318ti Handling

[Pedro]

Well, after browsing through a few threads, I think I can´t find an answer for this from someone else´s question, not sure if I should post this in the project section, but anyway, here it is, my "problem" with my car´s handling!

 

I've got a BMW Compact, it´s a 1997 318ti, with stock engine, so fairly underpowered at just 140hp, not sure how much it weights since I haven´t gotten the chance to put it on cornerweights yet, it´ll happen soon enough I hope. This car was fitted with a welded in rollcage, with no crosses or anything, just to improve structural rigidity in case of a nasty crash at speed, but also had a rear strut bar welded between the struts, and saw the front of the car reinforced with a Sparco top strut bar, as well as a BMW cross brace that fits under the engine.

 

Here are some pics

 

 

 

 

Rear strut bar:

 

 

 

 

Front upper strut bar:

 

 

 

 

Front lower cross brace:

 

 

 

 

I replaced the normal seats with Sparco racing ones, and moved the battery to the boot, that´s how far I went with lightening it, or relocating weight. The rear seats, however, will be removed when used on track, and as such when it get on the weighting scales, those will not be in the car. That´s so that I can use the racing seatbelts, wrapped round the rear strut bar.

 

Regarding suspension, I changed all bushings except the anti sway bar ones, for 75D urethane ones, the rear bushings have an adjusting capability that allows me to adjust both camber and toe, although not individually nor very easily. For example, if I want to increase rear camber by 0.5º I will surely affect toe. This car has got a rear trailing arm suspension and the only way to adjust angles is to fit these bushes, or fit adjustment plates, I feel I should have fit the plates, but was in a hurry and didn´t feel like welding stuff there. It´s important to have rear wheel adjustment here because it varies a lot with height adjustments, and since I´ve got height adjustment ability ...

 

 

This is better explained in this article, if anyone feeling like a read: http://www.e30m3performance.com/tech_artic...sh/ecc_bush.htm

 

 

Fitted a KW Clubsport coil-over kit, with front and rear adjustable shock absorbers as well as ride height, and front camber plates.

 

 

I´ve got a thread over on FastClassics where I describe the entire thing, loads of crap talk and brainstorming as well, but it´s there, if you feel like going through it all: http://www.fastclassics.co.uk/phpBB3/viewt...?f=2&t=1168

 

 

After all this word being done a new set of Bridgestone RE050A tyres were fitted, and the thing was aligned, as you can imagine what chaos it was after being completely torn appart. I usually use high pressures, as I really hate feeling the tyres start to bend, so for now I´ve got 2.5 Bar pressure on all four tyres. I am trying to get it sorted without resorting to messing with tyre pressures, until I need to. I feel like it´s sort of cutting corners if I start with tyre pressures without trying to solve things by adjusting the suspension first.

 

As an aside comment, it left me mad that the mechanic at the shop where I bought the tyres declared it a lost battle from the start when he saw the rear bushings, after an hour of going nowhere I brought the car home and have since had it adjusted twice by my local mechanic, using a used machine, with both of us contributing, things take a lot less time and go a lot better when one just feels like doing something and helping.

 

After it´s second alignment, it´s now more or less like this:

 

Front Wheels:

Camber between 1.41º and 1.42º

Toe In = 0º

Caster: 3.38º on the left and 4.05º on the right

 

Rear Wheels:

Camber: 1.44º on the left and 1.42º on the right

Toe In: 0.11º on each wheel

 

 

The front caster is a bit wonky because it´s got excentric bushings, like the M3 ones, and I guess one might be slightly rotated in relation to the other, but I feel it´s OK, don´t you? I really don´t feel like taking the arm off to rotate the bushing.

 

This was the second alignment, on the first one the car was set up pretty much to factory specs, with 2.00º rear camber and 1.40º on the front. I wasnt impressed with this at all, and felt that it would ruin my tyres without actually helping handling at all.

 

Now to the question at hand, the shocks were assembled with the reccomended adjustments, meaning they´re set right at the middle of the setup, both bound and rebound on the front, and the rear at 75% stiff bound, and 12 clicks from fully closed rebound, there's 19 clicks to go.

 

Torsion bars are stock ones.

 

 

 

 

Thing is, car didn´t feel balanced, it felt like the front was not holding the road properly, you´d enter a corner, either fast or slow, bumpy or smooth, and it´d understeer, the car was never balanced even before loosing grip, the front always felt like a step ahead of the car when loosing grip. You´ld settle into a tight long corner at, say, 85kmh, like a motorway entrance, at the limit of grip the front wheels would skid, at this point if I tighten the steering wheel it´ll have no effect, same as if if I let go of the accelerator it´ll have very little effect on the overall balance.

On fast bumpy motorway like situations, at 190kmh, it´ld feel the same, understeery.

On low speed smooth roundabouts it felt like crap, the front would plow ahead while the rear would have supreme traction.

 

I could throw the car sideways into a roundabout and after half a second it´d be understeering. At first I messed with tyre pressures as I imagined that since the tyres were new they might not be scrubbed in yet, lowered pressure on all 4 wheels to about 2.2bar, all that changed was that now I could feel the wheels bending, the exactly the same would happen, only delayed by tyre bending.

 

Returned to the higher tyre pressures. Oh, forgot to say I´m using 225/50ZR16 tyres on 7.5 rims

 

It´s important to say that before any work had been done to it, the balance on roundabouts was perfect, it had worn tyres, good ones though, and a standard m-technic suspension with mushy bushings, felt a bit wavy at high speed but slow speed and balance was very good.

 

Decided to mess with the front shocks to see if a little more softness would help. Softened it all up, both bound and rebound were turned to 2 clicks away from full soft, this is next to as soft as possible. After a couple of frustrated attempts, because of traffic, I finaly got a moment with low traffic and was happy to see it had a positive effect.

On the same corner, at the same speed, the front wheels don´t feel so loose, they will still understeer if I turn the steering in at that point, but when lifting off at that same point one can now feel the front wheels digging into the ground, and the rear livening up, as it should. Car is now overall better balanced. I got the same feeling on another corner, this one downhill, but since I didn´t have a clear memory of how it was there before, decided to discard it.

Haven´t got a chance to test it at speed yet, as the weather´s been wet at night.

 

I´m trying to make this as scientific as possible, approaching the same simple corner at the same speed at checking for changes.

 

I still think the car should hold the road better, and should feel more meutral, but it´s a work in progress. Going to use the conerweights of a rally team, but they´re a bit busy lately, on account of races that involve a bit of logistic, so currently waiting for my chance to use their equipment, as a special favour, this kind of equipment isn´t very available for street use here.

 

This is as it stands now, any advice, opinion or suggestion is very appreciated.

 

 

 

 

 

EDIT: Sorry, I went on a bit, didn´t I?

[Tony]

Now that's what i call a "full report", nice one.

 

Apart from the fact i would use more rear camber and castor the rest of the calibration is good.. I feel like you the majority of the problem has been the damper ratios and since there is no common base for the modified car then all that can be done is what you have done "experiment".

 

Nice car by the way

[Pedro]

Would you start by having the rear wheels with more camber than the front ones? How so?

 

Is castor the same as caster? Sorry, english isn´t my first language.

 

 

 

 

Nice car by the way

 

Thanks !

[Tony]

Would you start by having the rear wheels with more camber than the front ones? How so?

 

Is castor the same as caster? Sorry, english isn´t my first language.

 

 

 

 

Nice car by the way

 

Thanks !

 

On the straight the deeper rear camber allows a secure fast road balance compared with the lesser front camber not dismissing the front camber gains due to the aerodynamics.

 

Corner on you have a "push" during Yaw. This lift off transition during weight transfer is not balanced since the inner wheels castor (castor/ caster is the same) does not sweep forward enough.

 

The diagonal castor sweep progressively lifts the outer wheels tyre saturation limits (corner on) leaving the driver to feel how loose the exit is using the steering and throttle to control corner out.

 

By combining the sweep corner out with the axle squat you will find transition much more palatable.

[Pedro]

First of all, thanks, that was exactly the kind of advice I was hoping for, very cool stuff, as I´m really starting to get interested in this, I´ll need a little help processing it though, as I simply don´t know if I got the meaning right. First time I read it I simply looked like this:

 

 

 

On the straight the deeper rear camber allows a secure fast road balance compared with the lesser front camber not dismissing the front camber gains due to the aerodynamics.

 

So a rear camber deeper than the front will provide stability on the straights, will this help eliminate some weaving, or floating, of the rear at high speed strong braking on straights?

 

In what way does front camber provide aerodymic gains?

 

 

 

 

Corner on you have a "push" during Yaw.

 

You lost me here , is this as you´re lifting off and turning in, the first moments the car starts to change direction?

 

 

 

 

This lift off transition during weight transfer is not balanced since the inner wheels castor (castor/ caster is the same) does not sweep forward enough.

 

The diagonal castor sweep progressively lifts the outer wheels tyre saturation limits (corner on) leaving the driver to feel how loose the exit is using the steering and throttle to control corner out.

 

You´re saying that a higher castor would increase front wheel limits at the moment the outer portion of the tyres are loaded during cornering. So by increasing castor I´d get a better steering feedback about the front axle´s limits as well as more grip?

 

 

 

 

By combining the sweep corner out with the axle squat you will find transition much more palatable.

[Tony]

Hec of a topic to put into text but hay this is wim.

 

The aerodynamic effect on the front will allow front camber gains on the straight proving a secure balance, simple down force.

 

Point to remember during weight transfer "corner entry" the steering action transports the front camber values so you cannot have aggressive front static camber positions since they are affected so violently dynamically.

 

YAW is the moment of command or action, so corner on YAW is pushing.

 

The castor sweep is a hard one to visualize since this action is so dynamic Geometrically.... Nevertheless i will try to explain.

 

Imagine the wheels to be our limbs and corner entry is like someone taking a stroke when swimming, as the left arm drops the right leg kicks. At the chassis the castor increases "inner" decreases "outer" and lifts the outer wheel diagonally. So basically the inner corner lowers the sprung area of the car during transfer.

 

If the castor value is statically to low then the sweep is denied belaying the fluid motion during transition and since the car is in a moment of transition "corner on" then the diagonal lift is a direct reply to tyre saturation.

 

Look at this image>

 

 

A perfect example of a rigid chassis and a low castor sweep.

[Pedro]

This is precious stuff, you´re a star

 

Let me see if I got it. On a corner to the right, the front left wheel is the most important, so having a higher castor will make this wheel's footprint move outwards away from the car during that corner, basicly increasing that wheel´s leverage over the forces it needs to handle (same grip by the tyre itself but more leverage on the car's body), also it´ll help the opposite rear wheel, the rear right wheel, retain some weight on it.

 

I might have gotten it completely wrong, I´ll have to digest it over the afternoon and give it some more thought tonight.

 

 

 

In any case, my castor angle isn´t adjustable, I know I´ve now got more than when stock, but can´t increase it now.

 

 

 

 

What I am having some trouble visualizing is this:

"Point to remember during weight transfer "corner entry" the steering action transports the front camber values so you cannot have aggressive front static camber positions since they are affected so violently dynamically"

 

What are you saying that the steering action does with the camber angle? And what is the problem with agressive camber angle? Sorry, this is really very interesting but I still lack a lot of understanding.

[Tony]

It's a difficult topic to explain because there are many valid questions to ask and answer......

 

Lets talk about this cars situation.

 

 

The position is pretty much perfect but as a Geometric snapshot we have problems, so let's analyze the image.

 

The car is FWD, corner in during Yaw with the rear inner wheel off the ground "why?"

 

I know this car is fully track spec so totally rigid between the un-sprung and sprung chassis. It's position in Yaw is inertia, lift off oversteer, weight transfer and the polar centre is to far forward "why?"

 

The castor!! .... Historically a sweep of three or four degrees will lower the sprung parallelogram sufficiently so that the diagonal shear lowers the inner wheel to the ground (remember the stroke and kick example)

 

With a correctly advanced castor sweep the attitude of the un-sprung chassis would combat the polar centre/inertia ratio during transfer and still allow lift off oversteer as an advantage.

 

Additionally from this image what can you see from the front camber positions?..... Clue

 

1: They are even

2: They are uneven

3: They are bias

[Pedro]

I get it, so you´re talking about increasing castor as a means of putting the front wheels a little further from the center of gravity, lowering the pitch axle and reducing the amount of change in weight coming off the rear wheels as the weight shifts to the front? Or did I miss it completely?

 

 

The front wheels in that picture, well, the outer wheel seems to be with positive camber when pushed that far, maybe when under maximum load it should stay perfectly vertical, or at least pushing the tyre surface flat on the ground instead of twisting the inner edge of the tyre off the ground, if there is no tyre flex it should have 0º camber at that moment the picture was taken if that is maximum load?

When stopped, camber is almost non existing?

[Tony]

I get it, so you´re talking about increasing castor as a means of putting the front wheels a little further from the center of gravity, lowering the pitch axle and reducing the amount of change in weight coming off the rear wheels as the weight shifts to the front? Or did I miss it completely?

 

 

The front wheels in that picture, well, the outer wheel seems to be with positive camber when pushed that far, maybe when under maximum load it should stay perfectly vertical, or at least pushing the tyre surface flat on the ground instead of twisting the inner edge of the tyre off the ground, if there is no tyre flex it should have 0º camber at that moment the picture was taken if that is maximum load?

When stopped, camber is almost non existing?

 

Kind of... the COG is not an issue here, more the polar center and the dynamic consequences.... As the image suggests the car is in transition, you are correct that the camber transfer is opposing and dynamically correct at this moment of Yaw.

 

The transient initiator of the castor sweep is the Ackerman angle, then by default the camber disparity (negative-to-positive) during Yaw, this is complicated since the transient inertia adds front outer camber gains dynamically so actual static analysis is near to impossible.

 

All reads desperate but it's not.... Geometric laws are consistent, the only variant is the destination

[Pedro]

I have a few doubts about weight distribution.

 

Before I changed a thing on the car it went for an MOT, a quick glance at the brake testing area screen told me the weight for each axle, 712kg on the front and 621Kg on the rear, that´s with little more than half a fuel tank, and a driver weighting about 10kg less than me. Since then I´ve replaced front seats for racing buckets, fitted a lightweight cage, moved the battery to the boot, and will remove sound stuff and rear seats before taking it to be cornerweighted, haven´t got a clue how the overall weight is now, but I'm more worried about balance than overall lightness, I could loose a good few pounds myself

 

The guy that is doing me a favour of letting me use his garage and his skill as a favour to my brother is a racing mechanic, they race all kinds of categories, from rally to classic barchettas, he probably knows a bit about this. I will probably give him some input of how I think the car drives now and let him do his thing, but would still like to get some concepts right.

 

We´re going to adjust ride height on each wheel to try and balance things out on the scales, will try and not lower it too much from a "stock" look though, as I don´t want this car to stand out at all from other road users.

 

Right now, if I had to take a guess I´ld say there´s too much weight on the rear, simply because when the mass is stable and weight transfer is over, once midcorner, if I push it then, the front wheels feel like they´re not as "loaded" as the rear ones, they tend to scrub the ground while the rear ones have a lot of traction and feel very planted.

 

My doubts are this:

 

With a front bias weight distribution will the front be more or less planted going into corners when comparing to 50/50, will the car tend to oversteer or underteer on big roundabout with slowly increasing speed and no harsh accelerating or braking inputs?

 

On the other hand, if I have a weight distribution weight with a rear bias, wouldn't the car tend to easily go into oversteer once lifting off? I'd have thought that a bigger mass on the rear would cause a lot of rear momentum when lifting off midcorner as well as providing more traction when under load from the engine?

 

Sould I aim for a 50/50 weight distribution, if possible, and disregard considerations of how the car feels now or should I try and

 

I know a lot of other variables affect this, but lets only consider weight distribution changes from ride height adjustments for now.

 

Can´t thank you enough for your help Toni, really learning new stuff here, if you decide to visit the Nurburgring this August, or northern Portugal, I´d like to buy you a few beers.

[Tony]

A little pause for thought.... At the weight displayed you have

NSF@ 356kg

OSF@ 356kg

NSR@ 310kg

OSR@ 310kg

 

At an average track and wheel base this equates to a DI 0f 1.78..... not good

 

Have a read here> http://www.wheels-inmotion.co.uk/forum/ind...p?showtopic=794 then we can talk further.

[Pedro]

I had never heard of this Dynamic Index until I read those threads, seems it quantifies the responsiveness or handling of a car by considerating weight distribution, placement, and the actual chassis size, clever stuff, as most of what I find here is.

 

However, appart from being able to say my car as a certain DI, and therefore being able to look into a clear number that tells me if it´s good or not, thus having a clear means of comparing different cars, I can't seem to be able to find a use for this index when tuning a car´s chassis, not if I´m not willing to alter wheel positioning and basic axle length or wheel distance. Wouldn´t it be much simpler if, for limited tuning purposes, instead of discussing DI one discussed simple weight distribution, as including this in a more complex equation will only dilute it´s importance amongst other factors that we are unable to change?

 

What I think I got, though, is that a 50/50 weight distribution, in theory, makes for a more controlable and rewarding responsiveness, so maybe that´s the direction I should point it, regardless of the current distribution I should tweak the front and rear ride heights, within reason, to direct weight distribution into a 50/50 direction, without consideration of the current perceived handling deffects, and then assess the "damage"? My problem is that I´m having this done as a favour and don´t want to abuse my welcome by spending a few afternoons doing several adjustments after a trial and error work process, limited I know, but it almost always is. What I was asking about is if I should try and give it a front or rear bias to counteract my handling problems, instead of heading towards 50/50.

Obviously teaking ride height to affect weight placement has it´s obvious limits and affects suspension work in a way other than weight balance, thus maybe the side effects of displacing weight this way might even do more harm than good, no?

 

 

Another question, in theory is it better to have a slight ride height difference between two wheels of the same axle and an equal weight load on both, or the same exact ride height but a different weight load?

[Sagitar]

Another question, in theory is it better to have a slight ride height difference between two wheels of the same axle and an equal weight load on both, or the same exact ride height but a different weight load?

 

The question implies that there are only two variables, whereas there are at least four.

 

If we use a simple spring balance as an analogy, the distance between the point at which the balance is located at its top end and the point at which the weight is attached at the bottom of the balance, is (within the elastic limits of the spring) a function of whatever weight we hang on the balance i.e. there is a pair of dependent variables that are 1) weight applied and 2) distance between the fixing points. This system is precisely analogous to corner load and ride height on a vehicle with a sprung chassis.

 

But, if we change the spring rate in the balance we see a different distance between the fixing points for the same weight. Further, if we change the physical size of the top and bottom hooks on the balance, we again change the distance between the fixing points though all else remains the same.

 

Likewise, we can alter ride height by changing spring rates or pre-loading or by altering the size of the components that locate the spring between the sprung and unsprung parts of the chassis.

 

So, a simple trade-off between ride-height and corner-weight is only an issue in circumstances where neither the spring rate nor the geometry of the fixing components is variable. All the changes have their consequences and the static consequences i.e. what happens to the chassis geometry as a result, are less important than the dynamic consequences i.e. what happens to the behaviour of the chassis system as a result.

 

System dynamics are very complex. They may lend themselves to analysis using mathematical modelling and powerful computers, but much of what is done in the "enthusiast" field tends to depend upon trial and error and the development of "expertise" by individuals with lots of experience. A range of "rules" have evolved, but often they are based upon observations of performance rather than direct measurement. It's noticeable that in Formula 1, drivers who test vehicles are valued for their ability to describe chassis performance in a way that enables effective adjustment of chassis geometry etc.

 

It is sometimes very difficult to differentiate between the "beliefs" of experienced enthusiasts and the laws of physical dynamics as derived from basic physics. If something is repeated often enough, people begin to believe that it is an inviolable "law" rather than a practical guide that helps in most circumstances.

 

Universal truths are very hard to find. Circumstances alter cases. Stability and agility are, by and large, opposed concepts. It seems so obvious when put that way, but it tends to be forgotten when we are searching for a set-up that will allow the vehicle to continue in a straight line without driver intervention, but still give easy and quick responses in desired manoeuvres.

 

I am aware that I am banging on, but I have followed this thread with some interest and it has demonstrated many of the problems that we have when trying to communicate clearly about sophisticated technical issues where science and popular belief (by enthusiasts) get intermingled.

 

Jargon can get in the way. Some people e.g. the medical profession, use it because it enables a more precise description than everyday English. Others, e.g. lawyers and accountants, use it to protect their privileged position and in an attempt to stop laymen from understanding too much. Enthusiasts get into the habit of using it because it's useful shorthand, but on a site like this, one of its dangers is that it excludes the newcomer and creates opportunity for misunderstanding.

 

My plea therefore is for less jargon, more simple English and at least an explanation or a reference when we use a word that has a special meaning in context and is not intended to convey the normal dictionary meaning. In particular, I find acronyms a bind. Let me give one simple example that has cropped up in this thread - ARB. We all knew that it meant anti-roll bar (didn't we?), but my first call on that particular acronym, as an ex aircraft industry man, is Air Registration Board. It's commonly used an an abbreviation for air-bag and my acronym directory shows about 150 different uses. Is it really so much trouble to type anti-roll bar (ARB) when we use it for the first time?

 

A good friend of mine used to say - examine your English for COIK (clear only if known). If someone can understand your meaning only if they already know what you are trying to say then the meaning is not clear.

 

By now, I may well be standing in a glass-house and throwing stones, so I will stop . . . . . :rolleyes_anim:

[Pedro]

Likewise, we can alter ride height by changing spring rates or pre-loading or by altering the size of the components that locate the spring between the sprung and unsprung parts of the chassis.

 

So, a simple trade-off between ride-height and corner-weight is only an issue in circumstances where neither the spring rate nor the geometry of the fixing components is variable. All the changes have their consequences and the static consequences i.e. what happens to the chassis geometry as a result, are less important than the dynamic consequences i.e. what happens to the behaviour of the chassis system as a result.

 

In my case, I´m not going to invest any more in suspension parts until at least next year or until I think I´ve reached the limit of my current parts, which I´ve just started to try, so when I ask about the height versus weight preferences, I am obviously willing to go for a compromise.

 

About the technical talk, I can sometimes struggle a bit and have to read the same post a few times to make sense of it, even then sometimes I am unable to until a second or third explanation, mainly because I can compute a new concept surrounded by a sentence I understand the meaning of, but a sentence with 4 or 5 new concepts is pretty slow reading for me, sometimes not because of the most demanding technical terms, but also the simple concepts that I never dealt with when thinking in Portuguese. It´s a learning process, and I find this good reading because it´s challenging and I´m actually learning.

 

Would love a simple reply with clear instructions of how that get me the perfect setup for my driving, but this is the second best thing :rolleyes_anim:

[Tony]

"Sagitar"........ Superb post :rolleyes_anim: