panhard bar suspension FRC 1

[bhart]

I am writing a program to output the SS wheel loads for my racecar. The car has front and rear solid axles w/ panhard bars. The vehicle is a left turn only racecar and therefore the CG is offset. What I have determined, and what I would like to confirm, is that the FRC (force RC) for a panhard bar suspension is at the height where the line of action of the panhard bar intersects the vertical line dropped down from the CG. The way that I determined this is to assume that the panhard bar at a given end of the car supports all of the lateral force at that end of the car, then when the line of action crosses the vertical dropped from the CG, the vertical component of the panhard bar force (F_y*tan(theta_panhard)) will be pointing directly toward the CG. Therefore at this point, only the lateral force creates a moment about the CG, i.e. a body rolling moment, and so this would be the point at which you could exert all of the lateral force transmitted by that end of the car (in opposition to the actual lateral force), and it would not cause body roll (no net moment).
Does my FRC locate the point described in the definition below?
From SAE J670e, 9.4.28 "Roll Center - The point in the transverse vertical plane through any pair of wheel centers at which lateral forces may be applied to the sprung mass without producing suspension roll."
I'm assuming that this is the definition that most of you are operating off of these days. Another point I would like to get clarified, does "suspension roll" mean the same thing as body roll in this case.
I'm pretty sure that my results do locate the equivalent point of application of the lateral force on the body, but I'm not so sure, because of the wording, that this is what the SAE definition describes.
Of course I have done my own research on this particular problem, but the trouble I usually have is that all of the RC construction techniques that I have found seem to have a lot of implicit assumptions that are not clearly stated, plus they usually go on to imply that the KRC and the FRC are the same when the definitions are applied to modeling, I know this is not true except for very special circumstances.
Now that I can see that there are many people (some are apparently professionals) who share the same opinion I have on the KRC (kinematic RC) --its a very limited tool-- I was wondering if I could get some EXPERT advice on my particular approach.
Thank you in advance,
bhart

 

[GregLocock]

Sorry about the delay, I needed to check some stuff out.

The procedure you describe appears very similar to the way I'd get the KRC, and it ignores the weight transfer due to the springs (shocks) and sta bar.

I'll give the quick and nasty description of how fbrch is calculated, but I need to do some more work to understand it.

Mt=total roll moment on vehicle from contact patch loads during a steady state turn

Mrs=roll moment on sprung mass (ms) by springs etc

Ay=latacc

then fbrch=(Mt-Ms)/(ms*Ay)

However I'm still finding this roll centre height stuff confusing so probably don't count as an EXPERT.


Cheers

Greg Locock

 

[bhart]

Greg,
I see your point very clearly, and your method of calculation makes a lot of sense. A few things still confuse me though:
#1. How do you define the ms*Ay term for one end of the vehicle, assuming that you are trying to find the FRC at one end of the vehicle.
#2. How can I possibly know the SS contact patch loads w/out already knowing the FRC height? In my case, I'm trying to predict the contact patch loads using an FRC approach, so it is not valid for me to start with contact patch loads and back out the FRC.
Thanks,
bhart

 

[GregLocock]

#1 This method is based on using a full vehicle model in a dynamic event. For a single axle I'd use the proportion of the sprung mass for that end of the car.

#2 Do you need to know the RC to get the CP forces? In the simple case of two identical axles in a symmetric car then you can easily solve for the CP loads.

Off the top of my head taking moments about the ground Ay*ms*CGZ=Track*(CPFZouter-CPFZinner) MUST give the CP loads for a single axle in a steady state corner. (Tell me if I'm wrong!)

I don't think it is that much harder with different roll stiffnesses at each end. It probably needs an iterative solution, admittedly, in that case. (ie assume body roll theta and lift h, work out spring forces and arm forces, compare with Ay*ms, and go around the loop).


Cheers

Greg Locock

 

[bhart]

OK, I'll have to admit, I lied a little when I said I didn't have CP forces, because I have actually got them since I first posted the thread, using strictly force techniques... its almost as simple as you made it sound but with a lot of asymmetry (different track and offset at both axles - left turn only), I didn't actually rely on any FRC approach in particular, but now what I'm concerned w/ is getting the definition right so that I can locate the FRC.
With the method that you have suggested --(Mt-Ms)/(ms*Ay) -- (and presumably use), you locate one FRC point that characterizes the whole vehicle, is that right? And you also calculate all of your moments about a longitudinal axis passing through the CG?
Thanks,
bhart

 

[GregLocock]

As I said, I need to do more research because I haven't actually used this method, by hand. It is the default approach used by my modelling program ADAMS/Pre, but I need to run through it myself to see how it works out.

That simple equation certainly looks as though it comes up with one FRC, yet the results are displayed for each axle. So I think the program treats the car as two, linked, axles.

I'm looking around for some easily accessible discussion of this - without much luck so far.
Cheers

Greg Locock

 

[bhart]

Greg,
I looked in the ADAMS/Car .pdf help stuff and found this concerning the Car definition of RCH.
"Roll Center Location
Roll center location is the point on the body where the moment of the lateral and vertical suspension forces exerted by the suspension links on the body vanishes....
Method
ADAMS/Car applies unit vertical forces (perpendicular to the road) at the tire contact and measures the resulting contact patch displacements in the vertical and lateral direction (front view). ADAMS/Car projects lines perpendicular to the contact patch displacements for both the left and right patches. The roll center lies at the intersection of these lines.

ADAMS/Car reports errors when the motions of the left and right patches are parallel (just as it occurs in a fully trailing arm suspension). Therefore, the projected lines have no intersection. ADAMS/Car also reports an error when the motion of the left and/or right patches is very small for a unit vertical force (for example, the suspension is very stiff)."

Looks like the age old problem that I am constantly running into w/ researching this subject... a source will define the RC in terms of forces/moments, and then when it is shown how to locate the point a simple kinematics approach (instantaneous center of rotation) approach is taken. Anyway, I thought I would let you know what I found out about the ADAMS/Car method of RC calculation... which might very well be completely different than PRE's interperetation... I looked in PRE's help file stuff but could not find a section detailing the calculation procedures like Car's help has.
talk to you later,
bhart