Weight transfer,roll moment

[ivan182]

I have some problem understanding other things that i came across...

1.question about load transfer (lateral)...
In 'Tune to Win' it is written that higher roll stiffness wich arises from springs does not influence the overall lateral load transfer,but higher roll stiffness that arises from antiroll bar does affect overall load transfer....
At the same time in Milliken equation for lateral load transfer does not take in account where the roll stiffness of pair of wheel come from...
so what is right?

2.question is about roll moment....
I dont understand how higher roll moment on one end of car results in faster lateral load transfer and if roll moment is higher at front than at rear , the car will have tendency to udersteer.....(Tune to win, Fundamentals of Vehicle Dynamics)
I understood that lateral load transfer arises from : instantaneous transfer through suspension arms and from compression of springs wich couses roll moment (this lasts for a certain period of time).....
So doesnt that mean that if roll center is lower on front than at rear in dependence of CG (higher roll moment), less transfer will be accomplished instantaneous through suspension arms and more will come from compression of springs. On the other hand in rear instantaneous load trasfer trought suspension arms will be higher and rest of transfer from roll moment will be accomplished at lower roll angle than at front and the car should have tendency to oversteer . Does this have any sense?
For example above static weight transfer is 50%-50% and roll rates are same front and rear.

[exFSAE]

Originally posted by ivan182:
In 'Tune to Win' it is written that [...] roll stiffness that arises from antiroll bar does affect overall load transfer.

Go find that quote and share it here, I am skeptical that Mr. Smith did indeed write or mean that.

[ivan182]

page 38.....here is the part where he wrote that...

http://imageshack.us/photo/my-...es/810/83162354.jpg/ (http://imageshack.us/photo/my-images/810/83162354.jpg/)


but i would like if you can explan it to me,not ask more questions....

[exFSAE]

Incorrect OR poorly worded by C. Smith then. Though the direct connection DOES have the effect that a small vertical disturbance on one corner is instantly transferred across the axle, rather than a delayed and softened secondary effect through an independent suspension.


Originally posted by ivan182:
but i would like if you can explan it to me,not ask more questions....

Well I'll "explan" this - you'll likely get a better UNDERSTANDING of the topic when someone initiates a dialog and makes you THINK about it and truly own the answer, rather than have it spoon fed to you.

[Z]

Ivan,

Every book ever written has either some blatant mistakes, or more often some poorly worded sections. Carroll's wording in that section is misleading.

More relevantly, the questions you are asking have been discussed, dissected, debated, deliberated, etc., etc., etc., ad nauseam, on these pages. The reason you are not getting many responses is that it seems you haven't made much effort to check the many threads that cover this subject. Use "Find" and key words roll, centre, center, moment, transfer, etc.

The last thread covering this subject in detail is barely a month old.

Z

[ivan182]

I don't expect direct and whole answer, but I would appreciate any kind of guidance than.

I tryed to find topics on forum,but didnt find something that helped me understand this...

[Z]

2.question is about roll moment....
I dont understand how higher roll moment on one end of car results in faster lateral load transfer and if roll moment is higher at front than at rear , the car will have tendency to udersteer.....(Tune to win, Fundamentals of Vehicle Dynamics)
Ivan,

I think the problem here is a lack of any CLEAR DEFINITION of what is meant by "roll moment". This is not just you, almost everyone these days fails to CLEARLY DEFINE what they are talking about. http://fsae.com/groupee_common/emoticons/icon_smile.gif

IMO, at any instant the car's body can be considered to have one "force screw" (aka "wrench") acting on it as a result of forces acting from the road to the four wheelprints. In other words, we only ever have to consider one linear force, and one couple, acting from the road to the car body.

I think you are referring to "moments" calculated from either the spring forces, or the control-arm forces, about some unspecified point, and at only one end of the car. That is, you are referring to some sub-components of the total "couple" acting on the car, but I am not sure which sub-components.

So I suggest that you first go to page 3 (Open FSAE Discussion) and read the thread "Roll Center Migration", mainly pages 2 and 3. I had a lot to say there about lateral load transfer. Some key things to keep in mind while you read that thread;

1. Total Lateral Load Transfer (from inner wheels, to outer wheels) depends only on lateral acceleration, CG height, and track (to first approximation). BUT(!!!), this TLLT can be resisted entirely by the front wheels, or entirely by the rear wheels, or in some other ratio (called the "Lateral Load Transfer Distribution" LLTD).

2. Roll angle of the front of the car is (usually) the same as roll angle of the rear. (Flexible chassis cars can be made to work, but that is another, long, long, story...)

3. The end of the car that carries the greater proportion of TLLT is usually the end that "slides" first. So if front carries most of TLLT, then understeer.

4. While reading above thread, think about a Formula V. As set-up in US and Oz, these usually have only one rear "heave" spring, so NO rear spring resistance to roll. The KRC at the front is near ground level (horizontal n-lines), and at the rear is at the centre of the differential (say ~0.3m high, for quite steep n-lines). So all LLT at front is through the spring-dampers, and all LLT at rear is through the control arms (or n-lines).

If you have more questions, please just one question at a time, and as clearly defined as possible.

Z

[ivan182]

Thank you for advice! I red this thread that you have suggessted and it help me.
It was logical that higher roll center means more load transfer via "fast" mechanical route and less by a "slow" suspension spring/anti-roll bar/damper route, but in some books (written by people involved in racing) i red opposite is true (Tune to win,Fundamentals of Vehicle Dynamics) so it was hard to ignore it....i just had to be sure.

I have a question about influence of antiroll bar on weight transfer,is it significant or not and how can I calculete it?

and...thing that i wolud like you tu explain me..
''Thus for a typical vehicle with higher rear roll force centre than front, the rear tyres load up faster during turn-in. This helps reduce the phasing between yaw and lateral acceleration and is generally A Good Thing. ''
I dont quite understand the concept of ''phasing between yaw and lateral acceleration''...faster load transfer on rear generally means oversteer witch is unstable condition,why is it a good thing? (may be my english is not that good so i disnt translate this phasing thing quite right...)

[Z]

Ivan,

It is a shame that although vehicle dynamics is a relatively simple subject (just classical mechanics, not quantum mechanics), there are very few books that explain it clearly and correctly. The books "written by people involved in racing" tend to be the worst. I generally keep all the technical books I have ever bought. The only such books I have ever thrown away are "suspension tuning" books written by people in racing (all those appear in the "Book List" thread http://fsae.com/groupee_common/emoticons/icon_rolleyes.gif).

You ask "I have a question about influence of antiroll bar on weight transfer,is it significant or not and how can I calculete it?"

ARBs have two main roles.
1. Reduce body roll (and hence adverse camber change), while allowing heave (or "bounce") stiffness to remain soft.
2. Adjust handling balance by increasing the %TLLT at one end of the car.

Generally speaking (ie. to keep it short), for FSAE I would suggest starting with no ARBs. If, after much testing, you have too much oversteer which you can't cure by any other means, then a quick fix is to fit a front ARB. If too much understeer, then fit a rear ARB. So leave space for them, but only fit one if you have to. Note that ARB's have significant disadvantages, though not too bad in FSAE. Also, a simple "U-bar" at floor level is best design (easiest, low CG, good load paths, etc.).

You "calculate it" from basic mechanics (again, don't trust the formulae in the books).


"Thus for a typical vehicle with higher rear roll force centre than front, the rear tyres load up faster during turn-in. This helps reduce the phasing between yaw and lateral acceleration and is generally A Good Thing."
I dont quite understand the concept of "phasing between yaw and lateral acceleration"...faster load transfer on rear generally means oversteer witch is unstable condition,why is it a good thing?...
That quote was via Ben, originally from Damian Harty, so I can't be sure what he actually meant. Here is my interpretation.

Consider a car traveling straight at 50m/s (100+mph). The driver "instantly" turns the front wheels 10deg. It takes about one revolution of the wheel (2m, 1/25sec, 40ms) for the tread and sidewalls to deform to the shape required to carry the lateral force at 10deg slip-angle (say Fy max). Meanwhile, the rear wheels are just rolling along.

Once the front wheels have developed side force they can start yawing the car. Some time later (many more milliseconds) the rear wheels will be angled into the corner. Only then do they go through the above "tyre relaxation" process (weird term?) and start developing their side force.

So the "purely lateral" acceleration (ie. from both F&R wheels) necessarily lags behind, or is "out of phase with" the "yawing" acceleration. I think Harty's reasoning is that by giving the outer rear wheel more of the faster geometric (n-line) LLT it can catch up to the slower elastic LLT going onto the outer front wheel from body roll. I guess good drivers traveling at 100+mph can appreciate the N (?) millisecond difference.

Note, also, that a similar (perhaps stronger) effect comes from the car's mass distribution, mainly from the yaw dynamic index kz^2/ab, or "dumb-bell/wheelbase ratio = Ca" that I have discussed elsewhere. With these = 1, the behaviour is as described above. Simplistically, the car can be considered as two half masses, each directly above their pair of wheels. The only way the front wheels can change the slip-angle of the rears is by moving the nose a significant distance sideways.

However, with a "cannonball" car (ie. with Ca<<1), as soon as the front wheels start pushing the nose into the corner, the car pivots about its centre (in the ideal case) and pushes the rear outwards, thus starting the rear slip-angle process. On the other hand, with Ca>>1 (very long dumb-bell >> Wb) the opposite happens and the rear wheels first have to push the rear of the car outwards. Then, only some time later, do they start to push into the corner, giving a long phase lag between yaw and lateral accelerations.

Z

[DougMilliken]

Originally posted by Z:
... I guess good drivers traveling at 100+mph can appreciate the N (?) millisecond difference.

What Z said. A couple of additional comments--
+ There are some pictures & plots of transients in "Chassis Design", (Olley Notes), starting on page 232. Might be helpful if you have trouble following the written description.
+ Since tire response (relaxation length) is roughly independent of speed, the tire response *time* gets longer at lower speeds (ie, FSAE dynamic events).

[ivan182]

Thank you very much for answer Z!