VWVortex.com - McPherson Suspension Geometry on our MkIV ....... shall we one more time?
 Username or Email Address Do you already have an account? No, create an account now. Yes, my password is: Forgot your password? Stay logged in

Thread: McPherson Suspension Geometry on our MkIV ....... shall we one more time?

1. Hello everybody.....
Here is an attempt to simulate (virtually) how exactly our suspension works, what happens when we go lower, what happens when we go into a curve, etc..... Let's start with something very simple. Just trying to rebuild as accurate as possible the front suspension geometry first. Once we have the right data, it will be pretty simple to simulate what exactly happens when this whole thing starts working. I guess it will be easy to see how spring rate and spring height affects the whole process, how much camber change we have at any given angle, etc...... but let's start with the very basic. Let's verify first whether our virtual model is accurate. Mike and Ian provided some of the dimensions here, but would be nice if we have few more data collected from others, as their data was slightly different. I know it is not easy to get it, but we can try. Unfortunately I do not have a stock car, so I can't help much on the "stock" basic model. ....... Have a look, please:

2. How can I help? I can take some measurements if you want, but I am not stock, as you are aware. I have camber measurements if you want them (pre camber plates) for a point of reference if you think they would be useful. Nice renderings BTW...they rock.

3. Scott, actually you can help a lot! Look at the second picture, where all the dimensions are and try to take those same measurements from your car. Then I will do the same from mine, so we will have some data to compare how much all the picture changes with (you) and without (me) camber plates. The camber gain you got is not going to be the same through the whole range, and using this model we can find out the exact values at every position.....

4. this is also a heavly +? castor setup which cause big changes during turns which can't be simulated in 2D
http://www.rctek.com/handling/....html
http://www.rctek.com/handling/....html

Modified by oldmanTDI at 9:25 PM 4-27-2004

5. Oldman, you are running to fast ..... The who point is to go slowly, so EVERYBODY understands what exactly happens, why and what. I was waiting for this one to be posted later on, when we have full pretty picture on how this works in 2D first..... but not that you brought it up, here it is the next step:

6. sorry, I was not going to bring it up as lowering the car increase castor angle, a good thing. now all the slammers will use this as ammo... ''
Nice pictures!

Modified by oldmanTDI at 10:57 PM 4-27-2004

7. I haven't had a chance to take measurements yet guys...sorry. I will be able to tomorrow though. I have some hefty papers to write, which are due tonight. Then I'm finished with school for now.

8.  Quote, originally posted by oldmanTDI » sorry, I was not going to bring it up as lowering the car increase castor angle, a good thing. now all the slammers will use this as ammo... '' Nice pictures! Modified by oldmanTDI at 10:57 PM 4-27-2004

If I remember correctly, increasing caster angle is good for high speed stability, bad for turn-in response.

9. Scott, take your time, I do not think we have to solve this in two days
So, nobody said much about the above dimensions, which makes me think you all agree! Of course, later on I will find a way to verify them as to try to have the most accurate numbers, but now let's concentrate on the concept and on the camber curve itself. Yes, it had been beaten to death and I know there are people who do not need to see pictures as to figure out this, but I am also sure there are guys who are very close to capture the situation, but small "picture" help would give the final push and the puzzle will get solved easier, so let's try this one....
In the following picture we used the simplified model from the first post and simulated "lowering" from the stock (let's call it 0") to about 7", going by steps of 1". All the different geometries are overlaid on top of each other, using some transparency, so it is easier to see the "path" that the important components follow. Pivot point 1 is basically the inner side of the lower control arm, the one that represents "the car". Then we have pivot point 2 that is the ball joint between the LCA (lower control arm) and the strut assembly (remember, we simplified the whole strut assembly and wheel in one piece as to be easier to picture it) and that pivot point is the one that we are moving by 1" every time and record the angles. Then pivot point 3 is the upper part of the shock, the one that remains constant and doe not move up and down. Part 4 is the most important for this post, basically represents the wheel. It is pretty clear now how the angle between the plate (wheel) and the vertical axis changes as the whole assembly moves up (car lowered). Yes, in this picture the point 1 and 3 remain there for the whole time, as it was easier to visualize the changes at the item 4 in this way, but in real life, the points 2 and 4 remain there (wheel always stays on the road) and the points 3 and 1 move up and down together....

As you remember, for the previous drawing, we started with camber -0,5 degree (negative). Here are the numbers we got from this simulation, lowering inch by inch...
0" = -0,5
1" = -0,835
2" = -1.036
3" = -1,097
4" = -0,997
5" = -0,684
6" = -0,114
7" = +0,775
It is very visible that the best angle is when the assembly is lowered by about 2" to 3" from stock, but CAREFUL, this is static camber at the moment! This is exactly where the LCA is perpendicular at the rest of the assembly (strut, etc), which results from the given geometry in maximum negative camber. The ultimate goal is to have this situation there (max negative camber) when the car is not static, but when in corner and load is pushing the car down (lowering it in a way) more than what you need......
This picture may help understand better why exactly lowering is not a good idea - because once you are at let's say 2" below stock, you are around the max negative camber you can get, but remember, this is while you are in straight line, where negative camber helps you little to nothing, you just do not have use of it...... Then let's imagine now you are cornering hard, with your car already lowered at 2" and you have those not-so-stiff aftermarket springs too.... what happens? You are easily at the situation shown when car is lowered by about 6 to 7". You are about to bottom, no more suspension travel because the springs are soft and because you transferred already more than 1000 lb on that outer front wheel. The assembly is fully compressed and the car is in full lean on top of that, because now even the roll is increased. So, we are not far from the assembly being as shown at the end of the 7-th inch of compression, in which situation you have already positive camber (static already + 0,775) plus the fact that the car is not static, but leans on that same wheel, so the whole assembly is rotated around the roll center, with this sending the tire in even more negative camber....... you are on the edge of your tires and at that point it does not matter how sticky they are, because you are not utilizing them as much as you could.
Reminder - this whole picture is without involving caster yet and without being sure of the actual dimensions of LCA, strut, etc.... Also without knowing how much travel the front assembly allows. I used 7", but it may be more than that, which means even worse news as the camber curve is not linear. Later on I will try to do two different scenarios, with softer lowered spring and stronger tall spring, so to see them next to each other as direct comparison may help more to visualize the situation..... ciao

10. Peter, please go though the math on why the Camber just does not go to + I always thought that it did as the arm moved up, by looking at your calculations is does not it in a sinosudial wave? dunno. Can we see the equation?

11. Lou, actually it goes to + on the 7-th inch of static lowering. If you look at the numbers, an eventual "curve" would look exactly as you said..... give me few minutes, I will log the data in an excel format and do a graph....

12. Here we go.....

Keep in mind that I am absolutely not sure our cars are starting with -0,5 degree (negative) camber! My eye may be fooling me, but the company's Golf which is stock and is on the parking lot outside looks like it has some little bit of positive camber by standing there! Again, I may be terrible wrong as "it looks like" only.... If the camber is more positive than the given -0,5 degree I use, then the situation goes south faster (earlier)..... I think what we need to do here is to build the whole picture (both sides of the suspension) and connect them like they are connected in the car. Then find the curve (path) on which the roll center moves through cornering and do a more serious simulation, so we know exactly how much weight goes on the outer tire, and at which angle of compression that occurs, so we can find "virtually" which spring rate and spring height would be the ideal compromise for the front....

13. This is a great thread. Could you add the location of the roll center to the results as well? Great job!!! [IMG]http://*****************.com/smile/emthup.gif[/IMG]

14.  Quote, originally posted by pyce » Here we go..... Keep in mind that I am absolutely not sure our cars are starting with -0,5 degree (negative) camber! My eye may be fooling me, but the company's Golf which is stock and is on the parking lot outside looks like it has some little bit of positive camber by standing there! Again, I may be terrible wrong as "it looks like" only.... If the camber is more positive than the given -0,5 degree I use, then the situation goes south faster (earlier)..... I think what we need to do here is to build the whole picture (both sides of the suspension) and connect them like they are connected in the car. Then find the curve (path) on which the roll center moves through cornering and do a more serious simulation, so we know exactly how much weight goes on the outer tire, and at which angle of compression that occurs, so we can find "virtually" which spring rate and spring height would be the ideal compromise for the front....

Peter, this is super! I was wondering how your little "project" was working out -- very well done!!
Before you do the more in-depth simulation (e.g., in 3-D, and/or with accurate location of the roll center in a corner), may I please ask you for a quick & dirty plot? It'd be nice to see the actual camber of the outside wheel even with a rough approximation of wheel travel in a corner. Basically, could you take the output of the above plot, and add to it the tilt of the car body itself. It's a rough approximation, but you could use the following equations:
Let T = track width (what is it - 58" or so?)
Let theta = vehicle roll in degrees (range: 0 to 6 degrees)
Let delZ = vertical wheel travel (deflection due to cornering)
Let Zs = amount that suspension has been lowered, relative to stock
Let delC = change in wheel camber (relative to body) due to wheel travel
Let Cs = static wheel camber
Let C = actual camber (tilt) of wheel, relative to ground
(In your plot, you've basically calculated (delC + Cs) as aa function of delZ, correct? If so, then....)
delZ = (Zs) + ((0.5) * T * sin(theta)) (I know, I know -- a very gross approximation!)
C = theta + Cs + delC, where delC is what you've calculated as a function of delZ
Would it be possible to take the above, and use them to plot C as a function of theta? (Perhaps at varying values of lowering, indicated by different-coloured curves?) If so, it'd be really nice to see how the angle of the wheel in a corner changes with lowering! (But if it's too much of a pain, please don't bother). Thanks very much, and once again this is outstanding!
- W

15. Winston, I will try to do what you requested......
Meanwhile, does anybody has his/her car in the air right now and could kindly provide the following dimension? It is essential so we go ahead later with the roll center simulation. Thanks everybody!

Also, how do we calculate, as precisely as we can, the center of gravity? Plus, how much weight do we calculate for the front of the car. I got mine at about 1840 lb on the front wheels, but in a dynamic situation we know that more (or less?) would be transfered on those two McPherson assemblies..... Thanks again.

16.  Quote, originally posted by pyce » Also, how do we calculate, as precisely as we can, the center of gravity? Plus, how much weight do we calculate for the front of the car. I got mine at about 1840 lb on the front wheels, but in a dynamic situation we know that more (or less?) would be transfered on those two McPherson assemblies..... Thanks again.

Hi Peter,
I'm sure Dick Shine will know where the CG is, but 20" above the ground would be a plausible round number to start with. If you still have access to scales, and if you can tilt the car while it's on the scales, then I can calculate the CG height for you; basically, we need the L/R weight distribution (or weight at each tire) before and after the tilting. The necessary amount of tilting might be a lot, though, so hopefully someone can just tell you the CG.
For the front weight, did you subtract out the weight of the unsprung mass? I'm sure you did -- just checking!
For dynamic weight transfer, once we know the CG location, we can calculate that for different g-forces.
This is cool -- thanks for doing all the work!
- W

17. Hi Pyce.... maybe some alignment specs from my 2003 Jetta/Golf/GTi Bentley would be of help!
Measurements are in Degrees, Minutes ' , and seconds " if I recall my notation correctly.
Front:
Total Toe (wheels not pressed): 0 Deg +/- 10' for all suspension codes
Camber: -30' +/- 30' for Base Suspension; -33' +/- 30' for Sport Suspension; and -35' +/- 30' for G70 (180 HP 1.8T) Suspension Code Sport Suspension [20AE/337 I guess]
Caster (not adjustable): +7 deg 40' + 30' for Base; +7 deg 50' + 30" for Sport; +8 deg +3-' for Sport G70
Rear:
Camber: -1 deg 27' + 20' for all suspension codes
Total Toe (with specified camber): +20' + 10' for Base; +25' + 10' for Sport; +29' + 10 for 1JC (180 HP 1.8T) Suspension Code Sport [again - 20AE/337 I thnk]
The above does not include the rear Jetta wagon specs nor did I include the specs for the "Heavy Duty" suspension (1GB, 1GW, G10, G39, & G48 front suspension codes and 1JG, 1JB, & 1JN rear suspension codes)
A small side note: Checking the Vehicle Data Label for my car results in not being able to find any suspension codes that match my Bentley. I understand that my GTI was made late in the 03 model year and my Bentley appears only to go up to mid 2003 from what I can tell. So, whatever suspension codes I have listed above are probably not complete.

18.  Quote, originally posted by Ceilidh » ..... C = theta + Cs + delC, where delC is what you've calculated as a function of delZ .....

Winston, I believe that we need to know the exact roll center at every degree of roll the body does, only then we will have super accurate total camber curve. The one I did not is very quick and very dirty and may be misleading, because I used the degrees I got earlier and added 1 for every 1 degree of body roll, All these related to the 7 different setups with lowering from 0" (stock) to hypothetical 7" lowered vehicle, which we know it can't happen. If you need a more accurate graph, we need to find the roll center for every of those situations first...... Here is what I have:

Rich, thanks for the data, I see we are on the right track! I will adjust the caster accordingly. Thanks again!

Modified by pyce at 3:51 PM 4-28-2004

19. Here is another one. Run the simulation of camber change due to body roll (theta from 0 to 7 degree, by 1 degree increments) which is the pink line. Then the old blue curve that represents suspension travel camber change. Values are from travel from 0 to 7 inches, increments by 1 inch) and then the Yellow line represents that total camber change due to the two variables acting together. It does not make much sense, as again, we do not know yet how much weight would transfer at every degree of roll and every inch of travel. I mean, the two values have to be simulated together, not separately as I am doing it now and I can not go ahead if we do not have accurate weight numbers and roll center path. Or maybe I am terribly missing something somewhere?

20.  Quote, originally posted by pyce » It does not make much sense, as again, we do not know yet how much weight would transfer at every degree of roll and every inch of travel. I mean, the two values have to be simulated together, not separately as I am doing it now and I can not go ahead if we do not have accurate weight numbers and roll center path. Or maybe I am terribly missing something somewhere?

Hi Peter,
Just got in from the lab -- I'm so tired my eyes are crossing , so I'm not much good right now for real thinking. But roughly speaking the yellow curve looks plausible - it shows the outside front tire staying upright under initial roll, and then adverse cambering as the roll increases. I'll run a quick check on a spreadsheet tomorrow (night?) and will try to send you the results; I agree, there's something a little odd about these curves, though at the moment I'm too befuddled to see it. Thanks again for doing all these plots -- they're really useful!
- W

21. This thread is awesome ... showing it to all my suspension-geek buddies.

22. Pyce-
You're the MotherFing man.
-Mike P

23. I've seen stock MK4s range from 0.0 to -.5 in negative camber.....-.25 might be a better starting point.....not sure if it will make too much of a difference.....

24. Agreed with Oldman.....I would think that the static camber will immediately begin to head towards positive with any lowering....I think something might be wrong....

25. I disagree. Even though the LCA may be angled upwards relative to the ground level (it was slightly that way on my stock Mk4), the angle between the LCA and the strut is less than 90. As that angle approaches 90, it pushes the bottom of the strut outwards, increasing negative camber. It is only when the angle has exceeded 90 degrees that the bottom of the strut is pulled back inwards, making the camber more positive.