reducing body roll around corners

[ekfour97]

hi all, so ive read all 90 pages and still couldnt really find the answer im looking for....whats the best way to stop body roll in a stock ek4 around corners at high speed. i want to stay flatter around the corner , is it harder springs , thicker sway bars or more bars perhaps....cheers

[mrwillz]

combination of coilovers or shock/spring combo
sway bars, strut bars

[JetSir]

as above ^^^^^^
whack on all the bars u can get =D

[chargeR]

Do not buy strut bars. They will not reduce the amount of body roll your car has. They will however probably affect your perception of the roll of your car and your perception of the handling response.

One way to reduce body roll is to increase your roll stiffness, which is easily done with either stiffer springs or larger diameter sway bars. Roll stiffness can also be altered with suspension geometry changes but this is quite difficult and out of the realm of what is easily achievable with bolt on parts.

Another way to reduce body roll is to lower your centre of gravity by lightening or removing parts that are high up in the chassis. This will reduce the roll moment car's mass applies to the suspension and hence reduce roll angle.

[ekfour97]

thanks , so when i have my car lowered choose harder springs and harder shocks....maybe coilovers , but are illegal in s.a unless engineered at a cost of $600 for certificate . will have to compare prices b4 i make the decicision , coilovers would be gd especially with dampner adjustment so not to hard for street but just right for track and motorkhana which i want 2 do next year....cheers

[JohnL]

Not many of the commonly available 'coilovers' have particularly good dampers incorporated into them. Some are at least OK, but some are crap, and the ones with adjustable bump and / or rebound valving rates typically have very little actual difference between full soft and full hard (I've seen a number of shock dyno readouts). This is often despite often having a huge number of adjustable valve settings (i.e. 'clicks', with very little stiffness change from one click to the next...). There will be exceptions, but don't expect them to be cheap or readily available through other than specialist suspension companies.

The springs are what hold the car off the ground, but the dampers ('shock absorbers') are more important than the springs in their influence on handling etc. A very good choice would be a set Koni dampers (or Bilstein) fitted to stock or stiffer springs (maybe stock springs now, stiffer springs later, as finance allows).

The Konis are in effect 'just' replacement dampers (though superior ones), so should allow you to avoid any legal issues associated with aftermarket 'coilovers'. I use the term "coilover" advisedly, keep in mind that any spring / damper unit that can be removed from the car without disassembling the spring / damper from each other, and, where the damper does not also act as suspension location device (as is the case with Mac Struts, where the damper rod locates the hub etc), is technically a coilover, not just the aftermarket products that are advertised as "coilovers".

[EG5]

Also better suspension bushings all around help alot
rear trailing arm bushings , front and rear lower arm bushings , sway bar bushings etc

[ekfour97]

thanks will look into konis and bilstein and new bushings all round

[JohnL]

Do not buy strut bars.

Yes, do fit suspension tower braces (aka 'strut bars'), but only if they are very rigid (typical Ebay offerings are not and as a result many are merely ineffective bling).

Tower braces do a lot to significantly improve the steering / handling response, but not if they flex (as so many do...). With most poor quality braces, the lack of rigidity is mostly in the brackets that attach to the tower tops, but it can be in the tube itself, especially smaller OD / thinner wall tubes, or flattened section tubes (the tube stiffness is largely defined by the smallest cross sectional dimension).


They will not reduce the amount of body roll your car has.

Chassis flex in a localised area (say at the front or rear of the chassis, where the suspension attaches) will result in slight additional effective roll motion in the rest of the chassis because the chassis flexure acts somewhat like additional suspesion deflection. However, the effect is very small unless the flexure is substantial.

A chassis can never be too rigid, but can easily be too flexible. Anything done that actually increases chassis stiffness is a good thing, so long as it doesn't also add too much mass.


They will however probably affect your perception of the roll of your car and your perception of the handling response.

Good (rigid) strut braces make an actual difference, not just a percieved difference. Having daid that, for a road car all that really matters is how the car subjectively feels to drive.


Roll stiffness can also be altered with suspension geometry changes but this is quite difficult and out of the realm of what is easily achievable with bolt on parts.

Note that more than relatively modest lowering of the ride height can have very substantial affects on the suspension (and steering) geometry, nearly always bad.


Another way to reduce body roll is to lower your centre of gravity by lightening or removing parts that are high up in the chassis. This will reduce the roll moment car's mass applies to the suspension and hence reduce roll angle.

Removing heavier parts can leave the car rather spartan for day to day use. Lowering heavier components is usually quite hard to do. The CG is most often lowered by means of lowering the ride height, but it's easy to get carried away here, and many people do...

Lowering the CG will always result in less weight transfer (good), but may or may not result in a change in roll motion. This is because as the ride height is lowered the geometric roll centre also lowers, and it may well lower more than the CG does.

In this case it's possible that roll motion may increase rather than decrease. This may happen because with a lower GRC a lesser % of the total weight transfer will occur 'geometrically' (through the suspension linkage) and more will transfer 'elastically' (through the springs and anti-roll bars, ignoring dampers which only act transiently, as important as this is).

Less weight will transfer with a lowered CG, but it's possible that all / or most / or much of this reduction will be in the amount transferring 'geometrically', and that some increase in the amount transferring 'elastically' may possibly occur even though less weight transfer in total is occuring. It just depends how far the GRC drops as the ride height is dropped, i.e. if the GRC lowers more than the CG, then more roll may result despite less weight transfer.

[chargeR]

Chassis flex in a localised area (say at the front or rear of the chassis, where the suspension attaches) will result in slight additional effective roll motion in the rest of the chassis because the chassis flexure acts somewhat like additional suspension deflection. However, the effect is very small unless the flexure is substantial.

A chassis can never be too rigid, but can easily be too flexible. Anything done that actually increases chassis stiffness is a good thing, so long as it doesn't also add too much mass.



You are correct that a strut tower bar will have a slight effect of the roll of the chassis, since there will be flexure between the chassis and the suspension pickup points. I didn't consider that this amount was significant enough to warrant mention, but I will be more cautious in the future considering that you are lurking about to pick my posts apart ;).

My opinion is still that chassis bracing, to the degree that most will attempt it, is generally not the best way to spend money for performance. If you disagree that is fine :).



Removing heavier parts can leave the car rather spartan for day to day use. Lowering heavier components is usually quite hard to do. The CG is most often lowered by means of lowering the ride height, but it's easy to get carried away here, and many people do...

Lowering the CG will always result in less weight transfer (good), but may or may not result in a change in roll motion. This is because as the ride height is lowered the geometric roll centre also lowers, and it may well lower more than the CG does.

In this case it's possible that roll motion may increase rather than decrease. This may happen because with a lower GRC a lesser % of the total weight transfer will occur 'geometrically' (through the suspension linkage) and more will transfer 'elastically' (through the springs and anti-roll bars, ignoring dampers which only act transiently, as important as this is).

Less weight will transfer with a lowered CG, but it's possible that all / or most / or much of this reduction will be in the amount transferring 'geometrically', and that some increase in the amount transferring 'elastically' may possibly occur even though less weight transfer in total is occuring. It just depends how far the GRC drops as the ride height is dropped, i.e. if the GRC lowers more than the CG, then more roll may result despite less weight transfer.

I mentioned removing/lightening parts located high in the car to lower the CG because it has the desirable effect of reducing the roll of the car without affecting the suspension in any way ;). At no point did I recommend lowering the car/suspension itself.

All good points :thumbsup:.

I am surprised you haven't mentioned to the OP that reducing body roll just for the sake of it isn't the greatest way to arrive at a good handling car. The terminal roll angle of a car during a steady state corner is just one aspect of a cars performance. Some roll is of course necessary to make a car progressive and easy to drive.

[Bludger]

I will be more cautious in the future considering that you are lurking about to pick my posts apart ;).I LOL'ed

[JohnL]

I didn't consider that this amount was significant enough to warrant mention,

It probably isn't a significant effect (roll motion effect) in the majority of cases, but then I think that should be stated, not that it has no affect...


but I will be more cautious in the future considering that you are lurking about to pick my posts apart ;).

Apologies for the pedantry, but I like things to be clear, it helps avoid misunderstandings.


My opinion is still that chassis bracing, to the degree that most will attempt it, is generally not the best way to spend money for performance. If you disagree that is fine :).

There are all sorts of braces available, or that can be made. Some are next to useless, some somewhat effective, and some very effective. The area around / near / between the the suspension towers is on most cars two of (front / rear) the least rigid areas of the chassis, so stiffening here is IMO rarely not worth doing.

I know you were specifically talking about their affect on roll motion, but the impression you gave (at least to me) was that tower braces were not effective nor useful additions, basically a waste of money. I would agree that some are, but well designed (rigid) braces do make a very noticable difference, at least with some cars.

There are many people who will tell you flatly that tower braces are worthless and make no difference, but there are also many people who's only experience of tower braces is the bling from Ebay. If I remove my car's tower braces it instantly feels significantly softer in it's steering and handling responses.


I mentioned removing/lightening parts located high in the car to lower the CG because it has the desirable effect of reducing the roll of the car without affecting the suspension in any way ;).

Utterly agreed, and do if it you can.


At no point did I recommend lowering the car/suspension itself.

Didn't mean to put words in your mouth. I was extrapolating to the most commonly used means (by far) that people use to lower the chassis mass (since very few people will attempt to lower or lighten individual components). This doesn't mean that most people who lower the chassis are actually doing it in an attempt to improve the chassis dynamics (if they are honest with themselves...).


I am surprised you haven't mentioned to the OP that reducing body roll just for the sake of it isn't the greatest way to arrive at a good handling car. The terminal roll angle of a car during a steady state corner is just one aspect of a cars performance. Some roll is of course necessary to make a car progressive and easy to drive.

It's probably a mistake to encourage me to say more...

I know what you're saying, and I both agree and don't quite agree.

If reducing / eliminating body roll is considered to be an end in and of itself then we are not unlikely to end up with a rather strangely or even dangerously handling car. But, most cars (other than pretty serious and focussed high performance cars) do roll rather excessively and can benefit in a number of ways from reducing it quite a lot.

OP, you should be aware that the manner in which changing the roll stiffness at one end of the car (whether with a stiffer anti-roll bar or stiffer springs) will cause a reduction in grip at that end of the car, whilst increasing grip at the other end (this is called the 'roll couple'). This means that front vs rear roll stiffness can be and is used as a means to adjust under / over steer tendencies.

From a handling perspective this is the primary purpose of changes to roll stiffness, any changes to the way the car feels to drive is only a by-product. But having said that, for a road car the way the car subjectively feels to drive enjoys a much higher priority than for a race car, up to a point...

What happens is that as you increase roll stiffness (say at the front end only), you also increase the lateral weight transfer at that end of the car, i.e. when cornering the stiffened front end of the car will transfer more weight (from the IF wheel to the OF wheel), which will tend to result in less front end grip.

At the same time, there will be a decrease in the amount of lateral weight transfer from IR to OR wheels (even though no change was made at the rear end), resulting in an increase in rear grip. When we add the loss of front grip and the gain in rear grip, we end up with increased understeer created by an increase in front roll stiffness. This also works in reverse for an increase in oversteer / decrease in understeer (by decreasing front roll stiffness or increasing rear roll stiffness).

Now, there is at least one other factor that complicates the rather simplistic formula above. Some cars (most commonly those with MacPherson Strut front suspensions) suffer from such severe camber change (relative to the road surface) with roll motion that neither wheel (say front wheels) can present it's tread to the road in a manner that creates a large contact patch, and this results in poor front grip / understeer even if front roll stiffness might be low relative to rear roll stiffness (i.e. even if front WT were high relative to rear WT).

With such cars it's quite possible often to reduce understeer by increasing front roll stiffness because this prevents the wheels tipping over so much when cornering hard, thus increasing the size of the contact patch and improving grip because of that. This is even despite the increase in lateral front WT / decrease in rear WT caused by the increase in front roll stiffness. However, if the suspension geometry is good (typical well designed double wishbone suspension...), this is much less likely to be a consideration because with body roll the wheels don't lean over enough to cause a big enough problem.

Note too that very (excessively?) roll stiff cars will tend to be quite stiff in general, and while they are likely to be very responsive / reactive they may well have trouble negotiating less than racetrack smooth surfaces without sometimes quite abrupt / erratic / unpredictable / substantial changes in grip, to the point that the car may possibly be quite dangerous at open road speeds on real world roads, especially if the driver is driving near the limit of adhesion...

With very roll stiff cars such unforgiving behaviour may not be too bad in the dry, but will tend to be very much worse in wet / slippery conditions...

[chargeR]

Can't be bothered splitting up your post to agree with it, but good post ;). I think perhaps my bias against strut braces comes from driving one of the more modern Hondas, which if you believe the propoganda, are much stiffer than the older models. This may be contradicting what I said earlier, but I believe that strut braces have a functional effect and are certainly not worthless, but they seem to be much a much too common first step in suspension tuning.

Great info on roll stiffness and reducing body roll :thumbsup:.

[Sydneykid]

Strut braces add rigidity to the body with the express purpose of maintaining the static alignment settings under dynamic loads. The are only effective where they contribute to the rigidity where suspension components are mounted. For example linking the front strut towers with a brace efectively doubles the strength of the mounting points for the upper wishbones on an say EF, EG, EK or DC2.

Since strut braces are not directly connected to the suspension, they do absolutely nothing to reduce body roll or pitch or dive. The best devices for controlling roll are the antiroll bars (go figure), commonly called swaybars or rollbars or stabiliser bars. As any suspension engineer will tell you, they are also without doubt the most cost effective handling improvement you will make.


Cheers
Gary

[JohnL]

Strut braces add rigidity to the body with the express purpose of maintaining the static alignment settings under dynamic loads.

Gary,
I don't agree, despite what may be read on the internet. IMO, while the tower braces do act somewhat usefully to stabilise dynamic suspension geometry (i.e. lessening dynamic camber changes under hard cornering), I don't think this is their primary affect.

If it were, then tower braces would only have an affect once well into the corner and substantial lateral forces have developed, and it would manifest as grip rather than as ‘response’, which is what it creates (i.e. the mainly noticeable affect of good tower braces is to increase steering and handling responsiveness). The actuality is that tower braces have a significant and noticeable affect even when lateral forces are very low or even zero.

The additional chassis stiffness imparted by tower braces acts (mostly) to increase the effective torsional rigidity of the chassis, that is to reduce the partial chassis twist that occurs at / near the towers when one diagonally opposed pair of wheels becomes significantly more loaded than the other diagonally opposed pair. This causes all vertical forces to be elastically absorbed within the springs and dampers, not within a flexing chassis (or at least less within a chassis flexing less).

The more torsionally flexible the chassis is, the more vertical road force is absorbed within the chassis itself, which makes the transfer of weight (and the ‘jacking’ of weight) less immediate, less precise, less damped, and less predictable. This also makes the suspension less responsive to changes of spring, damper and ARB rates (a very flexible chassis can be fairly unresponsive to other than quite large spring, damper and ARB changes, while a very rigid chassis is typically much more responsive to smaller changes).

You can feel these affects of a good tower brace even when the car is not cornering. For example, if you were to hit a significant bump with say the LF wheel; with the tower brace in position you can feel the force imparted into the chassis being more ‘faithfully’ transferred diagonally through the chassis to the RR suspension, and feel this more strongly than you can without the tower brace. Without the brace more of the force is ‘lost’ in the chassis flexure, and the car tends to feel a bit more ‘doughy’ in it’s handling

This is immediately apparent when you fit the tower brace and drive the car on a bumpier road surface, and if it isn't then I'd suggest that the driver has an insensitive 'buttometer' or (not at all unlikely) that the particular brace is crap (i.e. for 'crap' read 'not rigid enough'). The benefits of this increased chassis rigidity also enhance the manner in which weight transfer occurs when cornering.

The are only effective where they contribute to the rigidity where suspension components are mounted.

'Added' chassis stiffness is beneficial anywhere between the front and rear axle lines, and indeed some braces fitted in front of the front axle line or behind the rear axle line have the potential to increase torsional chassis stiffness.

This is why with closed and open top variants of a given chassis the open top versions always require a lot of additional (and heavy) torsional reinforcement in the floorpan, i.e. to recoup torsional rigidity lost when the roof was hacked off (not a good thing from a structural point of view). Even so, the open top variants are rarely as stiff as the closed top variants, and as a result rarely handle as well...


For example linking the front strut towers with a brace efectively doubles the strength of the mounting points for the upper wishbones on an say EF, EG, EK or DC2.

I think you're assuming that the lateral forces (created by hard cornering) at the upper wishbone (or top of the damper rod with a Mac Strut) is primarily responsible for the lateral displacement of the suspension towers when cornering. I would contend that while this is not insignificant, it's the vertical forces imparted through the springs and the damper that are responsible for most of the lateral displacement of the towers under load.

The towers as sections of the chassis are very rigid in a somewhat vertical orientation, but much less so laterally. However, the vertical force applied to the towers through the spring and damper is not (usually) in a direct alignment with the axis of the tower’s greatest rigidity, so, the towers deflect in the weakest orientation (i.e. laterally) in response to a primarily but not purely vertical force.


Since strut braces are not directly connected to the suspension, they do absolutely nothing to reduce body roll or pitch or dive.

I would contend that tower braces are ‘directly’ attached to the spring / damper seats, so are directly involved with providing a ‘solid’ resistance to the forces passing through the spring / damper. Take the brace away and the vertical spring / damper forces will increase lateral displacement of the tower, and this compliance will act as if it were an undamped ersatz spring in series with the spring proper (albeit a very short travel spring relative to the spring proper).

This will see the unbraced / torsionally flexible front and / or rear sections of the chassis acting to some degree as if they were poorly controlled ersatz ‘suspension’ relative to the stiffest section of the chassis, that section being the passenger compartment. The less rigid the front and / or rear sections the more ‘roll’ like motion will tend to occur in the stiffer passenger section of the chassis. However, any additional roll like motion of the passenger compartment created by flexure of the front or rear sections of the chassis will be very small, unless the flexure is quite substantial.

.
The best devices for controlling roll are the antiroll bars (go figure), commonly called swaybars or rollbars or stabiliser bars. As any suspension engineer will tell you, they are also without doubt the most cost effective handling improvement you will make.

I don’t disagree with you here.

[JohnL]

but I believe that strut braces have a functional effect and are certainly not worthless, but they seem to be much a much too common first step in suspension tuning.

I suspect this may be because the majority of people who fit tower braces really only do so in order to have something to show when they open their bonnet, and don't really care too much how it affects the handling...

[saikou27]

hey johnL whats ur opinion on the honda oem strut brace from the dc2? i put one in my EK and it seemed to make a difference but it could have just been placebo. would u say the oem brace is just as worthless as an ebay bling one and that its worth upgrading to a stiffer more functional one?

[JohnL]

hey johnL whats ur opinion on the honda oem strut brace from the dc2? i put one in my EK and it seemed to make a difference but it could have just been placebo. would u say the oem brace is just as worthless as an ebay bling one and that its worth upgrading to a stiffer more functional one?

I haven't seen one (other than vague memory of seeing a photo or two quite a while ago), but Honda employs very good engineers, I'd be surprised if it weren't a good thing. Do you have any photo's?

[saikou27]

please excuse the quality of my fone camera

http://i422.photobucket.com/albums/pp304/saikou27/Image000.jpg
http://i422.photobucket.com/albums/pp304/saikou27/Image008-2.jpg
http://i422.photobucket.com/albums/pp304/saikou27/Image009.jpg

from the looks of it, the tubing has about a 25mm diameter

[JohnL]

To my eye it looks to be quite well designed. It's unfortunate that the tube is bent, but clearances are obviously an issue in that engine bay. Bends weaken the tube in both compression and tension, so if bends are needed then the OD of the tube should go up and/or the wall thickness substantially increase.

I would expect that brace to be quite adequate, though such a brace can't realistically be too rigid. If I owned the car I doubt I'd be spending the $ to change it (not that I'd ever buy a tower brace, I design / make my own).

[Sydneykid]

QUOTE]
Gary,
I don't agree, despite what may be read on the internet. IMO, while the tower braces do act somewhat usefully to stabilise dynamic suspension geometry (i.e. lessening dynamic camber changes under hard cornering), I don't think this is their primary affect.
What is their primary effect then? They have no effect whatsoever on roll, they don't increase (or decrease) the spring or damper rates, they don't increrase (or decrease) longitudinal rigidity, they don't increase (or decrease) vertical rigidity etc etc. The only thing they do is increase lateral rigidity by linking the 2 strut towers.




If it were, then tower braces would only have an affect once well into the corner and substantial lateral forces have developed, and it would manifest as grip rather than as ‘response’, which is what it creates (i.e. the mainly noticeable affect of good tower braces is to increase steering and handling responsiveness). The actuality is that tower braces have a significant and noticeable affect even when lateral forces are very low or even zero.
As soon as you input steering there is an immediate weight transfer, both laterally and diagonally. This weight transfer together with the steering input loads the outside tyres, mostly the front. A strut brace that links the towers effectively doubles their lateral rigidity hence helps maintain the suspension geometry. In simple terms you don't lose camber due to the inwards flex of the strut tower.




The additional chassis stiffness imparted by tower braces acts (mostly) to increase the effective torsional rigidity of the chassis, that is to reduce the partial chassis twist that occurs at / near the towers when one diagonally opposed pair of wheels becomes significantly more loaded than the other diagonally opposed pair. This causes all vertical forces to be elastically absorbed within the springs and dampers, not within a flexing chassis (or at least less within a chassis flexing less).
The strut brace that links the towers only acts in one plane ie; it maintains the relationship (distance) between the strut towers. It has almost zero effect on twist and it most certainly has absolutely zero effect on vertical flex.




The more torsionally flexible the chassis is, the more vertical road force is absorbed within the chassis itself, which makes the transfer of weight (and the ‘jacking’ of weight) less immediate, less precise, less damped, and less predictable. This also makes the suspension less responsive to changes of spring, damper and ARB rates (a very flexible chassis can be fairly unresponsive to other than quite large spring, damper and ARB changes, while a very rigid chassis is typically much more responsive to smaller changes).
Absolutely, but I'm not sure what this has to do with a strut brace.




You can feel these affects of a good tower brace even when the car is not cornering. For example, if you were to hit a significant bump with say the LF wheel; with the tower brace in position you can feel the force imparted into the chassis being more ‘faithfully’ transferred diagonally through the chassis to the RR suspension, and feel this more strongly than you can without the tower brace. Without the brace more of the force is ‘lost’ in the chassis flexure, and the car tends to feel a bit more ‘doughy’ in it’s handling
I would put that down to placebo effect, you have paid good money and spent time installing a strut brace on the car so you truly believe that you can feel it.

I prefer the more measured test, the same as what we apply when designing a new roll cage. Put the car on axle stands, jack up 1 corner and measure the movement at the other 3 corners. Then add your strut brace and repeat the test. Based on over 30 years (yes, I am that old) of doing this on race cars I can guarantee you won't be able to measure any increase in rigidity from the fitting of a simple strut brace.




This is immediately apparent when you fit the tower brace and drive the car on a bumpier road surface, and if it isn't then I'd suggest that the driver has an insensitive 'buttometer' or (not at all unlikely) that the particular brace is crap (i.e. for 'crap' read 'not rigid enough'). The benefits of this increased chassis rigidity also enhance the manner in which weight transfer occurs when cornering.
A simple strut brace that links the towers does its job, it will help maintain lateral suspension geometry. Once you get past that you are getting into the relms of roll cages, the domain of 3 dimensional triangulation. Quite simply there is insufficient space under the bonnet line to facilitate vertical triangulation in a strut brace. That's why we extend the roll cage elements through the firewall to the strut towers.





'Added' chassis stiffness is beneficial anywhere between the front and rear axle lines, and indeed some braces fitted in front of the front axle line or behind the rear axle line have the potential to increase torsional chassis stiffness.
I dissagree for the reasons posted above.





This is why with closed and open top variants of a given chassis the open top versions always require a lot of additional (and heavy) torsional reinforcement in the floorpan, i.e. to recoup torsional rigidity lost when the roof was hacked off (not a good thing from a structural point of view). Even so, the open top variants are rarely as stiff as the closed top variants, and as a result rarely handle as well...
Absolutely, but I'm not sure what this has to do with a strut brace.



I think you're assuming that the lateral forces (created by hard cornering) at the upper wishbone (or top of the damper rod with a Mac Strut) is primarily responsible for the lateral displacement of the suspension towers when cornering. I would contend that while this is not insignificant, it's the vertical forces imparted through the springs and the damper that are responsible for most of the lateral displacement of the towers under load. The towers as sections of the chassis are very rigid in a somewhat vertical orientation, but much less so laterally. However, the vertical force applied to the towers through the spring and damper is not (usually) in a direct alignment with the axis of the tower’s greatest rigidity, so, the towers deflect in the weakest orientation (i.e. laterally) in response to a primarily but not purely vertical force.

I'm not sure that I understand the logic here. If the spring/shock unit is mounted vertically then the only effect it can have is a vertical one. Obviously that doesn't translate to lateral displacement of the shock towers. This is of course the case with double wishbone Hondas. In a MacPherson strut design there is some angling of the strut, but that is mostly rearwards for caster and SAI reasons. In order for lateral loading to be a consideration, the MacPherson struts would have to be at a considerably greater angle than their packaging would allow.

More importantly spring/shock units, particularly MacPherson struts, don’t tolerate angular change, that’s what bends shock shafts. Hence they are designed to operate in one plane, straight line compression and extension.

In simple terms, any lateral loads transferred from spring shock units are so miniscule as to be irrelevant in a road car based chassis.







I would contend that tower braces are ‘directly’ attached to the spring / damper seats, so are directly involved with providing a ‘solid’ resistance to the forces passing through the spring / damper. Take the brace away and the vertical spring / damper forces will increase lateral displacement of the tower, and this compliance will act as if it were an undamped ersatz spring in series with the spring proper (albeit a very short travel spring relative to the spring proper).
This will see the unbraced / torsionally flexible front and / or rear sections of the chassis acting to some degree as if they were poorly controlled ersatz ‘suspension’ relative to the stiffest section of the chassis, that section being the passenger compartment. The less rigid the front and / or rear sections the more ‘roll’ like motion will tend to occur in the stiffer passenger section of the chassis. However, any additional roll like motion of the passenger compartment created by flexure of the front or rear sections of the chassis will be very small, unless the flexure is quite substantial.
Perhaps an opportune time to talk about spring rates, or more importantly relative spring rates. In a DC2 Integra race car for example we use front spring rates around 800 to 1000 lbs per inch, (14 to 18 kg/mm) ie; some 4 to 5 times higher than standard. From memory a standard, new, DC2 has a chassis rigidity around 30,000Nm/degree of twist. A simple mathematical comparison with a bit of high school Pythagoras thrown in would show full deflection of the spring, around 75 mm on a DC2, would take a force of 1000 to 1300 kgs at the spring. Obviously in this comparison suspension leverage and movement ratios (to the tyres) are irrelevant. A full force bump impact of that size would deflect (compress) the spring 75 mm while the same sized impact would flex the chassis around 1 mm.

I don’t know about you, but my butt isn’t sensitive enough to feel 1 mm of chassis deflection in comparison to 75 mm of spring deflection. Of course that’s the standard deflection, how much rigidity would a simple strut brace add to that? Obviously I believe 0%, but let’s say you believe 10%. Then that would mean you could feel the difference in the above example between 1 mm and 0.9 mm ie; 0.1mm. Which would be a similar feeling to changing the spring rate from 1000 lbs/inch to 1001 lbs/inch. Now I work with some pretty good race car drivers and they would have difficulty in picking the change from 1000 lbs/inch to 1050 lbs/inch, let alone 1/50th of that.

In conclusion, the numbers tell us that any vertical rigidity effect you are feeling from a simple strut brace is placebo. There is quite simply no other effect that can possibly be felt, other than the lateral maintenance of geometry.



Cheers
Gary

[beeza]

Doing this significantly reduced my body roll,significantly!

http://www.ozhonda.com/forum/showthread.php?p=2495923#post2495923

[Sydneykid]

Doing this significantly reduced my body roll,significantly!

http://www.ozhonda.com/forum/showthread.php?p=2495923#post2495923

The only bushes that reduce roll are those in the lower control arms, front and rear, where the shock/spring units bolt on. Which I notice you didn't do.

What you are feeling is the decreased later and longitudinal movement from installing the Mugen hardness up bushes. Very worthwhile doing for that reason, we ran them in the Bathurst 12 Hour in the EG Civic in 1994, but zero contribution to reducing roll.


Cheers
Gary

[beeza]

The only bushes that reduce roll are those in the lower control arms, front and rear, where the shock/spring units bolt on. Which I notice you didn't do.

What you are feeling is the decreased later and longitudinal movement from installing the Mugen hardness up bushes. Very worthwhile doing for that reason, we ran them in the Bathurst 12 Hour in the EG Civic in 1994, but zero contribution to reducing roll.


Cheers
Gary

I was gonna say 'What?' then I see you know what your talking about :D hehehe

"The only bushes that reduce roll are those in the lower control arms, front and rear"

Gotcha :thumbsup:

I have the Front LCA to put in,I think I will have to get the rear ones aswell,what do ya reckon?

Thanks Gary!

[mocchi]

The only bushes that reduce roll are those in the lower control arms, front and rear, where the shock/spring units bolt on.

Cheers
Gary

why is this? can you please explain?

thanks.

[Sydneykid]

why is this? can you please explain?

thanks.

The spring and shock units operate through those bushes and, being factory soft and far from new, they distort under load. This effectively lowers the spring and damper rates until the bush is fully distorted. You have around 2 to 3 mm of wheel travel that doesnt fully act on the spring and shock. So you get 2 to 3 mm of body roll. By fitting new hardness up bushes you take some of that distortion out of the bush so that the spring and shock act earlier in controlling the roll. That's why in race cars we remove the rubber bushes and replace with polyurethane, or in top end race applications spherical bearings, which I would never recommend for a road car.

It's small but it's there, most people wouldn't feel it.

Cheers
Gary

[beeza]

It's small but it's there, most people wouldn't feel it.

OMG!!!!

This is absolute MUSIC TO MY EARS GARY!! :thumbsup: :thumbsup: :thumbsup:

When you are fully in tune with your car,any change is noticeable!

[EG30]

If this ek4 was my daily road car that I want good handling I would:

put stiffer springs ( no less than 8kg Front 6kg Rear, up to 11kg front 7-8kg rear ) with good dampers to match

stiffer bars front and rear from EK9 and rear subframe reo from ASR ( yes cost lot more than whiteline but you wont regret it in the long run )

bushes need to be in good shape thru out the car of course

as for the strut braces, something like ultra racing fixed 2 points front and rear would do fine.

I went thru many different combos in my eg which was a daily and now track car ( wont bored you with the details by writing a book ), learnt it's a balancing act and you cant get away with soft springs and stiff bar or soft bar with stiff springs in our cars without major compromises.

[Sydneykid]

[QUOTE=EG30;2533660]If this ek4 was my daily road car that I want good handling I would:
put stiffer springs ( no less than 8kg Front 6kg Rear, up to 11kg front 7-8kg rear ) with good dampers to matchThe standard spring rates are actually pretty good, they really only need good quality shocks, Konis or Bilsteins are what I use.




stiffer bars front and rear from EK9The front is OK but the rear swaybar is not big enough, plus it really needs to be adjustable.



and rear subframe reo from ASR ( yes cost lot more than whiteline but you wont regret it in the long run )Since I helped in the design of the Whiteline kit, I just gotta ask why?




bushes need to be in good shape thru out the car of courseand adjustable. It's a wate of time and money doing the above without the ability to optimise the geometry.




as for the strut braces, something like ultra racing fixed 2 points front and rear would do fine.The front strut brace for sure, its actually does something. But the rear is a waste of time and money, they do nothing.




I went thru many different combos in my eg which was a daily and now track car ( wont bored you with the details by writing a book ), learnt it's a balancing act and you cant get away with soft springs and stiff bar or soft bar with stiff springs in our cars without major compromises.
I agree on the swaybars, the rear particularly, but dissagree on the springs. Optimising the suspension geometry is far more important that upgrading the springs rates.


Cheers
Gary

[weeman]

sydney kid, can you elaborate a bit more on optimising the suspension geometry and what the best way to go about that would be

[Sydneykid]

sydney kid, can you elaborate a bit more on optimising the suspension geometry and what the best way to go about that would be


Why to optimise?
It's pretty simple actually, maintaining the tyre's contact with the road (or track) is the most important thing in improving handling and braking and acceleration. Setting up the suspension geometry is the way to achieving that. That means the dynamic as well as the static wheel alignment.

If you have a 225 tyre then it makes sense to use all 225 mm of it as often as possible. It's pretty stupid to have a 225 tyre and only use 165 mm of it, you might as well have saved yourself some money and bought a 165 tyre.


How to optimise?
Start off wiht adjustable caster, camber and toe for the static alignment. Then working on the dynamic, for example its waste of time spending a lot of effort on setting up the static toe if the car has massive bump steer.


What to optimise?
Like, most Japanese cars Hondas never have enough caster, so that's the first thing I always work on. At least 6 degrees of caster, 8 is better.

Then I set the front static ride height up such that the driveshafts are flat, parallel to the ground. No use wasting horsepower on ridiculous drive shaft angles. Then set up the rear ride height to match, I usually start off with 10 mm of rake (nose down), measured at the sills. We do play with rear ride height a lot as it affects the handling balance, so its a good tuning aid.

I do set the race cars up on the corner weight scales, so any ride height adjustments at the track are done in spring pairs.

Next is front camber, around 1 degree is about all I use for a road car, on the race cars we use up to 4.5 degrees. On the rear around 0.75 degrees for the road and up to 2.5 degrees for the track. Each track is different and some days also require camber tuning. I always use a pyrometer as tyre temperatures are the best indicator of set up.

Lastly toe, on a road car a little front toe out is OK, around 2 mm per side, not too much as they tram line rather badly. On the race cars we run up to
6 mm toe out per side. Rear toe out is something we adjust a lot, especially on Production Cars where we don't have the ability to adjust swaybar settings.

Once I have the static settings where I want them I then work on the dynamic, removing bump steer, optimising the camber curves, maybe changing the Ackerman. To set up a new race car from scratch that I am unfamiliar with usually takes about 4 to 5 days at 10 or so hours a day. Obviously once I am familiar with the car and its adjustments I can do a pre race meeting set up in around half a day.

That's about all I have time for right now, work beckons.
Cheers
Gary

[mocchi]

Why to optimise?
It's pretty simple actually, maintaining the tyre's contact with the road (or track) is the most important thing in improving handling and braking and acceleration. Setting up the suspension geometry is the way to achieving that. That means the dynamic as well as the static wheel alignment.

If you have a 225 tyre then it makes sense to use all 225 mm of it as often as possible. It's pretty stupid to have a 225 tyre and only use 165 mm of it, you might as well have saved yourself some money and bought a 165 tyre.
dont quite understand what can cause this garry. wrong camber?

How to optimise?
Start off wiht adjustable caster, camber and toe for the static alignment. Then working on the dynamic, for example its waste of time spending a lot of effort on setting up the static toe if the car has massive bump steer.
i learnt about bump steer. thanks for the information gary.

What to optimise?
Like, most Japanese cars Hondas never have enough caster, so that's the first thing I always work on. At least 6 degrees of caster, 8 is better.
im assuming this is with eg? i saw your autocross practice with the eg.
btw is that + or -?

Then I set the front static ride height up such that the driveshafts are flat, parallel to the ground. No use wasting horsepower on ridiculous drive shaft angles. Then set up the rear ride height to match, I usually start off with 10 mm of rake (nose down), measured at the sills. We do play with rear ride height a lot as it affects the handling balance, so its a good tuning aid.

I do set the race cars up on the corner weight scales, so any ride height adjustments at the track are done in spring pairs.

Next is front camber, around 1 degree is about all I use for a road car, on the race cars we use up to 4.5 degrees. On the rear around 0.75 degrees for the road and up to 2.5 degrees for the track. Each track is different and some days also require camber tuning. I always use a pyrometer as tyre temperatures are the best indicator of set up.

Lastly toe, on a road car a little front toe out is OK, around 2 mm per side, not too much as they tram line rather badly. On the race cars we run up to
6 mm toe out per side. Rear toe out is something we adjust a lot, especially on Production Cars where we don't have the ability to adjust swaybar settings.

Once I have the static settings where I want them I then work on the dynamic, removing bump steer, optimising the camber curves, maybe changing the Ackerman. To set up a new race car from scratch that I am unfamiliar with usually takes about 4 to 5 days at 10 or so hours a day. Obviously once I am familiar with the car and its adjustments I can do a pre race meeting set up in around half a day.
can you please explain what it means by camber curves?

btw, according to longacreracing (http://www.longacreracing.com/articles/art.asp?ARTID=13):
In order to accomplish zero bump the tie rod must fall between an imaginary line that runs from the upper ball joint through the lower ball joint and an imaginary line that runs through the upper a-arm pivot and the lower control arm pivot. In addition, the centerline of the tie rod must intersect with the instant center created by the upper a-arm and the lower control arm.

can this be achieved in civics? im not sure where the a-arm pivot and lca pivot is... :(

edit: found this (http://www.hotrod.com/techarticles/chassis/hrdp_0411_wheel_alignment_guide/index.html), which is very good for me to learn.

thanks alot for the help.

[beeza]

^^ Good Find!

And Thanks Gary,keep the knowledge coming plz mate!

[weeman]

thanx for taking the time to answer my question gary, very informative

[Sydneykid]

Why to optimise?
It's pretty simple actually, maintaining the tyre's contact with the road (or track) is the most important thing in improving handling and braking and acceleration. Setting up the suspension geometry is the way to achieving that. That means the dynamic as well as the static wheel alignment.

If you have a 225 tyre then it makes sense to use all 225 mm of it as often as possible. It's pretty stupid to have a 225 tyre and only use 165 mm of it, you might as well have saved yourself some money and bought a 165 tyre.
dont quite understand what can cause this garry. wrong camber?

For straight line (acceleration or braking) traction is optimised with zero camber. Caster is a good substiture for camber as it's only there when you turn the steering wheel, I like to call caster "camber on demand".

It would be be nice to run zero camber, we pretty much do on the Formula 3, but on a production based car you need camber to compensate for the roll. As the body rolls around a corner you need caster and camber to maximise the tyre contact patch. In a race car with totally adjustable camber curves there is an argument for what we call 1 to 1, where 1 degree of body roll = 1 degree of additional negative camber. This is hard to achieve on a production car, for example an EF/EG/EK where it's less tham 0.5 to 1 on the front and close to 1.5 to 1 on the rear. On the race cars I try and bring this closer together. within the regulations I can get 0.75 to 1 on the front and around 1 to 1 on the rear. This makes the car far more predicatable as there isn't the huge dynamic camber difference front to rear.


How to optimise?
Start off wiht adjustable caster, camber and toe for the static alignment. Then working on the dynamic, for example its waste of time spending a lot of effort on setting up the static toe if the car has massive bump steer.
i learnt about bump steer. thanks for the information gary.
It's so important, especially when changing say the height of the car outside its design parameters, but very seldom mentioned by the sellers of super low suspension.



What to optimise?
Like, most Japanese cars Hondas never have enough caster, so that's the first thing I always work on. At least 6 degrees of caster, 8 is better.
im assuming this is with eg? i saw your autocross practice with the eg.
btw is that + or -?
I don't currently have an EG, my previous EG was red, so was it old footage? Or maybe one of the race EG's I work on, white or orange?

Caster is +, I haven't seen - caster since before radial tyres.


Then I set the front static ride height up such that the driveshafts are flat, parallel to the ground. No use wasting horsepower on ridiculous drive shaft angles. Then set up the rear ride height to match, I usually start off with 10 mm of rake (nose down), measured at the sills. We do play with rear ride height a lot as it affects the handling balance, so its a good tuning aid.

I do set the race cars up on the corner weight scales, so any ride height adjustments at the track are done in spring pairs.

Next is front camber, around 1 degree is about all I use for a road car, on the race cars we use up to 4.5 degrees. On the rear around 0.75 degrees for the road and up to 2.5 degrees for the track. Each track is different and some days also require camber tuning. I always use a pyrometer as tyre temperatures are the best indicator of set up.

Lastly toe, on a road car a little front toe out is OK, around 2 mm per side, not too much as they tram line rather badly. On the race cars we run up to
6 mm toe out per side. Rear toe out is something we adjust a lot, especially on Production Cars where we don't have the ability to adjust swaybar settings.

Once I have the static settings where I want them I then work on the dynamic, removing bump steer, optimising the camber curves, maybe changing the Ackerman. To set up a new race car from scratch that I am unfamiliar with usually takes about 4 to 5 days at 10 or so hours a day. Obviously once I am familiar with the car and its adjustments I can do a pre race meeting set up in around half a day.
can you please explain what it means by camber curves?
refer above



btw, according to longacreracing (http://www.longacreracing.com/articles/art.asp?ARTID=13):
In order to accomplish zero bump the tie rod must fall between an imaginary line that runs from the upper ball joint through the lower ball joint and an imaginary line that runs through the upper a-arm pivot and the lower control arm pivot. In addition, the centerline of the tie rod must intersect with the instant center created by the upper a-arm and the lower control arm.

can this be achieved in civics? im not sure where the a-arm pivot and lca pivot is... :(

At standard height all the Civics that I have worked on, that's most models, have pretty much zero bump steer. Honda engineers are pretty smart, they design for that within the normal range of suspension travel. The problems start when the car is lowered so much that the suspension travel is operating outside its design range. You can pretty much guarantee that if you need shorter shocks then you are trying to run it too low and you will get bump steer. It's one (there are others) of the reasons why I dislike adjustable length shocks, there is no indicator of pending trouble, people just keep screwing them down in total ignorance of the consequences.



edit: found this (http://www.hotrod.com/techarticles/chassis/hrdp_0411_wheel_alignment_guide/index.html), which is very good for me to learn.

thanks alot for the help.
No problem
Cheers
Gary

[bennjamin]

Since I helped in the design of the Whiteline kit, I just gotta ask why?


Hi gary. I dont want to open up another can of worms....but if you truly did herlp in the design of the whiteline kit (rear swaybar i ask...) can i ask why it is faulty and poorly designed.

1. The subframe reinforcement , while currently slightly improved sitll lacks back plates or braces for the mounting to spread the force along. ( like BEAKS or ASR kits) I have personally witnessed over a dozen failed kits , with one being my own due to this issue.
2. Kits often have parts missing , and always have LOWER than OEM tensile rated bolts.
With an aftermarket part that is under perhaps more than OEM loads , how is it acceptable that the included fasteners are LESS than OEM rating ?

Sorry if i have harped on abit. But Whiteline itself fails to answer these questions and frankly has lost ALOT of customers due to this.



The poor under rated fasteners isnt limited to its rear swaybar kit either , most other parts ive seen are under rated. I had a front castor kit from whiteline that failed within months of fitting (at whilteine) due to a snapped bolt. Wouldnt of happened if it was OEM or above rate.

[Sydneykid]

Hi gary. I dont want to open up another can of worms....but if you truly did herlp in the design of the whiteline kit (rear swaybar i ask...) can i ask why it is faulty and poorly designed.

1. The subframe reinforcement , while currently slightly improved sitll lacks back plates or braces for the mounting to spread the force along. ( like BEAKS or ASR kits) I have personally witnessed over a dozen failed kits , with one being my own due to this issue.
2. Kits often have parts missing , and always have LOWER than OEM tensile rated bolts.
With an aftermarket part that is under perhaps more than OEM loads , how is it acceptable that the included fasteners are LESS than OEM rating ?

Sorry if i have harped on abit. But Whiteline itself fails to answer these questions and frankly has lost ALOT of customers due to this.



The poor under rated fasteners isnt limited to its rear swaybar kit either , most other parts ive seen are under rated. I had a front castor kit from whiteline that failed within months of fitting (at whilteine) due to a snapped bolt. Wouldnt of happened if it was OEM or above rate.

Unfortunately I helped design/specify the swaybar, but I wasn't consulted on the heavy duty mounting kit.

I will say that I have had 24 and 26 mm rear swaybars with that kit on 3 EG Civic race cars since 1999, and I've not had a single problem with any of them. Maybe the issue is not with the mounting kit itself but with the installation and other factors. For example if the car is too low and the swaybar and its mounting will be forced to operate outside their design parameters.

Fasteners wise, I agree with you, I would prefer higher tensile fasteners, but as with all things there is a cost factor. The engineer can specify whatever he sees necessary, but the accountant will always win if the cost is too high.

Cheers
Gary

[bennjamin]

Fasteners wise, I agree with you, I would prefer higher tensile fasteners, but as with all things there is a cost factor. The engineer can specify whatever he sees necessary, but the accountant will always win if the cost is too high.

Cheers
Gary

Cheers for the reply - the account won in the short term in this instance. The failures i saw (and lived through) were a mixture of failed bolts and torn subframe)

[Tegzieboi_BAR]

Upgrading swaybards and obviously the coilovers were definitely my best mods for reduction of body roll, as someone has already said, if ur on a budget then swaybars are the way to go.

Perhaps look into ITR Shocks (+LCA's) with a/market springs (koni seem to be quite good n not overly expensive) if ur after bang for your buck performance... but then again as u mentioned u were keen to get into racing so perhaps its best to go all out right away n get urself a decent set of coilovers.

my 2 cents

[vinnY]

itr shocks and rear lca's aren't gonna fit his ek(unless he wants rear itr shocks with function7 rear lca's to use them)

the most effective mod i did to my ek was a rear asr + 22mm sway bar
though supporting mods do make it nicer to drive

[Limbo]

i've seen over a dozen of this happen also and have to agree with Ben, The mounting brackets & bolts are really poor quality. The only good part of the kit is the sway bar itself.
I've got the first batch, without the extra reinforcement, but i've seen the updated models and the bolts and additional reinforcement are still sub standard. Certainly nothing to write home about

I've fixed this on my car by running the ASR brace in combo with the Whiteline kit and it certainly makes the difference. I've seen alot of those fix it with thie problem also


Unfortunately I helped design/specify the swaybar, but I wasn't consulted on the heavy duty mounting kit.

I will say that I have had 24 and 26 mm rear swaybars with that kit on 3 EG Civic race cars since 1999, and I've not had a single problem with any of them. Maybe the issue is not with the mounting kit itself but with the installation and other factors. For example if the car is too low and the swaybar and its mounting will be forced to operate outside their design parameters.

Fasteners wise, I agree with you, I would prefer higher tensile fasteners, but as with all things there is a cost factor. The engineer can specify whatever he sees necessary, but the accountant will always win if the cost is too high.

Cheers
Gary

[ekfour97]

thanks guys for all the info.....asr brace and 22mm swaybar coming very soon followed by coilovers/camber kit i reckon , cheers

[ekfour97]

got asr and ek9 swaybar today will comment when done

[aznalex]

wow what an intense learning experience this was..
anyone wanting info regarding strut bars and sway bars..
this is where your gonna get your feed!

got a few questions up my sleeve but will probably make a new post :P

[ekfour97]

so its been well over a month and all i can say is oh yeah ...... asr and swaybar made a massive difference . car handles like a dream .
ive since added coilovers , and you guessed it , no more body roll ........ thanks to all

[quangsuke]

sorry to hijack the thread, but does anyone have any sound knowledge on c and b pillars?
how about the floor bar?

thanks

[mnc]

I dont think they are that effective. Just bling factor says many.

Your best bet is to get stuff like a solid piece 3 point strut brace or a 4 point rear strut like the ultra racing ones =]