth it.

For autocross, roll center correction is only allowed in really high end classes. I imagine the same is true for NASA/BMWCCA/SCCA...

[sped here]

Running no rear bar is more commonly seen on an autocross course.

I personally haven't seen anyone running track events without a rear bar. I'm sure it's possible, just not very common.

Depends... is it raining?

[MJockey]

My car has JRZ RS1s, 680#F, 780#R, Eibach Adjustable Sways, 26mm Front(full stiff) and 24mm Rear(full soft). I run 245mm tires F & R. I would like to dial out some of the oversteer on corner exit, to allow me to put the power down sooner. I picked up a stock 328i 15mm rear bar at a salvage yard. As far as end links go, is my only choice the stock rubber/aluminum ones?

[illinipo]

My car has JRZ RS1s, 680#F, 780#R, Eibach Adjustable Sways, 26mm Front(full stiff) and 24mm Rear(full soft). I run 245mm tires F & R. I would like to dial out some of the oversteer on corner exit, to allow me to put the power down sooner. I picked up a stock 328i 15mm rear bar at a salvage yard. As far as end links go, is my only choice the stock rubber/aluminum ones?

If the steady state balance is already acceptable, you might instead consider increasing rear toe in?

[SG_M3]

If the steady state balance is already acceptable, you might instead consider increasing rear toe in?

exactly, MJockey how much rear toe are you running?

[yokoseiki]

My car has JRZ RS1s, 680#F, 780#R, Eibach Adjustable Sways, 26mm Front(full stiff) and 24mm Rear(full soft). I run 245mm tires F & R. I would like to dial out some of the oversteer on corner exit, to allow me to put the power down sooner. I picked up a stock 328i 15mm rear bar at a salvage yard. As far as end links go, is my only choice the stock rubber/aluminum ones?

You asked a question and got a few more questions as answers, I like J/K

But to actually answer your question, no you don't have to use the stock ones.

Turner has adjustable ones like this http://www.turnermotorsport.com/p-53...inks-pair.aspx

Also you can get male and female spherical rod end bearings like these https://www.pegasusautoracing.com/gr...GroupID=RODEND and make your own if you don't like the rubber joint.

[MJockey]

Thanks for the replies.

exactly, MJockey how much rear toe are you running?

I'll have to check when I get home. How much would you recommend?

But to actually answer your question, no you don't have to use the stock ones.

Turner has adjustable ones like this http://www.turnermotorsport.com/p-53...inks-pair.aspx

These say " Note: these links will not work with factory or non-adjustable Eibach sway bars."
Remember the stock bar has no holes at the ends.

[SG_M3]

I'll have to check when I get home. How much would you recommend?

3/16" total toe in. 0 toe in my experience leads to on throttle/exit oversteer.

[MJockey]

3/16" total toe in. 0 toe in my experience leads to on throttle/exit oversteer.

The alignment sheet says .29 degrees total toe. That said, it's all on the left (L 0.23, R 0.05). Looks like it's back to the alignment shop.

[yokoseiki]

Thanks for the replies.



I'll have to check when I get home. How much would you recommend?



These say " Note: these links will not work with factory or non-adjustable Eibach sway bars."
Remember the stock bar has no holes at the ends.

OMGosh I didn't read your entire post. The easiest thing to use are a couple of split collars with a tab tig welded onto them. Then you would attach the rod end to that. Makes sense?

Personally I would weld 2 tabs per side so the bolt ends up in double shear. A single tab could induce a moment on the bar and twist it.

[wazzu70]

There seem to be some basic principals overlooked in this thread. It was mentioned before, then glossed over. There is a significant difference in a car that gets all its roll resistance from springs and one that gets assistance from an ARB. The difference comes in heave (up/down). If you arent on a very smooth track and you get all your roll resistance from spring rate...have fun on that ride! With very high spring rates your car is also overly upset when you hit the kerb.

Since most tracks I have been to would not fall in the smooth as glass category, I wouldnt consider getting all my rate from springs alone in the front.

Due to the nature of MacPherson strut geometry, especially on a lowered car (roll center displacement) an ARB is almost essential. The spring rates you would need to run to replace the roll resistance provided by an ARB would be way too high to be practical. Even with roll center correcting ball joints, an ARB still makes a lot of sense for the same reasone even though they are less influential.

On the front end there is also the misconception you get better cornering grip with full independant suspension (no ARB). Yes you get more load transfer to the outer tire, but is front grip best when outer tire grip is maximised? What if you had the front weight split 80% to the outer tire and 35% on the inner tire....thats greater grip than 100% on the outer tire!

The rear end is different as there are many possible scenarios. You cannot at all compare an E30 semi trailing arm rear suspension to the E36 and similar rear suspension. Both behave very differently! On the E30 its common to run a light rear bar or disconnect the bar to keep the inner wheel on the ground for traction purposes. The E36 rear suspension has a much better ability to roll without lifting the inside wheel so disconnecting the rear bar does not achieve the same effect to a similar extent.

What differential you run also plays a role. Torsion and open (if anyone runs them) cannot put power to the ground if one wheel has no traction. In this case you want to maximise the ability for both rear wheels to be on the ground at the same time. A clutch LSD does not have this issue.

The same principal applies to the E36 rear as the front though. The ARB helps even the load between the sides and in most cases maximises the grip over getting one corner fully planted!


I dont know it all and I am not a professional in this area but have been setting up cars and studying the subject for years. There is a lot that goes into it, but if you mess up some basic principals your logic will set you down the wrong path every time!

Unfortunately there is no magic bullet. The age old answer of "it depends" still applies.

[yokoseiki]

There seem to be some basic principals overlooked in this thread. It was mentioned before, then glossed over. There is a significant difference in a car that gets all its roll resistance from springs and one that gets assistance from an ARB. The difference comes in heave (up/down). If you arent on a very smooth track and you get all your roll resistance from spring rate...have fun on that ride! With very high spring rates your car is also overly upset when you hit the kerb.

Since most tracks I have been to would not fall in the smooth as glass category, I wouldnt consider getting all my rate from springs alone in the front.

Due to the nature of MacPherson strut geometry, especially on a lowered car (roll center displacement) an ARB is almost essential. The spring rates you would need to run to replace the roll resistance provided by an ARB would be way too high to be practical. Even with roll center correcting ball joints, an ARB still makes a lot of sense for the same reasone even though they are less influential.

On the front end there is also the misconception you get better cornering grip with full independant suspension (no ARB). Yes you get more load transfer to the outer tire, but is front grip best when outer tire grip is maximised? What if you had the front weight split 80% to the outer tire and 35% on the inner tire....thats greater grip than 100% on the outer tire!

The rear end is different as there are many possible scenarios. You cannot at all compare an E30 semi trailing arm rear suspension to the E36 and similar rear suspension. Both behave very differently! On the E30 its common to run a light rear bar or disconnect the bar to keep the inner wheel on the ground for traction purposes. The E36 rear suspension has a much better ability to roll without lifting the inside wheel so disconnecting the rear bar does not achieve the same effect to a similar extent.

What differential you run also plays a role. Torsion and open (if anyone runs them) cannot put power to the ground if one wheel has no traction. In this case you want to maximise the ability for both rear wheels to be on the ground at the same time. A clutch LSD does not have this issue.

The same principal applies to the E36 rear as the front though. The ARB helps even the load between the sides and in most cases maximises the grip over getting one corner fully planted!


I dont know it all and I am not a professional in this area but have been setting up cars and studying the subject for years. There is a lot that goes into it, but if you mess up some basic principals your logic will set you down the wrong path every time!

Unfortunately there is no magic bullet. The age old answer of "it depends" still applies.

Actually I think quite the opposite. I think most of the comments made by members here are fairly accurate.

I wouldn't over complicate things. Sway bars add roll resistance while not adding it in pitch/heave. I don't think one person made any comment that deviated from this simple, yet accurate description of their function.

Where things get wonky are what folks think the net effect on "balance" is and how much of a part the bar makes on it. The one comment I have to say is this. On any of the cars we are talking about, the net effect by the bar relative to the wheel rate is minimal. What are we talking about here, 2-3 cm rear bars? Honestly, you can deflect it with your hand. That's why you can remove it and increase the effective roll resistance through total rear platform and not adversely effect the pitch and heave of the car. That's why I removed my bar. I mean its a 2600lb race car with 600lb rear springs (coil over so equivalent to 1200 lb conventional?) That little 22MM bar wasn't doing much. But like I said before I absolutely want to run a rear bar, just with a greater range of adjustment and no binding. It just makes tuning your car so much easier.

[Evice]

I think dual spring setups, (main/tender not helper) are also adding roll resistance.. Something like a negative roll resistance on unloaded tire, becoming positive resistance on loaded side.

[illinipo]

Due to the nature of MacPherson strut geometry, especially on a lowered car (roll center displacement) an ARB is almost essential. The spring rates you would need to run to replace the roll resistance provided by an ARB would be way too high to be practical.

Have you actually done the math found in Milliken? I have, and for certain weight and tire combinations, it is absolutely practical. So practical, that I run on bumpy as hell parking lots without any bars at all. And I do have fun. Allow me to reiterate,

The easy way to pick rates is develop an understanding of how high of ride frequency you can get away with on various surface sticky-nesses and personalities. Sometimes you just cant get down to 1-1.5 degrees per G in roll while keeping a low enough ride frequency for the racing surface... then you need to run a bar.

If you can get the roll gradient you want without a bar, without exceeding an acceptable ride frequency for your tire and the surface you run on, no need for a bar! That is directly out of Milliken, and just plain makes common sense.

On the front end there is also the misconception you get better cornering grip with full independant suspension (no ARB). Yes you get more load transfer to the outer tire, but is front grip best when outer tire grip is maximised? What if you had the front weight split 80% to the outer tire and 35% on the inner tire....thats greater grip than 100% on the outer tire!

I'm not sure I totally understood what you were getting at here. Why would you get more load transfer to the outer tire without an ARB? The situation is actually the other way around, an ARB increases load transfer. That is why when you stiffen the ARB on the front, the car understeers more, and when you add more rear bar, it oversteers more. It's causing more weight transfer relative to the other end and reducing the cornering power of that end of the car.

The rear end is different as there are many possible scenarios. You cannot at all compare an E30 semi trailing arm rear suspension to the E36 and similar rear suspension. Both behave very differently! On the E30 its common to run a light rear bar or disconnect the bar to keep the inner wheel on the ground for traction purposes. The E36 rear suspension has a much better ability to roll without lifting the inside wheel so disconnecting the rear bar does not achieve the same effect to a similar extent.

What differential you run also plays a role. Torsion and open (if anyone runs them) cannot put power to the ground if one wheel has no traction. In this case you want to maximise the ability for both rear wheels to be on the ground at the same time. A clutch LSD does not have this issue.

The same principal applies to the E36 rear as the front though. The ARB helps even the load between the sides and in most cases maximises the grip over getting one corner fully planted!


I dont know it all and I am not a professional in this area but have been setting up cars and studying the subject for years. There is a lot that goes into it, but if you mess up some basic principals your logic will set you down the wrong path every time!

The fact that clutch LSDs can still put power down with a wheel in the air does not make it okay for one to set up a suspension to lift a wheel.

The ARB does the opposite of evening the load between the sides. It actually increases load transfer.


...I suggest you re-read the sections of Milliken covering load transfer before your logic sets you and anyone reading your posts down the wrong path.

[MJockey]

Have you actually done the math found in Milliken?

...I suggest you re-read the sections of Milliken covering load transfer...

What is this "Miliken" you speak of? I did a google search but didn't see anything related to car setups.

[kenndoggy]

What is this "Miliken" you speak of? I did a google search but didn't see anything related to car setups.

http://www.amazon.com/Race-Car-Vehic...&keywords=RCVD

[MJockey]

http://www.amazon.com/Race-Car-Vehic...&keywords=RCVD

Thanks, Would you say it worth the $70, or do you need to be a mechanical engineer to understand it?

[kenndoggy]

IDK, ther'es a lot of math in it. I'm starting with Carol Smith's 'Tune To Win'. After I finish that, I plan to attack the Milliken book next. However, the Millekin book is also used as a classroom textbook, and is necessary reference material. It is THE chassis & suspension bible.

[illinipo]

Tune to Win is a good introduction, and I'm glad I read it, but it is rife with conceptual errors and broken laws of physics that only a good engineer would notice. Smith's free body diagrams are rarely balanced and he relies upon "fake" forces which are actually momentum (mass times velocity). Instead of showing the force which changes momentum, he will use the momentum itself. Additionally there are a number of places where he clearly does not understand the basic principles of torque (force at a distance) and angular momentum.

$70 is a great price for Milliken, just a few years ago it was about $180. Check if your local library can get it on interlibrary loan.

[wazzu70]

Have you actually done the math found in Milliken? I have, and for certain weight and tire combinations, it is absolutely practical. So practical, that I run on bumpy as hell parking lots without any bars at all. And I do have fun. Allow me to reiterate,

If you can get the roll gradient you want without a bar, without exceeding an acceptable ride frequency for your tire and the surface you run on, no need for a bar! That is directly out of Milliken, and just plain makes common sense.

Yes, I know RCVD well. I think the issue here is you are talking about what could possibly happen with a combination that would not be a good setup for a BMW Macpherson strut suspension. There is a huge difference between what you could do to get your equations to balance and what actually makes sense. Garbage in, garbage out as they say. I can set up say an FEA analysis to give any result I want to see, you still have to ask yourself if the answer actually makes any sense!

How fast are you driving in these bumpy parking lots? 50mph is significantly different than 120mph! Yes you could run without a bar if you can get the desired roll gradient. Most cars I work with run around 600lb/in spring rates up front with Rcomp tires. Say you get 200lb/in additional spring force in roll from the ARB for 800lb/in total spring rate. Yeah a 800lb/in set of springs will give you the same roll gradient, but your car will be much rougher on bumpy parts of the track at high speeds.

Another issue with high spring rates to control roll without the use of an ARB is your spring rates will likely be too high for the tires in a practical scenario. When your spring rates are higher than is necessary to drive heat into the tires you overwork them and turn them into a greasy mess. You also get less traction when the tires bounce off the pavement instead of following the surface.

If your only experience is autocross the negative effects arent as dramatic after a 50 second run. When running a 30 minute race or a 12hr endurance race it matters a lot!

The general rule is you dont want more spring rate than you need to heat the tires and keep them in the proper temp range.

I'm not sure I totally understood what you were getting at here. Why would you get more load transfer to the outer tire without an ARB? The situation is actually the other way around, an ARB increases load transfer. That is why when you stiffen the ARB on the front, the car understeers more, and when you add more rear bar, it oversteers more. It's causing more weight transfer relative to the other end and reducing the cornering power of that end of the car.

You are right, my example wasnt very clear. It was based on speing rates you would actually run on the front of a BMW. With the ARB you will have a higher spring rate than you would have otherwise. Since the outer front wheel will have a higher rate in cornering you will even the load out better among the other tires. The situation is the same with a high spring rate in the corner, although practically I dont think you should run that high of a rate. An ARB isnt a magic device, its just added spring rate in roll only.

The fact that clutch LSDs can still put power down with a wheel in the air does not make it okay for one to set up a suspension to lift a wheel.

True. Just because it still can put down power does not mean you should set up the car to do this. Sometimes its a necessary evil though as lifting in one corner might slow you down a bit, but the potential gain in the other corners may be faster overall. You are optimising for the entire track, not a single corner.

The ARB does the opposite of evening the load between the sides. It actually increases load transfer.

You only have so much load. If you have a high rate in a corner to resist roll, the load is not erased, it is distributed to the other wheels. This decreases the load on say the outer front corner and distributes it to the other wheels. Seems like the corner forces are now more even than they would have been otherwise. Correct? Just because it is evening the load does not mean its at equilibrium. Thats what I meant anyway.

...I suggest you re-read the sections of Milliken covering load transfer before your logic sets you and anyone reading your posts down the wrong path.

I think I am due for a re-read anyway so thats a great suggestion. My suggestion to you is to not read one small section and take that as a standalone piece of the puzzle. The maths might make sense when looking at a particular area only, but it might not make sense for the entire function of the car. Everything needs to work together to be siccessful. You can't just say every part works individually, so its got to work great when its part of the system.

In the beginning you said you had read a lot but did not have experience. Experience is key to sanity checking your results. 1000lb/hr front springs might look good on paper....but go drive the car like that and see if it still seems awesome


The type of racing and the speeds make a big difference too. Autocross, which it seems you are relating more to is a funny beast.

[illinipo]

First let me say thank you for not blowing up. It's nice to have a discussion on a forum in this age of people getting all defensive. Also I apologize if there really was some offensive tone there... not having the best day.

You only have so much load. If you have a high rate in a corner to resist roll, the load is not erased, it is distributed to the other wheels. This decreases the load on say the outer front corner and distributes it to the other wheels. Seems like the corner forces are now more even than they would have been otherwise. Correct? Just because it is evening the load does not mean its at equilibrium. Thats what I meant anyway.

The spring rate does not significantly affect load transfer. The extent of body roll is mostly insignificant as well. That's the point that I think needs to be reinforced here. The only things that affect load transfer are weight, CG, and track. A bone stock Mazda 3 with 100lb springs has approximately the same load transfer as a Mazda 3 on 700lb springs at stock ride height, at least until the 100lb sprung car lifts a wheel.

In the beginning you said you had read a lot but did not have experience. Experience is key to sanity checking your results. 1000lb/hr front springs might look good on paper....but go drive the car like that and see if it still seems awesome

The type of racing and the speeds make a big difference too. Autocross, which it seems you are relating more to is a funny beast.

1000 lb/in, lol. I don't recall saying anything about springs that stiff up front (ride frequency around 3.35! roll gradient around 0.5deg/g!). That said, 1000lb rear springs could work very well on a car with super soft Rs and aero. I remember reading around here about true coilover cars running around 400-450lb springs, which I calculated to equal about 900-1000lb inboard springs.

I would certainly try to run street tires on 600/800 at 120MPH, or maybe 600/750 to bring the balance a bit more to understeer. This setup plenty fun at 80 and doesn't feel uncontrollable at all. I would definitely try R comps on these springs at 120MPH. It's all a part of finding out what frequencies you can run on the surface you run with the tires you run at the speeds you run.

I did the math and it happened to come to fruition in reality. It's not fair to completely discount either the maths or the real world. Either is less effective without the other.

[wazzu70]

Unfortunately the forum ate my reply. It said it autosaved....just no idea where. Said it posted and it didn't!

[wazzu70]

Well, looks like I can't get my post back. Seems the auto save only works in the same session....and it wasnt working then. Too bad, Im sure it was a brilliant post Ill just quickly summarize then...

No worries on not getting angry, it does not help the discussion and no one learns from it. Im in this to learn over anything.


For the 600lb/in front spring rate you mention, that is approximately what most cars I know run. Although, they run this with an upgraded ARB. This setup is also quite effective and race winning. Without a front bar you would need to run much higher rates which is what I was getting at with the high rates in my post. Its not a number out of the air without basis.

That is not to say the car might do OK with a slightly lower spring rate, or the current rate without a bar. The drivers feel confident enough in their spring/sway setup they wont let me make much change to that. I will have to push to let me try it and get feedback just for experiments sake.