What Do Sway Bars (Anti-Roll Bars) Do?

[BSD] When you increase roll stiffness, which is what aftermarket sway bars do, you are reducing the amount of body roll for the same amount of cornering force. This does 3 significant things with respect to cornering. They are:

  1. Increases weight on outside tires - decreases weight on inside tires
  2. Reduces amount of body roll which reduces camber loss due to body roll
  3. Affects front to rear weigth distribution

Effect 1) is not a good thing for handling. The coefficient of friction for your tires decreases as more weight is applied. Read that twice to make sure you got it. If your car weighs 2000 lbs and 1000 is in front and 1000 in back (500 each corner), your cf might be .8 for illustration purposes. Your total grip in front is 500 * .8 + 500 * .8 which is 800. As you go into a turn, some of the weight in front transfers to the side so you may have 600 on one front corner and 400 on the other. However, at 600lbs, the cf may be only .7 and at 400, the cf may increase to .85. (Each tire has different values here.) Your total grip now is 600 * .7 + 400 * .85 which is 760. This is a decrease in total front grip. When you add stiffer sway bars, you increase the weight transfer to the outside which further decreases the cf for that outer tire and, therefore, for that end of the car.

Effect 2) is usually the primary reason for sway bars. When the body of the car rolls over, it takes the suspension with it. This causes the tires to ride on the outside edge of the tire and not get a flat contact patch. NSXs are very good in this area, but can still be improved (I suppose).

Effect 3) is neither a positive or negative absolutely. It depends on your car. If your car understeers, it may be due to effect 1) on your front suspension. Too much weight is transferred to the front outside tire which reduces its cf and may be causing it to break loose from grip. A way to prevent this is to increase roll resistance in the rear with a stiffer rear sway bar. The stiffer rear sway bar would resist body roll and apply more force to the outside rear tire instead of applying so much to the outside front tire. Obviously, you can go too far (too stiff in rear) which would transfer too much weight to the rear outside tire and cause it to decrease cf too much and eventually break loose.

Sway bars are a car set up thing. More isn't always better and often more is worse. Adjustable sway bars are a big win, too, because you can play with the front to rear roll resistance and find out what is better for your car.

In my measurements with a pyrometer and stock size R1s at the track, I have not found that the NSX needs more negative camber or less body roll (effect 2) very much. My tire temps were pretty flat accross and not "too high" on the outer edge. My suspension is still stock (not counting shocks which don't affect this.) One thing I did find was that the inside edge of the inside tire was hotter than it should be. Not sure if sway bars will help this condition.

One quick comment about driving. Do you see why driving smoothly is sooo important? It is partly due to the fact that if you twitch or jerk the car around, you will be transferring more weight than is absolutely necessary. This, as described above, decreases the cf for the outer tires which decreases the grip of your tires. Also, jerky driving increases body roll and, therefore, you lose more camber which further decreases grip. Driving at maximum performance is a balancing game of holding the weight of the car in the right places as you go around the track while not transferring too much weight.

[BSD] The concept of a sway bar (or anti-roll bar) is to resist body roll. When a the car body rolls, one side of the car gets closer to the suspension parts (they get compressed) and the other side gets farther from the suspension parts. An anti-roll bar is a U shaped link that connects the suspension pieces on both sides of a car. There is one bar in front and one bar in rear for most applications. We'll focus just on the rear.

A sway bar is a U that is shaped more like

|_________| (top view)

Its force is based on the diameter of the bar and the length of the arms. In an NSX, the long center part of the bar is held in place by bushings that let it rotate and slide, but not move out of its centerline plane. Then, each arm is connected to the lower A-arm, one on the left rear, one on the right rear.

When the car goes over a bump, both wheels in back go up and then down at the same time. This moves both arms the same which causes the bar to rotate in its collars. The sway bar has no effect on bumps or dips that affect both rear wheels the same.

If only one rear wheel goes over a bump, it lifts the arm on that side up. The centerline of the bar stays in the same plane and the left arm stays where it is which is connected to the left rear A-arm. The moving of the right arm of the sway bar up causes torque on the sway bar which resists the movement up. And, as the wheel goes past the bump, the sway bar helps push the right A-arm back down.

The effect is a spring that resists differences in positions between left and right rear A-arms.

When the car goes around a turn, the body tends to roll over towards the outside. The outside suspension compresses and the left side extends. Relative to the sway bar which is held in its center line, the right arm bends up and the left arm bends down. This twists the sway bar and the sway bar resists this twisting and, therefore, resists the body roll.

There are many factors in calculating the forces at work here which is way too complicated for an email. I have books that explain it if anyone wants references. The point is that, all other things being equal, a stiffer sway bar will resist the car's body roll more than a weaker/thinner one.

Does Bruce have a problem? Well, in Bruce's case, the link that holds the A-arm to the arm of the sway bar is straight on one side and at an angle on the other. For the side that is at an angle, the movement of the A-arm will not be exactly the same as the movement of the arm of the sway bar. The sway bar link can change angles and move the sway bar's arm less than the amount the A-arm is moving. I think this will result in less spring loading do to less movement of the sway bar arms for the same amount of body roll. The overall effect is that the sway bar's effectiveness is going to be (slightly) reduced.

Although having angled links reduces the sway bars effectiveness, it still remains equal on both sides. In all cases, the left is pushing against the right and vice-versa. There is no way, short of welding the sway bar to the car, to have one side be more or less effective than the other.


OEM Anti-Roll Bar Data

diameter x wall thickness (all units are mm)


Front Rear
1991 NSX 18.3 x ? 17.5 x ?
1993 NSX-R 21.0 x 2.6 17.5 x 2.3
1995 NSX-R 21.0 x 2.6 19.1 x ?
1997 NSX 18.3 x ? 17.5 x ?
1997 NSX-S and Zanardi  18.3 x 3.0 19.1 x 3.0
1997 NSX S-Zero 18.3 x 3.0 19.1 x 3.0
2002 NSX-R 22.2 x 4.0 17.5 x 2.3

Note: The Type-R anti-roll bars are bent a little differently than others. Different years are also bent differently to accommodate other changes in the car, so they are not all interchangeable between model years.


What Sway Bars Are Available?


[CN - 99/4/28] Comptech NSX tubular adjustable sway bars are in stock. Gloss black powdercoat. Front bar has 3 points of adjustment each adjustment is +/- 20 lbs. Rear bar has 2 points of adjustment each adjustment is +/- 20 lbs

Fronts are 123% stiffer, rears are 220% stiffer. They should eliminate nearly all of your body roll without affecting ride quality.

Each kit includes front and rear bars, and poly bushes and brackets along with teflon grease. Reuses stock end links for reliability. Retail is $350. These bars fit all years of the NSX.

[AWL - 99/10/19] Keep in mind here, that stiffer is not always better, especially at the extremes.I went from Dali street bars to the Comptech sway bars (125% stiffer in front and 200% stiffer in rear), and there is a noticeable difference in stiffness. The car does not roll at all...but it is also unstreetable, in that the car is now very skittish over even small bumps and ridges, which causes the car to hop around when going through turns. My car feels like it has 500lb springs in it now. While the Dali and RM sway bars don't adversely affect straightline driving comfort, the Comptech sway bars are so stiff that they do...as the suspension is not less independent at each corner; the damping force on one side of the car is now working not just on that one corner, but also on the other side of the car, and thus the sensation (and reality) of increased overall damping.


Dali Racing

Series 1

Dali Racing's original Series 1  sway bars were RM's bars re-labeled and painted Racer Red. Easily installed using common tools. Available in 19mm and 21mm diameters. See the RM Racing section above for impressions of the bars.

dalistb.jpg (20357 bytes)


Series 1 Adjustable Sway Bars

[AV] It looks really trick, and I had a chance to drive his car for a few hot laps around Holtville. Though I didn't get to push the car 10/10ths (probably more like 8/10ths) I did not feel any tendency towards oversteer. This is with the stiffer Dali/RM front swaybar.  The adjustment on the rear swaybar is 1 hole towards understeer to make it "neutral." Mark says this means that he might have to make it smaller to get it dialed perfectly with the 19mm front bar, or make the front bar bigger, he has not decided.

The swaybar is a straight bar from left to right, with bolted on lever arms on either side. This is in contrast with the factory and the factory-like Dali/RM rear sway bar that is bent at several places to clear the cats and the other structures underneath the car.

[MJ] It will fit the '95-'96 models no sweat, or the '91-'94 without cats or perhaps with some minor exhaust bending. (I have not tried to do that)

A post I re-read today talked about one set of bars that caused the car to have a slight tedency to oversteer. If either the front or rear sway bars was adjustable, you could then affect the understeer/oversteer characteristic at least to some degree. Most race car drivers feel (from what I've read) that a little oversteer is faster than some understeer, but lots of oversteer is worse than understeer. I don't think I am ready to have a tail happy car... I'd rather have some degree of understeer which I'm used to. Further, if you get mentally lazy late in the afternoon and understeer starts to get you, it is fairly easy to deal with it. If oversteer sneeks up on you, it may just grab you and spit you out heading who knows what direction.

Us go-karters like a little oversteer. What I don't like is thesnap-back oversteer the NSX and other rear heavy cars tend to have. Anyway...

My experience is that I didn't feel the NSX to have anymore oversteer with the Dali/RM swaybars, as compared to the stock sway bars. At the track, I try to compensate for the tendency for this oversteer by lowering the tire pressures and using softer suspension settings at the rear. I run softer rear springs, and may sometimes try to adjust the Konis perhaps a turn softer than the fronts. But a lot of times adjusting the shocks aren't even necessary.

I agree with Brian that having a little understeer is a lot safer thanhaving oversteer on the street, as a less than attentive driver would be more likely able to effectively correct this. But the fastest way around a track is to have NEUTRAL steer, neither over- or understeer. And a track driver should be FOCUSED. If he feels he has less than optimal concentration, he should get out of the track and into the pits.


Series 2

[MJ] What's wrong with the Series 1? Harry was giving me shit about them being "too heavy", so I made them hollow, and some people said "not stiff enough" so I made them stiffer (well now there are two versions) and Al.T said, "I think they move around" so I welded the "collars" on.

Series 1
Solid bar stock, J value: 0.0311
front: 7.5 lbs
rear: '91-'94 = 8.0lbs, '95-'99 = 7.5 lbs
non adjustable front and rear.

Series 2 Street
Hollow bar stock, J value: 0.0301
front: 4.0 lbs
rear: '91-'94 = 4.0lbs, '95-'99 = 3.8lbs
4 hole "stiffness" adjustment in the rear only. welded 'collars' for locating bar in center of axle.

Series 2 Race
Hollow bar stock, J value: 0.0477
front: 4.0+ lbs
rear: '91-'94 = 4.0+lbs, '95-'99 = 3.9+lbs
4 hole "stiffness" adjustment in the rear only. welded 'collars' for locating bar in center of axle.

Pricing is the same for both size Series 2 bars, $325/set with shipping included.

OEM J values for comparison -- Front: 0.0238,  Rear: 0.0200

Remember that they are the same size (stiffness) front and rear because the goal is twofold: 1) to reduce body roll (looks are everything - right<G>) and 2) to enhance cornering and handling by taking out the built in understeer and moving the car's balance towards "neutral". I realize everyone may have their own definition of "neutral steering", but this is mine ON OEM YOKO TIRES (either 15/16 or16/17).

I emphasize this as by changing to tires that have the same compound front and rear (and therefore the same "stickiness" factor front and rear) like SO2's, SP8000's etc. you have already changed the handling balance of the car. If the new rears are wide enough (ala Beermans's switch to 275/35/17 rears) then IMO the stickiness factor is closer to stock than if you get the same size (245/40/17) as the OEM. YMMV etc.

I am running my Series 2 (21mm diameter) "race" bars, with Eibach progressive rate springs and Koni Sport Shocks. The car is now flat on the turns and even more nimble than before. Adding the Koni's firmed up the ride a bit too much for me. You may want to try the Bilstein or leave on the OEM if you are not SpeedRacer 100% of the time.

[BSD - 99/5/20] Recently I have been evaluating two sets of sway bars, the Dali Racing "street" bars and the Dali Racing "racing" bars. The street bars are less stiff than the racing bars.

The appearance of these items looks first rate. They are painted with a nice and shiny coat of red paint which looks really good under my red car. The front bar of each set has bends in it to mirror the stock bar. They have a single hole in each end for attaching to the drop links.

The rear bar in each set is also bent like the stock rear bar but each end has 4 holes in it so you can select one of 4 positions of rear stiffness. The 2nd stiffest hole is marked to let you know what the recommended position is.

I think the Dali bars are significantly lighter than stock. I never weighed them, but when I picked up all 3 sets at once, I could tell that the stock bar is heavier than the Dali bars.

Installing these bars is, for the most part, easy. I have done a write up on installation. Basically, it takes a jack and jack stands, plus a few basic hand tools, such as a 12mm and 10mm socket, a 17mm wrench and you'll probably need a set of metric Allen wrenches (6mm size I think it is.)

Basically, a sway bar on an NSX is held on with 4 12mm bolts and 2 17mm nuts, one on each end. That is it. However, in the front, you need to remove the spare tire, battery and battery holding brackets to get the sway bar in and out.

Adjusting the rear sway bar only requires the removal of the 2 17mm nuts, then selecting a different hole.

I started out by installing the street bars on my car. Immediately, you can tell that the car is less prone to roll. You feel it at turn in, because the car turns in slightly quicker. You feel it in the turn as well, because for any amount of turn rate, the car leans over less. On the street, the car just feels more sporty due to the less body roll. My van has body roll and anything less like a van is sporty feeling.

As an even better test, I took the car to Texas World Speedway for a PCA driving school. While lapping instructors in the instructor run group, it really made the car feel more controllable. Cars that oversteer in some situations are sometimes hard to control as you approach oversteer. The street bars made it easier to detect when the rear tire would let loose and made the driver have more confidence. I tend to stay away from oversteer because it is not safe and I'm not out here to win big money, but I found myself not being so concerned with oversteer. I knew I'd be able to drive on out of it.

As with any improvement, there can be a negative as well. For sway bars and the like, the normal down side is that if you stiffen up the rear end too much, you get power on oversteer. This type of oversteer does not cause accidents usually, but does keep the driver from getting on the gas as early in turns. That results in slow lap times. Crew chiefs have to decide on either more turn speed (stiffer) or earlier on the throttle (less stiff) and make that decision for each track or day. I did not notice much of a change in throttle on oversteer. Note that the rear sway bar has 4 settings so I can change this characteristic of the car.

As for other possible negative features of sway bars, a bumpy ride is not one of them. Sway bars connect the wheels on an end together (fronts to fronts or rears to rears). When you car hits bumps, say going up a curb, the sway bars have no effect.

Sometime later, I put on the Dali Racing "racing" sway bars. These bars are slightly thicker visibly than the street bars, but as you may know, small changes in diameter have a large change in torsional stiffness. With these guys on, the car is again noticeably stiffer. If you liked a little increase in the resistance to body roll, these give another improvement over the street ones. The improvement here would be obvious to any driver.

Also, the fun is improved as well. It is really fun to have your car handle like this. It makes the car feel lowered and like it is just above the ground. In turns, it really doesn't lean too much and you as a driver are aware of it. When you are not turning, however, you can't tell at all that the sway bars are on the car.

As a first test of these bars, I mounted some Yoko A022H tires (15/16) and went to a local track, the site of NSXPO '97, Second Creek Raceway. It was a test-n-tune day so I had all the track time I could handle. The racing bars instills even more confidence in the driver. The car reacts when you want it to react and doesn't waste much time shifting weight, rolling over, etc. As you know, rolling over changes the geometry of the suspension and the driver is always compensating for this. There is less driver re-adjusting after turn in with the stiffer bars.

Further, the combination of street tires and the stiffer bars made the car more go-kart like. What I mean is that the street tires reduced the grip that I am used to at the track which made the car more likely (easier :-) to slide into and out of turns. Like I said, I am not big into oversteer, but on this day, with this setup, hanging the arse end out a few degrees while powering out of the turns was certainly fun and easy to control. I think that stock bars (less stiff cars and also ones with bad shocks) are almost uncontrollable in this fashion.

As I checked the tires for wear, I saw what appears to be good things. That is, the tread blocks seem to have the same rubber scratching/sliding across the surface of the tire, as opposed to only being worn like this on the outside of the tread. It appears that a stiffer car wears the tire more evenly than stock at the track.

The track I was on is rougher than any street and we were going significantly faster. :-) Even over bumps with stiff shocks and stiff bars, the ride was great. Very sporty feeling.

I took some tire temps and noticed only a slight improvement. The outside edges of the tires were less hot relatively, then they should have been with stock bars. However, the inside edges still got hotter than you'd expect and resisting body roll did not fix this to an obvious degree with this brief set of measurements.

I should note, that while at this test-n-tune day, several AMA Superbike drivers were there. I was on the track with them and if only I had a video to show the NSX catching and passing road racing motorcycles. Basically, the NSX can brake with them and out turn them, but lose in acceleration. In fact, we lose bad in acceleration, as a bike not more than 15 feet in front of me pulled a wheelie out of turn 1 heading down the back straight.

One of the bikers was a young female. A group of the bikes, including hers, was parked near our cars. As we were leaving, my buddy's car did not have enough fuel in it to start. I asked one of the guy bikers if we could buy a gallon or so of gas from his container. Well, the fuel was owned by the girl. She said she would not sell any to us, but if she could have a ride in my car, she'd give my buddy the gas. Anyway, out on the track we went, street tires and all. All should could say was "Hell yes!" and "Rad!" and a few other slang biker terms for kicking butt on the track that I have since forgotten. We, on streets and with a passenger, were able to pass one of the 3 superbikers out there, but the two other guys could just out accelerate us too much. Anyway, my passenger was thrilled with the ride and was really impressed to see it hang with the bikes.

(Note: later in the day, the superbikes were able to run faster lap times. They had to get their tires up to full temp first and had to get the confidence to really take the turns hard. It *was* a rough track.)

I also had the racing sway bars on at NSXPO '99 at Leguna Seca. By now, I am used to the bars and the car feels really comfortable and natural. Well designed sway bars (and other mods) do that to you as a driver - they fit in so well that you get the benefit, but they become second nature.

On turns 3, 4, 5 and 6 at LS, I think that stiffer sway bars give an improvement over how stock would feel. The car does not roll as much and so when the car comes to the track out point, it is less of a transition from full turn to going straight. You know how race cars on race tracks follow the outside edge of the track right from the track out point... they don't over turn, then steer back out? Well, I found this easier to do with stiffer sway bars. I remember watching videos of me previously, in other cars also, not being able to just stay at the edge of the track. I think that stiffer sway bars help the driver to stay straight after turning.

One final thing. I had a new set of G-froce R1s at TWS and then used them again at LS. Both my wife and I were driving for both of these events. These tires *still* look like they can handle another event and the wear looks identical across the tread. Normally, a set of tires would be dead after 4 track events (8 track days). I have to assume that the sway bars are evening out the wear to some extent and that I am getting somewhat better tire wear.

If anyone does get 8 or more track days with a set of G-force R1s, then I'd be interested to hear it. The previous R1s would not last this long.

Anyway, I'm glad to have the Dali Racing "racing" sway bars. I hope to get back to Heartland Park and Hallett because I know my lap times there and can see how much of a difference bars make in lap times. Also, I may try putting the bars in their extreme settings (full stiff or full soft) to see how that feels at the track. I also may try a slightly less aggressive alignment over all. With sway bars, you may not need all that negative camber.

[DLJ - 99/6/21] After my first full country drive with the DALI *street* swaybars, I can give some amateur impressions.
Imagine driving straight, and then turning into a constant-radius curve. For a fraction of a second after the turn in, the car is settling over before it becomes steady in the turn. How do I know this? Because the time it takes to settle down is *different* from the stock swaybars! I was surprised to notice this, and hadn't expected to. Hence, the car feels more responsive to steering input.

Similarly, in a fast S bend, the transition from one curve to the next doesn't have the slight pause that occurs with the stock bars. To me, this means that the car feels more like a sports car, and less like a sedan-- though even stock it is
admirably flat on corners. It is *even* flatter now, most noticeable by the change in responsiveness.

There is a small increase is ride stiffness, but I'm willing to trade it for the responsiveness. FYI, I am using Koni shocks on the minimal settings.

[CA - 99/2/24] The DALI sway bars are great, they really flatten the unwanted body roll, and the rears are adjustable so you can "dial in" the desired stiffness, I have mine set at one setting away from max.


RM Racing

rmswaybars.jpg (16513 bytes)

[RM] 3/4'' 1045 c.f. steel, 45% stiffer than stock. Kit includes Polyurethane Bushings and features bright silver powdercoat finish. $355 a set.

[MJ] I had a set of RM sway bars in my car as part of a complete suspension package with Eibach Progressive Rate Springs and Koni Sport shocks.

They DO NOT affect the straight line ride at all, however, they do have an affect on the cornering ride as part of the NSX's independent suspension is now more "dependant" - i.e. one wheels motion affects the others more now that they are connected by a stiffer bar. I feel that the improved corner "feel" (much lower body lean than stock) is well worth it, AND that this is a much better method for reducing body roll than the 1000 lb springs (vs the 170 lb Honda factory OEM springs) that COMPTECH uses in the "Pro Suspension" to get the same effect. (I bought one and tried it as well, my kidney are still in shock) They have since lowered the spring rate they are shipping with the kit to 600 lbs in the front, still a huge OVERKILL in my book.

[DAD] Just wanted to post my findings after having Randy's sway bars installed. I was concerned that they would degrade the quality of the ride, especially since my NSX is used for more street than track use here, in pothole New

Well, I can tell you all that there is NO degradation in the quality of daily driving , and there is once again a big <g> on cornering and h'way exiting /entering . I might even say that straight line performance feels "enhanced" and reminds me of that "Tiger Paw" commercial, highway grip, straight line seems improved as well as cornering .

Would recommend them for ANY NSX owner!  Thanks Randy!!

[DNG] A little heavier than stock cause it's solid now. At low speeds, I really couldn't tell the difference. I would whip it left and right, and could notice a   little body roll. But take it at high speeds and wow, I can definitely throw it in the corners faster.

[CMI - 99/7/25] I just ran two days at Road America with RM' s bars installed for the first time. My 95T has Eibach Springs, stock 15/16 wheels with R1's. I believe this setup maintains the original handling characteristics of the NSX. Although the car still leans on the high speed turns, it still maintains the neutral handling of the of the stock bars. The new bars seemed to help track control, which allows you to go faster, then loose control. :)

[AT] I finally have had a chance to fully test the difference between the stock sway bars and the stiffer ones available from both Dali and R&M Racing. In a simple phrase, they are worth the change. My observations are presented in
the order that they have occured.

Scenario #1: Installing the sway bars is a very simple process for even the mechanically challenged. I drove the car to the nearest freeway on and off-ramps near my house. I immediately noticed that the car seem more stable during long sweeping turns. It felt smoother as if all the tires were working more equally during the cornering process and not just the outside taking on the brunt of the work. I have this one freeway overpass clover leaf near my house and I can just take one off-ramp and then get onto another on-ramp and back onto an off-ramp etc., all day long if I want. All four of these corners are different in their radius and how much elevation climb or drop you might have while on them. There is one particular off-ramp which is an uphill ever tightening right hand decreasing radius sweeper. I like this one because it forces the car into the turn as you climb giving you extremely good traction and if you have the nerve, you can take with a lot of speed. With my old sway bar setup, I would be taking this car in second gear and as I'm climbing and accelerating, I'm also having to correct for minor car attitude changes during the turn. With the new sway bar setup, the turn feels more stable. It's as if the inside tires are doing more of the work once done by the outside tires during the turn. I also seem to sense the car being more nuetral, less understeer while going into the turn than before. Now this wasn't a large change but something that definitely felt better.

Scenario #2: While driving from my wife's job site at the UC-Davis campus, there is this nicely paved left hand sweeper that I wanted to try and see how the car would react to some power during the cornering. I accelerated from first and quickly shifted into second during the middle of the sweeper and because of the slight center crown in the road, the car started to spin out to the right. I started to correct for it but it then snapped back to the left and for some dumb reason I let the steering wheel spin in my hands to help counter the spin. No flames please, it's an old bad habit that I'm constantly working on. What I did notice is that the car quickly corrected itself without any additional input from me with ever decreasing swings from right to left very quickly. With the old sway bars, this correction never occured. It the car snapped back from a swing in one direction, it would almost always spin out in the other direction no matter what I did. With the new bars it seem to lessen those oscillations which made them more controllable. An autocross was coming soon and there I would try to induce this action and controll it in a safer envrironment.

Scenario #3:
An autocross. I finally had the chance to try the bars out in a scenario which would really test the differences between the stock setup and the stiffer bars. I was using street tires for this run and I had 4 chances with two laps each run to push the car to the limit. The course was a fast layout. Designed for the Corvette club they needed the extra room and longer straights. The first run reinforced one feeling I'd experienced from the on/off ramps. The car seemed to understeer less going into the turns then in the past. However I also noticed that in hard sharp cornering it also wanted to oversteer more quickly. This could be bad but it was also more easily controlled than with the stock bars. I could sense the car oversteering sooner and could react quicker to control it. Second run, knowing what I did, I started to experiment with going into the corners a little faster, inducing some trailing throttle oversteer and then controlling the cars steering with both the steering wheel and the throttle. It worked, rough at first but controllable and my time improved. Run 3 I got a little carried away and got out of shape for one turn, but no spin out, which makes the next two or three turns more difficult. Caught a few cones and didn't feel that good about the results. Run 4 was a lot more fun. I got the speed up and looked further down the course. I was able to induce some trailing throttle oversteer in a number of corners and controlled them either with the throttle, the steering wheel or both. The time again improved and no cones were hit.

All in all the car now feels like it using all four wheels during hard cornering rather than working just the two outer tires of a turn. It feels flatter, smoother in a corner and this gives you a better feeling of the car. I also found out that using the stock tire pressures of 33 front and 40 in the rear for the Michelin tires was just too high. I am now using 30 in the front and 34 in the rear and may go even lower by a lb. or two. This is mod-upgrade I can recommend to just about everyone on the list. JMO-ICBW

[BC] I installed the RM Racing sway bars on my NSX this weekend. Cornering is so flat and solid now it's scary. You wail around the highway cloverleafs with a knot of fear in your stomach thinking "I'm going way too fast to make this turn" but the car doesn't care. The super tight and in control feeling as you exit a turn at full throttle is a step above what just the Koni shocks and Eibach springs deliver. I'm amazed it's taken so many years for the aftermarket guys (like Randy) to make these stiffer bars for the NSX.

Some have reported extra oversteer with these bars. The only downside I noticed is that hard braking with a snap 90 degree turn (like into a side street) can make the tail twitch, which might spin you if the pavement was wet - but then this isn't the sort of maneuver you would do with a race car, eh?

When installing the RM bars I noticed that the linking ball and socket joints were so stiff that they barely moved. The 8 year old grease was solidified until I worked it around a bit. The sway bar had a smooth action when I put everything back together. I wonder how supple these links must be to really work right? I'd bet every link in the suspension has the same arthritic condition on my old 91.

DIY Installation note: if the bolts are rusted, get a hex key ratchet attachment (or curl up into a fetal ball and die under your car).

[TMI - 99/3/7] I moved from 91' to a 95-T and was disappointed with the lean and how much looser the 95  was. I put on RM sway bars this weekend and as stated in other posts, it's amazing. The car handles flat with minimal change in the ride. For some reason that is not clear to me, the car is less creaky/flexy than before. This always bothered me compared to the '91. As has been said earlier, the sway bars are a "dead bang go do it" proposition. No compromise that I can tell, I wonder why they don't come this way?


FAQ Admin Note: The following comments are about the Series 1 Dali Racing sway bars, which were simply RM's bars re-labeled and painted Racer Red.

[WM] I installed the Dali Racing sway bars this afternoon. I did the installation by myself in the driveway with no lift, just a normal floor jack and the NSX jack. It took about 1.5 hours to do the front because I couldn't figure out how to get the old bar out without removing the spare tire, battery and the aluminum tray underneath. The rear was a piece of cake and took about 30 minutes from start (jacking up the back) to finish (putting all tools away).

Next I went for a ride through my favorite twisty section of road. There is a big difference. The car didn't lean much, if at all but I noticed that the handling felt a little twitchy and the car seemed to understeer a bit. At the end of the fun section of road, I pulled over and adjusted my Koni's. They were all set at 1 turn off full soft. I made the rears 2 turns off full soft and left the fronts at 1 turn off full soft. I then went back through the road again and it was better but I was still not that happy. I then adjusted the fronts to a half a turn off full soft and tried again. This was a little better still and I left them there. What does anyone out there set their springs to for normal driving? I actually prefer a little oversteer rather than understeer and think that you are supposed to have the rears harder than the fronts for this. Am I correct? Anyway, the sway bars are well worth it.

[NM] Dali Racing Sway Bars (approx. $300): One of the best $300 investments you could make. Really keeps the car leveled in the turns. Someone was telling me that they worked an NSX event last year and they could tell just by looking which cars had the sway bars and which ones didn't. After the brake pads - the best bang for the buck I have found. Mark tells me he sells these bars by taking people for a ride around a cloverleaf with the sway bars in. They start writing the check before the turn is finished. NOTE: Dali and RM bars are not yet available for the NSX-T so you may have to wait a few weeks as they are in development.

[WSC] New Dali Racing sway bars worked great

Sway Bar Collars

[AT] Well I've finally had a chance to modify the R&M Racing/Dali sway bars setup to see if it can be improved. It can and I think I've done it. The biggest thing I noticed was the ability to induce throttle oversteer at will and then control it to almost eliminate the dreaded snap-oversteer the stock setup has. After you start the oversteer siuation (or accidently induce it) you start to counter the oversteer by steering into it. With the new bars and my modification the car catches the start of the spin (oversteer) more quickly than in the past to stop the oversteering condition. You then counter in the other direction in anticipation of the car snapping back, usually wildly, in the other direction. Except this time it only snaps back very slightly and then it catches, again very quickly, as you start to accelerate out of the manuver and continue on your line. What a great feeling.

So what is this great modification to the sway bars. Something quite simple and easy to install on any sway bar setup. They're known as shaft collars and are available in steel, stainless steel, and aluminum. These collars prevent the bar from shifting side-to-side as they are also twisting during cornering. In this fashion the bar becomes even more efficient and reduces the amount of body lean the car will produce in a corner for the same given force. Again these collars can work for the stock sway bars or the newer bars from R&M and Dali. The cost for the collars are listed below:

Steel (4 ea.) $25 + $5 shipping
Aluminum $25 + $5 shipping
Stainless $50 + $5 shipping

I've gone with the stainless because of the potential corrosive environment they might face but I'm sure Aluminum will work fine for those who are worried about the weight difference. Aluminum is around 1/3 the weight of stainless or the steel units. The steel units have a minor anti-rust coating but I don't think they would stay rust free where the sway bars are located in the long term.

It's still too new to see if there will be any damage to any suspension parts for the long term. I seriously doubt that there will be any since it doesn't impinge upon any suspension part and the collars are even designed not to mar the sway bar shaft that they fit on. If you're interested in please send checks, inquiries, what type of material, and whether the bar's are stock, or R&M/Dali's bar. I will fill all orders within 2 days and answer any inquires via e-mail.

[AT} Well gang I know I've been touting the benefits of the Dali/RM sway bars especially with the sway bar collars. I can now truthfully say they've been tested under racing conditions with extremely good results. I don't care about all those spring rate numbers, although the information was very good and I'm glad we now have it. Nor am I going to get into measuring one car versus another on a skid pad which will again give us a more quantitative approach to this IMPROVEMENT. Yes I did say improvement. But compared to what I've worked with in the past with the stock sway bars on the street and the track, and now the Dali/RM sway bars with my collars again on the street and track, the latter is definitely the way to go.

We had an autocross course setup in Sacramento on an old B-52 stand-by parking paddock area which is made up of concrete and some asphalt. The course was made up of mostly short sweeping turns, a couple of long hairpins, a couple of short straights and a five cone slalom layout. Now I'm running on stock 94 16/17" wheels using Michelin MXX3 tires. So I'm not using race tires to make up for handling problems or possible poor driving techniques.

From the moment I started each run I noticed the car seemed more neutral as I entered into any corner and my speed was up while doing this. In fact so much faster than in the past that where I normally would have to be in first gear to get through a turn effectively, I could now be in second, carrying more speed, and use the car's more neutral stance to allow for earlier power application as I was exiting the corners. When I did apply power I found that I had less wheel spin than before and that when the car started to oversteer because of the power, I could quickly bring it back in with the steering wheel inputs. I mean the steering inputs were much smaller than in the past to correct for this oversteer condition. Now when the rear starts to come out, I add some steering input to correct, then more power and the car hooks up, squats down in the rear and goes where I want it to go. What a GREAT FEELING!!!! It's almost like I knew what I was doing.

Normally in the past where I found I got in trouble was in the slaloms. These are just a series of cones set in a straight line so many feet apart. In this case around 30 paces. Now going into the first one is rarely a problem but you should go through these cones with a rhythm or flow rather than just trying to accelerate past each cone. It's very easy to get the car's rear end to start to swing wider and wider afer each cone and by the time you reach the last cone, you end up either hitting it or possibly spin. This time the car just hung on and I could accelerate toward the next cone gaining speed a little at a time before I reached a long U-shaped hair pin turn after the last cone. It felt very stable and very neutral through each of the cones. It felt like all four tires were working to make the turn rather than just the fronts. FUN! FUN! FUN!

Oh, for those who think the NSX can't shift into first gear while moving, well I can attest that it can and does fairly easily. After the slalom cones I was in second gear going through them at around 40-50 mph, when the left hand hair pin was coming up which meant I would have to get into first gear to get the optimal handling and speed through this turn. After the last cone I applied the brakes while still going straight (never invoked ABS ever in any corner) and could feel the tires scrubbing off some speed, and while I was going around 20-25 mph, I shifted towards first gear. Now since I kept my grip on the shifter very light, I let the car do the work of getting it into gear for me. How did this happen, simple. As anyone who has raced and knows about heel and toeing, when you match the enginge revs. to the proper gear you've selected, the shift should occur very smoothly. As soon as I reached the right engine RPM, the shift just went into gear almost on it's own. It was just that smooth. You just have to keep a light grip and don't rush it. It will occur.

From that point on I used the steering wheel and the throttle to help steer me around that corner. I was able to apply power early and when the rear end wanted to come around a little, I quickly counter with the steering and the car responded. It would never do that with the old sway bars.

I could now attack the course where before I would often hesitate on some corners because of how the car handled. Also the course was bumpy in the transitions from concrete to asphalt often in mid corner. No problem with the new bars, it handled it with ease.

Gentlemen and you ladies who like to motivate the NSX like it needs to be, this Dali/RM sway bar setup is a no brainer. With my collars added to that, it's like icing on the cake. The car is so much more fun to drive it's just has to be experienced. And yes when the car is handling better, you also start to drive it better. Enough said. Order the bars and my collars.

[AT] According to Chad at Comptech he says the links are fairly hefty and the added stiffness of the sway bars we will be using shouldn't present a problem. For the bar to work perfectly, it should just pivot up and down on the sway bar bushings as the suspension is moving. While going straight down the road the bar will just rotate on both sides equally (almost) on the bushings which will act like bearings.

When you make a turn into a corner, one arm of the bar goes up while the other goes down. It's this twist and the resistance to the twisting action that the bar imparts upon the suspension links to reduce the tendancy for the car to lean in a corner. Now if part of that twisting engergy is lost due to shifting from side to side then it won't be as effective or work as quickly during a cornering action. The collars role is to keep the bar from shifting and thus allow the bar's resistance to twisting to work more quickly and effectively in reducing body roll during cornering.

All race cars which have any form of decent sway bars will have some way of keeping the bar from moving from side to side. This also allows the use of smaller bars for the same control of a larger bar which may be allowed to move in any direction it wants to. The use of urathane bushings also reduces the tendancy for the bar to twist, as compared to the stock rubber units. This is the reason why I decided to use the sway bar collars on the bars I have on my car and make them available to members of the NSX list. So far it seems to add that little extra degree of control during cornering. Hope this helps.

[NM] I got the collars from Al Terpak [email protected]. He sells them and you can install them on the aftermarket or stock sway bars. If you look at the sway bars, they attach to the suspension just behind the rotors. The bar tapers down and bolts on directly with lots of "play" for twisting in turns - the collars are round pieces that keep the sway bars in place - and ultimately much more effective. I did it myself and the whole job took about an hour to complete.

They come in steel, Aluminum, and I think stainless. I got the Aluminum ones for $30. They work very well and really enhance the ability of the sway bars to stay "lined up" and thus effective. I have had them for a couple of days and really like them with my setup of H&R springs and Dali sway bars. I believe the car is very resistant to "rolling" now and is the most I would do for a car I was going to use on the street. I have been in cars with the Konis too and everything but "full soft" is too hard for me and the streets of St. Louis (ranked near last in the nation).

Anyway - Al is a great guy and from what I hear - a very good driver. This may be the best mod for the dollar available and I owe him some thanks for making it available.

[MJ] The addition of the collars DO NOT stiffen up the bars. The keep them from moving back and forth (as in side to side) on their mounts and changing their point of leverage so that their ant-roll action is consistent from corner to corner. That is ALL they do. (which is certainly worth having them for)

[BSD] If the sway bar mount bushings let the sway bar slide sideways, why don't they just fix the sway bar to the car so it can't move at all? The reason is because if you fix the sway bar to the car so it can't move, such as by welding it, the sway bar now resists movement between the chasis and the suspension, not the left suspension versus the right suspension. The fixed bar operates just like a spring in this case and our cars already come with coil-over springs which are the best you can get. (Well, at least the configuration that "real race cars" use.) Sway bars have to be allowed to rotate to be effective.


What Shock Tower Braces Are Available?


ctechstbjpg.jpg (8396 bytes)

Billet aluminum, non-flex ladder-type design.

Dali Racing

[NM] Dali Strut Bar (approx. $175): Extremely cool looking and worth the money. For the non-financially challenged, upgrade to the titanium bar.

DC Sports

RM Racing

rmstb.gif (80873 bytes)

Alloy shock brace.  Available polished or brushed finish.  $275.



Back Home Up Next 

Web NSXPrime.com
Copyright 1997-2008 NSX Prime, all rights reserved. Read the Disclaimer! Page last modified: December 16, 2002