View Full Version : Getting a Handle On It
With some prompting from Roger I am starting this thread to have a spot to talk about handeling concepts and how we might be able to apply them.There will hopefully be some ideas we can take over to the editing side. I would also like to say thanks to KenK for all the work he has put in to the forum. Also for all the time he has spent figuring out the physics and writing up his guide. With out the guide I would still be floundering around as if I am not floundering around enough.
So to start off, being from the U.S. ( there goes the validity) and noticing that most forum members are from Europe we tend to have two forms of racing.In Europe you rally and road race. In the U.S. we go around in circles a lot, TV would have you believe that it is all on tarmac but most is on dirt. Both have distinct ways to set up cars. RT is a challenge to me because it is road courses on dirt and tarmac.
So here goes with the first topic.
I would think nothing of using four different shock set ups and four different spring rates to effect weight transfer, control wheel hop and effect bodyroll. and on dirt I am looking to roll weight onto the outside rear tire to get both side bite and forward bite off a turn. Thats right we use different terms to describe the same thing. so how would it be phrased across the pond? I am not trying to be a smart ass or making a here there compitition just throwing things out to get a disscussion going and get on common ground.
Hope this sounds interesting
CDK
Interesting for me! Firstly, as for credibility the U.S. has produce Carroll Smith :trophy: , for those of you who know who he is no explanation required.........for those who don't find out and read his books. I don;t think any other person alive (erm.. he is sadly not anymore :cry: ) has written more common sense about running and racing cars.
It was also U.S. racers (Andretti for one, & JV for another) who brought assymetric setups to F1, AFIK it had hardly been tried before Andretti and when JV wanted to use them I understand it was still a bit of an oddity even in the 1990's.
More to follow.
Pero_Grozni
12 March 04, 11:58
First off i should say that i never read anything on toppic of suspension setups for racecars. So i dont know a lot on this toppic. What i do know a lot of is mechanics. And from there is where i draw all my conclusions(i just think it through using physics spectacles).
So here to start it off- the absolutly biggest inflouance on a cars performance have the tires-no quiestion about it. They are the only part that are connected to the road(except in some WEIRD situation-upside down...). And tehy generate all of the forces that are put through the suspension sistems to the body. We have an interaction of body, suspension, tires and surface here. And in all of them we have some phenomenons. All transfer the forces "through" them.
So to get to the suspension- it is the link of the wheel-rims to the body conections. With that link you can control how the wheel rimm is positioned/orientated relative to the global coordinate sistem and the body. So all that the suspension needs to do is to ensure the optimal position of the wheel rimm to ensure that the tire contact patch with the surface is in optimal conditions(the correct size and under the corect deformation). Also in some cases you deliberatly can alter that orientation/position to ensure that there is enough heat generating to get the tire mixture to the corect temperature-so that it bonds better with the surface. It might be out of the optimal geometry setting but looking at it globaly it gives higher overall grip. I dont want to go very much more in detail here as there are a lot more influances present-we can discuss them later on.
So why are the suspensions as they are-well the tire without the spring/damper would bounce up and down over bumps... And with that the time of the tire being in contact would be shorter. And the average force on the contact patch during that time would be the same(energy laws). What would happen is that the force would grip a lot from time to time and almost nothing or nothing during the remaining time. With that the direction of the car the driver wants it to go could not be ensured. So to keep the tires in contact all of the time a M-K-B sistem needs to be added. M-K-B(mass, spring force constant(or function), damping constant(or function)). With that we can controll the frequency the rim moves relativly to the global coord sistem and teh body. There are two masses to be considered here(or even more) if we are looking at a single suspension strut. So all can theoretically be calculated and you can set the suspension to follow the desired function(shape of the ground) perfectlly. But there is a catch. As the suspension parameters are more or less fixed you can adjust it to follow one or a few extra basic functions(shapes of the ground) corectly. Usually that is a sin function, but modeificated are also possible. And for that to work as it should the shape of the ground must be consistent more or less. And as we all now that is never the case. So any suspension setting is always a compromise to follow the surface geometry the road has;BTW: I would like to comment more on weigth transfer later-just that i dont get to much confused now;
And therefor all the different suspension settings are necesary for different tracks. The better the surface of the track is the easyer it is to find the optimal settings. The more diverse it is the harder it gets. So for rally cars it is very hard to find the absolute optimal settings. And that is also why they dont use asymetrical settings(if some do than they reall know the road or the car properties are not simmetrical). The thing with the asimetrical settings is that for a driver that is dreadfull as turning left you get one responce and turning right you get different responce(again i left out a few thing connected with the tire temperature-but that are details). So for oval racing they are :up: . Also for todays track racing thay are also :up: . As for the nurby-well i doubt that they are an advantage there as the road is very diverse, althought most of the turning is done right-hence one side gets exposed more.
So if you think about rallying there are almost no consistent directions(always doing clockwise revolutions for example). So the advantages of the asymatric settings dissapear(again if the car has symetric-geometry/weigth distribution...).
Enough for now(need to cool down my brain :P ). More to come-details and weight transfer+chasis, suspension, rim, tire, bearing, surface... deformation
BTW: This is my oppinon curently so not necceserally 100% or even 95% correct. Will make edits if i figure i said something stupid.
Cheers PG :P
Phew, lots to think about........I'm still learning how to tie my own shoe-laces! :P
I had never thought of the suspension movment as a sin function, I suppose it must be. Am I right to assume that not only the shape of the ground needs to be constant, but also the vehicle velocity for anyone sin function?
I had never even thought about the possibility of rally cars being assymetric. On to the Nurburgring, in GPL I notice that some setups are assysmetric, could it be that despite the constant left/right nature of the place that a well chosen assymetric setup may give you the edge in a few key places (perhaps Galgenkopf is one - before you hit the long straight, and Bergwerk - both right handers). Although I would imagine that in real life a sysmetrical and stable setup would give a driver confidence to attack the place over the majority of the lap.
At some point in this discussion I would suspect that it will be important to define words like "bumpy". Is a road bumpy because it has lots of dips and crests, or is it bumpy because the surface is rough?
Pero_Grozni
12 March 04, 13:21
Yes the suspension movement is a sin function. The thing is that it is not a pure sin function, but has also stuff added. The dampers and road geometry get added here. So it is a variaton of the sin function. The variations can be very strangely shaped as the sin function is very"modable". Best is if we dont go too much in detail here as that would lead in total confusion-even i get confused thinking of it.
Well the important thing about it that when we have a simple m-k-b sistem we have three basic options for the function reaction on one input. One is oscilating and eventually leveling out, the middle is one oscilation and leveling out and the third is no-socilating, but just slowely leveling out. All three cases can be folowed by a variation of the sin function. The easyest way to test that out is go down to you car and push it down at one end. Usualy it will come back and level out. The oscilation of it is dependant on the damping and spring coeff. Usually when yuor shocks are broken the car oscilates like hell. And if you have good shocks it dosent oscilate at all. That are exactly the same cases i described above.
BTW: Sin funciton is through time! The responce when driving and looked on the distance covered is different. As i said all can be calculated for nice surfaces, but for undefined surfaces it is imposible(if the input on a spring is not known). So still best set ups are made by test/trial method and experiance.
I would go so far to say that be carefull when using any software for calculating setups. All those programs are just a set of equations for a model. And that model is not like real life. It is only a aproksimation of reallife proceses and oyu will get perfect setups for the model, but not optimal for real life. More stuff happens in Real life(more inflouances...)
As for driving the suspension responce is difficoult to calculate as the inputs are undefined. So mostly you can calculate a frequyency for the suspension sistem you would like it to follow. That is the frequency of the inputs(this is a lot more complicated as i am telling it, so take it with a bit of reserve). And lets say that you have a very uncompatible sets of bumps. Than in one frequency of inputs you could get your suspension into resonance=dreadfull handling-lots less time in contact. Therefore usually the suspension sistems are set up that way that they are not easy/possible to get into resonance. The problem is that that kind of suspensions are not so fast with the responce and therefore dont folow the road.
I will define "bumpy road" now.
I would say that this is a road that has the geometry changes bigger than 5mm in the RT y direction(i would suggest that al the directions we use are as in RT). So it is hard to quantify why a roda is bumpy(it could be the extra loose bits like stones as well as solid surface changes-the surface is never solid, but deformable so -carefull with that one).
So when we will declare something lets say that:
-flat is a theoretical surface-with no inputs on suspension
-smooth is a surface with really fine surfae(smaller y changes than 5mm).
-bumpy-from that size on-lets just use less/more bumpy
Why i choose 5 mm. Well i think that an average tire would "kill" such an input-damp it. So we would get no input on the suspension. I could be way off with the number, but for our talks it will do.
As you might know the tire is also a set of m-k-b sistems(at least 3 or more-depends how far we would go). And that is why i am talking of the rims. All that you fix with the suspension is the movement of the rim-if you change the geometry it could well be that you get a different responce out of the tires inputs onto the rims.
More to come
Cheers PG :P
Wow!! Covered a lot of ground there. I had to work late and its 5:30 am so I am going to take a little nap come back read and digest all of this before I reply.Thanks guys for picking up on the thread. PG from a quick read thru you covered it all now how do we take a stage a car and set it up for that? ( open question)
CDK
Pero_Grozni
13 March 04, 14:48
HAHA good quiestion. Here is where most of that knowledge fails and trial/error/win method is required. Hey if they cant do it in F1 without testing how can we. So testing will deffinitly always be number 1. Than datarecording and analysing the stuff mesured.
I think that for RT settings we should not make perfect settings. We should try and recreate the setts the original cars used.
As for real life-there is no better sensor/computer than a good driver. If he has also a bit of knowledge about mechanics(physics) than it is a winning combination. All the really great drivers were excelent testiung drivers as well(best example Walter Rohrl). Look at his today's job-test driver for porsche. And it is a known fact that GT series porsches are the best set up cars overall.
And such drivers are also able to adapt to any strange settings/car characteristics and get the best out of it(Walter Rohrls win in SanRemo 1985)-but they also know what is to be changed to make it faster. The problem is that suspension settings can be perfect, but the car will still be slow if the base concept is bad. (i dont really want to mention it, but the current WRC lancer and the past WRC lancer were/are messed up the moment they decided what bodyshell to use).
BUT as we are here disscusing how to use physics tools to get "quicker" cars-i will try my best and try to make some sence.-next post; now i have some testing/data analysing for RT to do.
Cheers PG :P
Just got up and am on my way back to work(busy weekend) when I get home tonight you will get some thought out comments back. Excelent description of drivers abilities I agree completely. One of the methods used when someone has a good car is to measure it up and make a copy but it never seems to come out the same even with jigs and CNC and laser stuff. (clarification) a lot of stuff here is tubular frames I will send some pics.
CDK
Pero_Grozni
13 March 04, 21:08
Just a quick reply to why it dosent come out the same.
Engineering is a precise science-or so it is thought.
Well being a freshman 3 years back on the UNI(mechanical engineering-construction nonlinear mechanics) i thought everything can be calculated and all is made like we want it to. Guess what by now i know that those facts couldnt be more wrong. Every year i know more and by the same time i know less. The tendency in science is that the more detailed you look at a certain problem the more complicated it gets-and the more stuff become unknown to you.
I must say that designing machines...(we say construction here so i will just say constructioun proces-thinking, evolving ideas, calculating, prototyping and so on) is acctually arelativly rough and undetermined proces. You never know what will really happen-yeah you know the basics, butr never the compleate detail. 10% mistake in calculations compared to reality is fenomenally good.
So to get back to why you never get the same machine with teh same plan. Firstly the undetermination in materials. Here a huge percantage of the differance will occur. Materials although very much deeply covered in some cases will always be undetermined(again the properties can be mesured/guarenteed to a certain level-the better the level-the better the probability of getting very simmilar responce will be). Next up is the manufacturing-not only geometry here also heat input is very important here(especially polimer materials). Than assembly-again a huge number of inflouances(i will just put out the temperature of the parts when assembling-jou can have a tourq wrench and tighten the sqrew as you did before. Well if the temperature is not the same than extension under temperature wont be the same). Ok than the conditions when in use...
See even with most advanced lasers, CNC machines, CAM, CAD software you always get different products(OK on some scale they appear the same ,but on a smaller scale they are very much different.
That is also why there are safety factors in standards.
One thing i should mention as well is the incorectness of teh equations used. Every equation is false and dosent compleatyl cove the reality. it depends on the quality of that equation(physics rule) of just how much of a safety factor you use.
Just wanted to say this to make you all think about it a bit more when using software, equations...Always question your results! That is my advice.
Setup discussion a bit later-still analysing data.
Cheers PG :P
Just got home four hours later then I expected and the sun will soon be up again. I start a thread get good responces and the real data posts go wild and I dont have enough time to read or reply to any of it. But I will not give up will get caught up and back on track soon. As a quick reply to your last post PG, the reason I am four hours late is I just spent 6 hours getting nowhere on a job I have done 50 times. No matter what I did it just would not work right this is life imatating art or in this case your post as everything you just said basicly just happened to me
CDK
Has anyone noticed this? Some current F1 cars have very odd front wishbone angles (view from the front). So, far I have only noticed this on Michelin cars BTW.
The Williams (I think Renault too) appears to have the inboard mount of the upper arm higher than outboard. "So what" I hear you ask ;)
Well, a traditional wishbone has the inner mount lower, so when the wheel moves up it assumes more negative camber - this helps keep the tyre at the right angle when cornering. This arrangement will ADD positive camber as the wheel moves. So, why would an F1 team want to reduce the static negative camber when the car rolls?
Here is my idea:
The cars might have such massive roll stiffness that they don't perceptibly roll at all, so the static camber setting is right for front end bite in a corner. BUT, maybe they have enough suspension movement when NOT in roll (PITCH) that this feature actually moves the negative cambered wheel upright under braking - thus maximising the tyre.
Another idea - with angles like that it MIGHT be that the roll centres are below ground level (again if very stiff in roll this may not be a problem). But having the roll centre so low any jacking force becomes "negative jacking" actually adding to the mechanical grip.
This might be akin to why sprint cars seem to work with very stiff roll bars. Maybe it is a case of using the bars to restrict wheel angles in roll because the suspension is better used keeping the wheels upright to maximise traction once the car is pointing where you want it to go. Of course sprint cars differ from F1 cars in that the go sideways - but I am guessing that they have an excess of power over traction and need all the drive they can get and can live with stiff roll bars in corners as long as they can get the power to the road.
Back for a moment, They are just Heavier bigger bodied sprint cars( you will know what I mean) No roll bars, Torision bar suspension with various forms of location bars and links thats both sprint cars and mods. Most is done with srings\coilovers and shocks, Traction is the biggest problem. But when they get good bite they will lift the front end right off which is pretty neat at over 180kph
CDK
Pero_Grozni
14 March 04, 10:03
Here is my idea:
The cars might have such massive roll stiffness that they don't perceptibly roll at all, so the static camber setting is right for front end bite in a corner. BUT, maybe they have enough suspension movement when NOT in roll (PITCH) that this feature actually moves the negative cambered wheel upright under braking - thus maximising the tyre.
Another idea - with angles like that it MIGHT be that the roll centres are below ground level (again if very stiff in roll this may not be a problem). But having the roll centre so low any jacking force becomes "negative jacking" actually adding to the mechanical grip.
Firstly i didnt notice that at all as i am not a F1 fan(modern F1 that is). The first idea-sounds very reasonable :up:
About the second idea-like i said i am not very fammiliar to the expresions in suspension setups. If you would tell me what EXACTLY is the rool centre of suspension and what is jacking force i can comment it(will think it through with physics googles :D ).
Also to give some thought on that-you said that you noticed on top michelin teams. It could well be that it is very dependant on the tire construction and mixture. In modern F1 the contact patch of the tire is anyway much more dependant on the tire properties than the suspension settings than in other motosports. You probably noticed that the F1 tires have quite big sidewalls compared to other motorsports. Here are my thoughts on that. The most of the suspension work in a F1 car is done with the tire. As it is underdamped they need to add damping to it. So they make the suspension struts. The springs are very stiff(downforce...). So the dampers must be very stiff and variable as well, to cope with the tires and the springs. And a thought and a reply to your second post here Roger-tracks in F1 are very smooth. So the bumps encountered are very small in the y direction. So that is why they can have such high own frequency suspension settings there. if they would still be doing Nordschleife(just thinking of that makes me scared :twisted: ) than there would be very different suspension settings on the cars. Probably a lot softer and higher as enywhere else.
What i always found interesting was the behaviour of Le-mans cars(modern ages). They always bounced a lot through corners(very fast oscilating). My theory-as i think that the downforce possible with those cars is even bigger than in F1 they need to be set up even stiffer to cope with it. And as the tires have much smaller sidewalls-are stiffer in the y direction(RT language)- the bumps cant be coverd by the tires so effectivly.
Yeah and you were absolutly right on the faster car-higher frequency of bumps inputs over the same road onto the suspension before .
And i just want to add that the own suspension frequency is dependant on both-the spring stiffnes and the damping-opend my old notes from 2 years back ;)
Cheers PG :P
Pero_Grozni
14 March 04, 10:13
They are just Heavier bigger bodied sprint cars( you will know what I mean)
What cars?(i dont know-where did i miss that)
Another thing-sprint cars-another expresion i am not quite fammiliar with. Are the touring cars ment by that? Guys here in Slovenia we only have Eurosport and DSF avalible for seeing races. So please update me a bit :bowdown:
CDK-were you working on reallife suspension settings last night?
I am very interested in the software avalible there-I have to have a look ;)
Cheers PG :P
Illustrations enclosed. As you can see the forces from the tyres act through the roll centre. And the car rolls about it's roll centre (at that end) - so the CofG would move laterally as the car rolls.
The roll couple is the lever that produces roll. If you have the roll centre and CofG at the same height you get no roll, BUT you get bad jacking forces that lift the body up (watch a VW Beetle cornering one day!).
If you had a very high roll centre and low CofG the car would tilt INTO a corner!!!!!!!! :D It would also have spazzy wheel angles and massive jacking forces.
The roll axis is the line joining the front and rear roll centres and this relative to the mass centroid axis has a big effect on the transitional responses at corner entry, mid & exit.
If you had car car with the CofG of the masses at the front of the car very low, but those at the rear very high (tall mid mounted engine), if the roll axis did not produce a similar roll couple between the masses at each end you would get very odd roll developing and wheels being loaded/unloaded very strangely.
Also, look at those pics (sorry about the quality) long enough and you will realise that the instantaneous pivot and roll centre are moving targets, so if the designer is not careful the roll centre moves up/down/sideways giving the car very unpredictable behaviour.
You may notice that I drew the line linking the outer pivots to the ground - the difference between this and the centreline of the tyre is the offset. If the two are the same very little force (apart from the tyres own self centring force) will be felt. When the offset is inboard like this a disturbing force on the tyre will create a pull strongly to the outboard direction. many modern cars actually have the offset outboard of the contact patch, this makes for self stabilising steering (good for Joe average if he gets a puncture), but also makes packaging the suspension harder without extreme "king pin" inclination angles (which will produce camber gain when turning the wheel).
The poor suspension designer also has to worry about the tyre scrubbing sideways on the road as it move up and down - just to complicate things further. :OMG:
What cars?(i dont know-where did i miss that)
Another thing-sprint cars-another expresion i am not quite fammiliar with. Are the touring cars ment by that? Cheers PG :P
I think that by Sprint cars CDK is referring to short dirt ovals and the short, nasty, powerful exciting cars that tear around them.
Pero_Grozni
14 March 04, 11:21
Thanks-very interesting :up: . Will report back and comment on it. ;)
John_Schoen
14 March 04, 11:21
Like this. :)
Pero_Grozni
14 March 04, 11:25
Looks like a BIG fun machine :up: .
I have something simmilar myself made out of an beetle, but i didnt use it since a while now(destroyed so many engines that i rather stoped+it was always underpowered wit the beetle engine). Will upload pictures of it soon ;) .
Cheers PG :P
In a earlier post I spoke about a new term being used called the moment center I believe it is similar to the CofG. But moves as the suspennsion travels. I am just learning about this so my facts are a little off right now.
Interesting point F1 and Indy cars have tremendus downforce built into them it is like they are sucked down on the road surface this certently changes a lot of how you design stuff.
Tires make so much difference it is not even funny.
I know I am throwing out different types of cars to talk about, but looking at Sprint and Dirt Mod suspensions is a interesting comparisson as they are both basic and complex.If I can figure out how to post some pics I will.
Pero that little sprint car in the pic is also 800+ HP and wieghs about 600kg you know you are not just going to the store for milk when you push down on the gas.
CDK
Pero_Grozni
14 March 04, 22:08
Moment centre-i will put my thoughts on this one. COfG of the body is fixed. CofG of the car is moving a bit with suspension movement(but really small changes). As for the Moment centre-yeh it could happen that it is in the car's COfG. But only when driving straight. It is very much effected by the forces on the car(with a car i mean everything-body, suspension, tires). So usually on a car a lot of forces are input(will talk of it as a rigid body). So in the COfG there is the the gravitationaal force. To keep it in balance four forces are put onto the tires(i neglected downforce...)Than there are the drive forces on the wheels. The counterpart force to that one is the inertial force in the COfG in the opposite direction to the drive forces. That one is dependant on the acceleration and car mass. Than if cornering you have side forces on the tires(the angle is 90 degrees to z axis-RT talk. It can be but it almost never is as the tires are deformable and the resulting forces direction became different. The resulting inertia force again is in the COfG and opposite direction. It again depends on the acceleration(side acceleration in this case). So as we covered the force part of it now we can go to the Moments part(tourqe if you like it). Tomake it more clear on a rigid body there are 6 possible movement options:
3 translations
3 rotations
And in every such movement inertial forces are present. Now you will say-what is with PG :confused: . The side forces are as if looking in rotational movement around the corner radious. - It is not so simple. That is true for a mass point, but not true for a riggid body. It is a extra case when that accurs, but that is if car has no couples of tourqes on the wheels. Ok that is a bit complicated to understand without a scetch, so i will explain in detail.
Looking in the car you dont see any radiuses... So how to calculate everything? Moments accure always where there are couples of forces present. And in a car that is always. So we now have 3 moments that rotate around x, y and z direction. The moments result from the forces onto the tires and the inertial forces in the COfG. So with every moment we get an inertial moment to counterpart it. If they are in balance no rotation accurs. If one is bigger there we have rotation. And that rotation is not around the COfG point(can be, but it almost never is). The rotation centre is than located in some other point that is dependable on the forces input onto the car. And around that point inertial moments appear. So inertial moments are very much changable with the position of the rotation point(that is why i disscused that what i did in the MOI disscusion a bit back). And that rotational point is probably what you call the Moment centre.
Ok i explained it simply here and on a case of a rigid body. The car can be aproksimated by a set of rigid bodys. And all those bodys have again 6 freedoms of movement. Some of those freedoms can be fixed so the calculation gets simpler. The funny thing i should mention that with the suspension the rotation centre's coordinate system is not always pointing in the direction of the coordinate system we use in the COfG. It is very much movable and influanced with the body movement compared to the tires(that is the suspension movement defined). So the suspension rooling line is in the direction the rotational Z axis of the coord. sistem is showing.
That is why we declare MOI in RT. And we kinda declare some middle MOI. With that MOI the inertia moments are calculated. And as the rotational centre is movable they are also changable. From that fact i drew teh conclusion that the most realistic behaviour we will get in RT is for the midengined cars as the mass dosent stick out on one part. So i think that the stratos can be farely closely " nailed" . As far as the front engined cars or wery rearengined cars go they will never behave as they really should. OK we can make them preety close, but the realism to the stratos will be of a "better quality"
So the rearengined cars will never really swing on you like they do and teh frontengined(RWD) cars will always "Protest" to turn.
Shourly most of you have allready noticed that kind of behaviour.
Just to get back to the physics part of this post-to calculate car behaviour without inclouding deformations we can allready get good results with minimum of 16 freedoms of movement. If we add some more than also the camber can be simulated. So i will comment on the schetces you drawn Roger a bit more later. All i can say now is that the roll centre you drawn is very dependable on the movement of the suspension and that is not only in the y direction it si also in the X direction. That what you drawn we normally call teh momentary rooling centre in mechanisems. It is theoretical and for that moment you drawn it in you can calculate the conditions on the body as if you have only rotation around the rolling centre. I guess it was used to make explanation of suspension forces more easily, but i rather do it as globaly as i just described.
To just tell how much freedoms of movement you should use to consider calculating all the suspension movements-i would say 52 or more if you would like to moddel the simplest deformations of body-suspension as well. That is a lot :twisted:
Cheers PG :P
Pero_Grozni
15 March 04, 03:18
Just thinking a bit when looking at the picture of the little monster(800hp+ WOW :up: :twisted: ) .
A nice wing :up: . What i am interested in is if it is determined by the rules how it is facing? As i imagine that those cars go sideways for most of the course(oval- right?) it would be much more efficient if facing into the direction of the "wind". The way it is positioned now the fluid currents behind it are probably very turbulent-higher drag and less downforce.
And the shape that is followed is at an angle to the shape if driven in a straight line. So if it is detrmined that it is supposed to be faceing that way perhaps some different crossections of the wing would help adding the downforce.
I wered a bit off toppic here, but i thought it was worth to mention it.
Cheers PG :P
The moment center is a new term to describe I think everything you just said. I am still trying to learn about it as I have only read a couple of articals and most times I have to see it live for the whole concept to kick in.
Very good about the wing it is a set size and the bascic location is set but there are both cockpit and external adjustments it is to help get thru the cornors and aplys its force that way more so then on the straights the cars are driven much straighter then you would think. you see the cars push down in the turns from the force of the wings.
Heres a new thought often a car that is to tight (undesteers) feels like it is loose (oversteers) to the driver. How can you find out and why did it happen. Anything to do with setups and driving are fine for the thread in fact I have been showing and mentioning what would be considered simple cars in some parts but getting them to work is an art and everything your talking about comes into play.
CDK
Pero_Grozni
15 March 04, 12:36
I will try to comment a bit on the tigth car being more loose for the driver.
The first think that comes to my mind is that the drivers have to drive it the way to provoke it much more to turn into the corner. And with provoking it gets more loose as they would find it when set up loose.
But this is no clearly defined argument so i will try some sence with a physics explanation.
It is connected with the tires and the weigth transfer. If you load the tires they break away at simmilar conditions(for the wide and the narrower tire). So as you have weigth transfer to the back the rear tire is very much closer to braking away than the front(which has lots of reserve for a loose car). So to maximise the overall grip(through a corner not when turning in) the car gets set up more towards tight so that you use also most of the front tire reserve before braking away. And now you have both tires very loaded and closer to their limmit(before they break away-will use these terms for now not to overcomplicate it). So if one end decides to brake away you have much less reserve to relay on with helping you recover the break away(also during a drift as a drift can have very stacionary conditions-if they get nonstacionary than i will call it braking away). So it gets harder for the driver when over the limmit and that is why they have comments that it is to loose.
Hope this makes at least a bit sence to you guys as it is very "thin" as everything that has to do with human perception of stuff.
EDIT: Just went through it and edited something (changed loose to tight) as it was wrong ;)
Cheers PG :P
PG, right on target and with a good tech explanation behind it. It is very hard for the driver to spot this because there is very little sense of the front sliding and then all of a sudden it loops on you. you think its loose(less letters to type then oversteer) so you adjust the car even more the wrong way and on and on. The way to check this is to free(loosen) the car up a lot and see what happens.With rwd you can control the back with the gas so you can work your way back set up wise a lot easier if you start out with a car with oversteer. What would you say is one of the easier items to work with in RT to loosen or tighten the cars up?
CDK
Pero_Grozni
15 March 04, 20:06
I would say COG Z position for a starter. But as that is fixed for a car than the diff settings come to my mind. Again they are not connected to the suspension, but are a huge influance and help when setting the car up to take hairpins and sharp turns. I must say that as far as the suspension goes i dont really know. Perhaps make the back end stiff-to get a loose car and the opposite for a tight car. Damper settings could also be the key-less damping in front -looser car. I must say that i didnt experiment too much with loose and tight car behaviour yet as i was working on the tires till a week back. And i would also say that the camber could be a huge inflouance, but as i didnt test it out much i dont really know how RT reacts to it. Will think it through a lot more and come up with a better post on it(has been a busy day today so my head is not too clear).
Cheers PG :P
Since you know I just had a busy weekend and am just starting to read everything over so I can understand it I will cut you some slack. However I must note that even on a busy day for you you still manage to come up with a lot of ideas and info(compliment). In the question I posed and as I read the posts both here and in editing there is a lot of emphisis on inital frame \suspension set up. In some ways with the intro of the dirt cars and some of the questions I am asking I am trying to lead you to what do we do with what we have during the pit stops during the race. We can edit what we are going to do with the sliders during the tune part between stages. With the Beta Healey I have the setups about as good as they are going to get well still retaining origional parameters but I have set my tuning abilites to help me adjust the car for different road surfaces. what might be some of the things you would work with during your pit stop?
If that was confusing I will try to explain better in the next post.
CDK
What would you say is one of the easier items to work with in RT to loosen or tighten the cars up?
CDK
I would say that once you have a basic suspension setup, then the ARB is a good tuning tool. If you place too much emphasis on the springs you might upset the way the car puts to power down.
I have found that the strongest influence is (naturally) the CofG location - although I do think that it is essential to calculate this to be as close as possible to the real car in order to get it's essential character. MOI is important to feel (although not balance).
Then you need to get the tyres right, just to get it drivable (P_G can help with some tables for TireTeam tires).
The diff settings make a big difference power on and power off.
For basic neutral cornering balance the ARB (although how much neutral cornering do we have in RT!)
I have noticed that the camber curves have quite a significant effect (but again these strictly should try to emulate the real car).
Interesting question you raised, ably answered by P_G. My other Alp' (GTA Turbo) is often described as an oversteerer. However, when I changed it's settings (springs/dampers/wheel swidths) I moved it all to settings that should in theory cause more oversteer, but got better handling. The reason I chose this is not far (I think) from what you are describing. I noticed that with the original setup, the car would initially understeer - then almost instantly snap into oversteer, so when pressing on if I felt understeer I was already beginning to apply opposite lock (that could catch out in-experienced drivers I ams sure!).
I added more front tyre and stiffened up the rear springs, now the understeer is far less and when oversteer happens (which it will with a big V6 hanging over the back axle!) it is easier to control. I also seem to have picked up better traction (might just be better quality dampers), which helps control oversteer under power.
Very interesting answer, I agree, I think once the car is initally built to spec then the key tuning aid becomes the ARB. In fact I have set up the Healey so that when you do tune suspension during stages the only change you make is to the ARB. This posts question --- which bar or bars am I adjusting and in which direction to get what result?
CDK
Pero_Grozni
16 March 04, 23:04
what might be some of the things you would work with during your pit stop?
CDK
Interesting-in what relation do you mean. If having a predefined addoncar that i shouldnt change i would go for the stiff suspension if it is not to rough on the stage. Usually with a stiff suspension in RT(suspension slider fully right) you get more downforce as it is made that way so that it is directly dependant on speed and momentary ridehight(we can allready calculate just how much downforce is produced at what speed with around 15% mistake). That and the ridehight calculaiton has been mine and Rogers work in the past week-and we have it all worked out. Everything will be inclouded in the next physics calculator version along with the TT tires tables.
The second relation that comes to mind si if i would make my own car and set three basic settings options. Curently i am thinking that i would not set to different ridehights so that i wouldnt be getting to much DFC inflouances on the settings. Than i would make the middle settings the way that tehy would behave friendly-not to tight and not to loose. And for the soft i would make it loose and for the stiff i would make it tight so that in both cases i would be able to follow the desired line and in the case of the stiff settings maximise the tire loading front and back. It is highly dependant on the drive(rwd, 4wd, fwd) concept though.
One thing to keep our heads thinking constantly is that in RT we should try and make simmilar settings that the original(real cars we are simulating) cars had. So if it was very loose in real life it is only apropriate that it is loose in RT as well.
What you raised is an interesting toppic though-how much faster cars can we make with only changing the extra parameters(like small changes to springs multipliers...) without tuching the basic set up.
That kind of testing is interesting and would help to make cars be as corectly fast compared to another car(to simulate reallife again)
Hope i made some sence :P
EDIT: ARB's tommorow
Cheers PG :P
With the Idea that a tuning session between stages would only alow you minimal time to work on things you would have to pick things that could be turned or swapped with out to much trouble, that would adjust your handeling for the style of the next course.My first choice was the ARB.By editing the car ini Ihave been able to pick what part of the suspension I am going to adjust ,At this point it is only the ARB using the suspension slider, so did I include both front and rear bar or just one do I change there rate equally or ratio it.and what might be some of the effects by going harder or softer with the bars. Remember I only think of two parameters does it understeer or oversteer. One of my favorite racers said take your car around the track keep going faster and faster till one end of the car breaks loose then fix it and then keep repeteing that exercise till you have a very fast car.
CDK
Edit When we build a car from paper up to metal we can calculate and design in different parameters that will basicly set the cars handeling. when we take a standard model and try to get it to perform we must work with what it came with hence the Idea of what can we do with what we have were the variable that changes is the surface we are on. with that in mind what can we do for example going fom dirt to tarmac or to a snow covered road with simple changes like springs or shocks or arbars.
Pero_Grozni
19 March 04, 12:54
Today-as promised ARB's and weigth transfer. :sherlock:
This toppic is not so easy and clear to explain as the others i described till now so it will be highly discutable and i ask you to quiestion my thoughts here.
Ok I will start with the basics of "weight transfer".
Here no weight or simmilar stuff gets transfered at all. Why we have more force on the suspension than when turning or braking or accelerating than?
It has to do with the equations for the balance of a riggid body.
I did write them down before, but now i wil do it again and in 2D so that we will be able to think of a side view of a car.
For a rigid body to be in balance(2D) all three of the next exuations must be corect.
All forces in the Z direction have a resultant of 0
All forces in the Y direction have a resultant of 0
All tourqes around X direction have a resultant of 0.
As we want to have an accelerating car the first one is not OK and the +z forces are higher. That dosent matter to much as we dont have any restrictions in that direction. But we have restrictions in the Y direction-teh ground. And that one ensures that the car dosent move in the y direction or rotate around the X axis in the Moment point.
Now as we have all forces balanced stacionary we have a known weight distribution between the wheels(forces on the wheels). But as soon as we start accelerating we have two forces more-the drive force in the contact patch of the drive wheels and the counteracting inertial force in the COfG. As the drive force is higher the car is accelerating. But this two forces form a couple of forces and form tourqe around the x axis. But as the car cant turn around the x axis there needs to be a counteracting tourqe to satisfy the third balance equation. The counteracting tourque is produced out of the Y forces on the wheels. So to counteract the tourqe the force in the front wheels must be smaller than in static conditions and the force on the back wheels is higher than the force on them in static conditions. Along with the gravitational force in the COfG they form anothe couple of forces-hence they coountract the tourqe produced by the driving force and the Z inertial forces.
Hope that i didnt overcomplicate this easy to understand fact-i had to describe the correct physics explanation of it.
All of the explained above is true for the both other two planes so if you look at the car from the front you can explain the " more force on the outer whells as well this way.
""""PG takes a deep breath""" :wc:
Now for the complicated part of it. Now we have suspension as well. What suspension does is to slow the delivery of the full weight transfer force on the ground. So the force gets added/taken away formt the tire slower. That gives the tire more time to adapt to the change in the force and the deformation of the tire is slower -wich gives you more grip during the transition. The transition is when the body is in movement and the tourqe part of the balance equation is not satisfyed for the body part of the car(it is for the car though) and the suspension part of the car. So when you accelerate the forces dont get transmited with their finnal values right away, but they wait till the suspension is compressed so that the body isnt rotating no more. So now the more the suspension has to move the slower is the transition-the more grip reserve the tires have during that transition. The transition can be slowed with the dampers, but the forces remain the same as in the nondamped state, beacause dampers add a
damping force when moved.
Now this is all fine and dandy, but does the suspension get compressed with the same speed with a soft spring or a hard spring. As both of the springs are made out of the same sort of material( the basic rigidity coeffient is very much the same for all sorts of steel). So out of the material side view the springs are not different in their stiffnes if they have the same shape. So the softer springs are shaped differently than the hard springs.
As the movement of small disturbances through the material is the same for both cases the differances in speed some out of the shape of the springs.
Also to go out of the energy part of the view-energy transformed or work can be calculated with the next equation" W= x*F"
W-work
x-movement
F-force(dependant on x)
So out of this equation it can be seen that the energy used for deforming a soft spring is much higher than the energy used to deform the hard spring-hence the soft spring needs more time to get compresed.
Now with that explained you get the basic idea of why the soft springs give better grip in the transition faze(when rallying that is a vast majority of the driving time).
So next to the ARB's.
First what do they do?
They deform when we have a differance of length of the left and right suspensions. With that deformation they counteract the suspension movement and efectivly we have something like a higher stiffnes coefficient on the outside suspension(outside part of a corner) and a smaller stiffnes coefficient on the inner part of the suspension. So loking it form the view of the explanation above the inner wheel has more momentary grip and the outer wheel less momentary grip, which is soemtimes good especially in ralying when you can have different grip from one side of the road to the other side and overall it is a higher overall grip through the stage-faster times. the other beniffit of the roll bar is that you can have softer springs to help forward traction, and dont have so much roll of the body-better feel of the steering for a driver.
So again the setup is very highly dependant on driving style and road conditions.
So as far as my answer for RT goes-you only change one ARB with the slider. You made the basic setup so that it was more or less neutral and to make it more loose you reduce the front rollbar stifnes and to make it tighter you incresse the front rollbar stiffnes.
Or you change both, but in a nice ballance between the two.
""""PG wipes his forehead and takes a breather""" :banana:
Ok that is enough for now. This is a very highly discutable post("thin")
so dont just take it for granted. And oppinions welcome.
Cheers PG :P
I liked all of the explanations but is it possiable that it is backwards. Every thing you said makes sense but it seems that the exact opposit is what happens. Very good call on my setup but I am adjusting the front bar again opposit from what you predicted. I am very impressed that you figured it all out. You did miss one equation the gas sloshing in the tank on weight transfer HaHa. Spring manufacturing is another whole topic as there are a lot of variables that go into making a spring , they sure look simple dont they.
CDK
Phew P_G terrific discourse.
ARB's - Is it not the case that, in fact, they don't "soften" the inside as you describe (particularily with reference to your brilliant spring explanation). Because the ARB joins the wheels together across the width of the car, when both springs compress by the same length the ARB does not come into play at all (you knew that ;) ). If the car hits a diagonal bump, you get different spring lengths so the ARB gets worked - this can make a car quite sensitive to diagonal bumps (so too much ARB is a bad thing on bumpy roads). Also when in roll you get different length springs (or rather they TRY to, but are restricted by the ARB) - in this case as the outer spring shortens it the ARB tries to pick up the inner wheel, so is not making the spring softer, but is actually removing load from that wheel and transferring it to the outside wheel. The ARB is actually adding it's spring rate to the outside wheel (thus adding to the roll resistance) and will eventually lift the inside wheel off the road if it is stiff enough.
Dampers - in some extreme cases the oil can be forced through the damper so fast that it resists movment in itself - I cannot quite visualise this, but is it the case that it is then preventing the spring from doing it's job properly so it will behave as if the car is stiffly sprung (for that instant that the damper piston velocity is greater than the speed the oil can move through the damper's valves).
Good post - the bit I like most is:
"Now we have suspension as well. What suspension does is to slow the delivery of the full weight transfer force on the ground"
People often foget this - the fact is (small movements of the CofG aside) that the total weight transfer hardly changes when you change springs, it is the speed of weight transfer that changes. So, a stiff car is very responsive to inputs, but will also give it's tyre less chance of doing it's job on certain surfaces.
Pero_Grozni
22 March 04, 11:26
Phew P_G terrific discourse.
ARB's - Is it not the case that, in fact, they don't "soften" the inside as you describe (particularily with reference to your brilliant spring explanation)
:dood:
I dont think that I made a mistake-i will try and explain.
Well the outer wheel does "feel" like the wheelrate would be higher, because the inner spring gets commpresed and that commpresion force"helps" to counteract the ground force on the outer wheel.
And the inner wheel gets loaded from the movement of the outer wheel.
What that does is that the inner spring gets compressed. So if we look at it as there would be no ARB's than to achive the same effect the inner spring would be softer and the outer spring stiffer. That is where i was pointing with my previous explanation.
But i did get it wrong in some other thing-it is not the wheelrate that changes-it is the spring load. So how does the load get changed-again we can look at it as if we had a spring, except the spring movement is now the differance between the outer and inner wheel movement. So if we look at the spring as i described above we again have a fast and slow deforming spring depending on the shape of the rollbar. But we musnt forget that the resisting force of the inner spring is what is the base point (joint) to the rollbar. And the rollbar than has a combined deformation movement depending on the force-the movement of deformation of itself and the movement of deformation of the inner spring. But whatever view we look at it the logical thing is that the with a stiffer ARB, the outer wheel would be loaded faster.
Damn-this is a hard thinking toppic so i really cant be 100% sure, but what happens in RT is not necesarily what happens in reallife and also there are so many inflouances that this is more theoretical-what happens in practical use is more??? :confused:
EDIT: I also gave my theory to a RT test. In one case i used no front rollbar and it the next case i mounted a really stiff rollbar.(tested on smooth tarmac with the 037 on same slider suspension setting).
And my theory worked out-with no rollbar the car was a lot more loose in the transition phase of turning and a lot tigther with a massive rollbar mounted during the transition phase of turning. Perhaps you are having ridehight/DFC connected influances when you set the car up with the slider? If the car has a lower ridehight in RT the downforce increases, and it could well be that this is the inflouance you are getting. Or perhaps your explanation of the driver feeling whether a car is loose or tigth comes into play here?
Cheers PG :P
Well I have been turning test lap after test lap changing bar rates and I am not getting as much of a change as I expected, One good part of this is I think I hit on a better set up then I had and I am now running much better lap times. You went the right way and tested on tarmac as I have been running Russia 1 as I know this section the best and can feel the differences. The part about loading the outside front tire is right on, this will make the car push (understeer) sometimes the thought is to get more weight on it ,it will stick better but then your trying to use one tire instead of four and four stick the best. I think technically PG and Roger have it down and testing is showing the therories are holding up. I not sure yet how the numbers we are inputing into the bar rates correspond to size or rates other then the obvious bigger number bigger bar. A small increase in the bar rate should yield big results but I am not getting this yet.Back to the garage for more testing,will report back
CDK
Real life or RT? :D In RT, I noticed that softening the front ARB gives more steering bite. I have not used it myself for ages, but my test-track has a skidpan and constant radius corners for this purpose.
The problem in RL & RT is finding that mid corner phase, so that corner entry/exit weight transfer are not confusing things.
In RL, the use of more front ARB and picking up more grip may in fact be two things? Could one be that the extra ARB is sharpening up corner entry by supporting the diagonal weight transfer better as you turn in. In which case there might be an argument for stiffer springs/bump damping? Or (circuit racing mainly I assume), could it be that the stiffer ARB is actually loading the tyre more, so is putting more heat into it - if that tyre was not working at the correct temperature before then it would appear that more front ARB is increasing front grip against all traditional beliefs? These examples could pehaps show that setting up a car just ain't that simple - there are so many factors in play (we have not even taked about brakes & diffs yet :P ).
Yes, P_G in RT the suspension slider will affect stiffness and ride height if the car builder has included alternative settings in the car.ini file (I am sure most will have done).
Probably for CDK to answer - bu any other takers?
How the heck does this happen?
I assume car is moving right-left across the frame (anti-clockwise on track). It appears to have opposite lock on, but the body roll!???!!!! :OMG:
loungekid
14 April 04, 17:19
I assume car is moving right-left across the frame (anti-clockwise on track). It appears to have opposite lock on, but the body roll!???!!!!
My guess is it has a lot to do with the crazy amounts of camber change that the wheels are exhibiting at full lock, massive positive camber on the outside, and massive negative on the inside.... The whole car leans as the front suspension flattens out like that...?
Also, a car that spends all of its time turning in only one direction is likely to have some amount of side-to-side disparity built in to its setup, i would guess....
-chris
I have been photo hunting and when static the car's don't seem to exhibit large amounts of tilt. Maybe there is some mechanical leverage when the suspension loads and/or when engine torque is applied to the axle.
Pero_Grozni
14 April 04, 18:44
No mates-you overlooked one important thing-the spoiler and the spoiler flaps(the big square things ;) ) Those ones produce so much drag that they tilt the car so much-i have noticed that(less evident) on the first pictures John uploaded. And obviously it is OK that way as the big square things help the laptimes-otherwise they wouldnt be mounted on the car anyway.
How they help-my theory:
1: naturally more grip to the inner wheels
2: Bigger drift angles possible and less opposite lock necesary.
3: Easier to catch slides and sustain them.
Of course for that to work the speeds need to be high, but at average 120-140mph i cant see any obstacles.
BTW: Roger-i think that it could be simulated ;)
Cheers PG :P
I have read that the assymetric wing end-plates do indeed contribute to car control. The drivers are able to "lean" on them in corners. It is a monster amount of energy if the wing plates are indeed making the car lean like that - but look at the bracing struts - if the plates were doing nothing they would not need the huge amount of support they are being given.
vBulletin® v3.8.7, Copyright ©2000-2013, vBulletin Solutions, Inc.