Side View Questions

[scott_rfr]

First, if anyone could point me in the direction of any literature or papers on anti/dive/squat etc and pitch centers that would be great.

Next I have a question on pitch centers. I understand that the height of the pitch center will affect the amount of load transfered longitudinally either by the springs or linages. However how is the location fore/aft interacting with the system. I came across the thread where Pat Clarke echoed Ron's suggestion of having the pitch center two wheel bases behind the car. The reasoning I can see for this is to keep the stability of the center from rapid movements under pitching motions. Is there more to this that I am missing?

Lastly, referring to anti dive/squat. These quantities will not affect the total weight transfer however they will effect the distribution of this transfer. So if I change my front or rear stiffness in bump this should have an effect on the distribution of the transfer correct? So would my elastic wt have a term of something like Kor/Kor+Kof where Kor,of= stiffness of front/rear axle in bump?

If you read through all of this thank you.

Scott

[scott_rfr]

First, if anyone could point me in the direction of any literature or papers on anti/dive/squat etc and pitch centers that would be great.

Next I have a question on pitch centers. I understand that the height of the pitch center will affect the amount of load transfered longitudinally either by the springs or linages. However how is the location fore/aft interacting with the system. I came across the thread where Pat Clarke echoed Ron's suggestion of having the pitch center two wheel bases behind the car. The reasoning I can see for this is to keep the stability of the center from rapid movements under pitching motions. Is there more to this that I am missing?

Lastly, referring to anti dive/squat. These quantities will not affect the total weight transfer however they will effect the distribution of this transfer. So if I change my front or rear stiffness in bump this should have an effect on the distribution of the transfer correct? So would my elastic wt have a term of something like Kor/Kor+Kof where Kor,of= stiffness of front/rear axle in bump?

If you read through all of this thank you.

Scott

[PatClarke]

Ahead, Scott ...ahead of the car =]
Pat

[VinceL]

@ Pat
Ahead? In that thread you talked about wanting it behind. http://fsae.com/eve/forums/a/tpc/f/125607348/m/4366064394/p/1

Is the reasoning for ahead to reduce unfavorable camber change under braking? In that case it would worsen camber change for the rear under acceleration. Is the reasoning for behind to make the undertray (where applicable) of the car work better in turn entry?

Or maybe it has to do with how your kinematic roll centers move up and down on turn in and turn exit?

Is any of this close?

[PatClarke]

Aren't I permitted to change my mind ? =] =]

Seriously, the front spring rate should be stiffer than the rear ?? Therefore when a vertical load is applied at the c/g the car should squat more at the rear, putting the instant centre somewhere ahead of the car.

Don't know what I was thinking way back when.

Cheers
Pat

[murpia]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by scott_rfr:
Lastly, referring to anti dive/squat. These quantities will not affect the total weight transfer however they will effect the distribution of this transfer. So if I change my front or rear stiffness in bump this should have an effect on the distribution of the transfer correct? </div></BLOCKQUOTE>

Scott, the only effect on the distribution of longitudinal weight transfer are the (very minor) changes in CG height / position and the wheelbase when the vehicle pitches. You can only exchange sprung and geometric transfer with antis, the axle total at the contact patches is always the same.

There is a line of analysis that says make your rear axle natural frequency &gt; than the front (with a typical rear-heavy car that means rear wheel rate &gt;&gt; front wheel rate). Then when you hit an axle bump disturbance the pitching effect rapidly becomes a heave effect, which is supposed to be less disturbing to the vehicle as a whole. Gillespie Chapter 5 contains one treatment of this.

Where this puts the pitch centre, I'm not sure...

But, for a typical rear-heavy car, roll distribution usually pushes you towards front wheel rate &gt; rear wheel rate (modified of course by anti-roll bars). So in the end ride may not be the best priority to concentrate upon.

Regards, Ian

[exFSAE]

I really wish I knew more about longitudinal weight transfer and pitch center/axis effects.

I don't even see what having the pitch axis well ahead of the car would do. Is that to say when the car pitches its pitching about that axis well ahead of the car? In which case the whole chassis would jack as well?

If your A-arms at static are all parallel with each other and the ground, in the side view, how do you even calculate the pitch center? Or is it not defined/at infinity?

[scott_rfr]

exFSAE I feel the same way as there is not much in print about pitch centers etc, at least that I can find.

Pat why ahead of the car? The way I view it is that your instant center longitudal and vertical distances define your anti-properties so what exactly does your pitch center contribute to? The way I always thought was that your left and right pitch centers defined you pitch axis which is what the car pitches about. So wouldnt having the pitch center in front make the car pitch about some axis in front of the car creating large jacking forces like exFSAE stated? I would imagine in this realm having the pitch center somewhere near the rear axle would be more benefical?

Also why does the instant center of the rear axle need to point towards the front when the car squats? Couldnt you angle the plane of one of your rear arms up towards the rear to create your anti squat thus putting your instant center behind the car?
Scott

[Ian Scott]

RT,s calcs are more to do with frequencies than the angle of A arms....Ian Scott.

[PatClarke]

Thanks Ian (Welcome aboard and how are you keeping BTW?)

Yes, RT and I were talking about frequencies and their effect on pitch and squat rather than geometric pitch control aka anti-dive and anti-squat.

Personally, I don't like geometric antis, as they really are binding mechanisms.

Pat

[murpia]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by PatClarke:
Personally, I don't like geometric antis, as they really are binding mechanisms.
Pat </div></BLOCKQUOTE>
Hi Pat,

I'm doing fine, thanks.

There is plenty of anecdotal advice that using antis (anti-dive seems to be singled out more often that not) can make your suspension 'harsh' or increase joint friction. I assume that's what you mean by binding mechanisms? No doubt poor quality sphericals or rod ends have more friction than one would like under load. Even the best have quite a bit. This is where flexures offer a distinct advantage.

The thing is, if you calculate the inboard joint forces for a suspension with and without antis, then the antis just redistribute the forces slightly. Much can be done to influence joint forces (and hence weight and friction) with suspension link geometry, so if you undertake that exercise for a geometry with some antis, you should be able to eliminate any 'binding' effects.

Also, I've seen few anecdotes criticising a high rear roll centre for binding up the suspension...

There is usually some truth in anecdotes, and this may have to do with inertia effects when accelerating the unsprung mass through the typical heave path of an anti-dive front suspension. I'd like to see an SAE paper or textbook treatment of that. Or maybe it's just that in the days gone by many cars had rocker arm front suspension and too much suspension joint friction...

Regards, Ian

[exFSAE]

I still ask... what happens when you put pitch (or roll) centers outside the chassis.

When the pitch axis is say well ahead of the front axle, does that mean under braking that the whole chassis will lift and all camber will migrate to positive? Conversely with the pitch axis well rear of the rear axle would, under braking, all camber gain be negative (with the chassis squatting).

The idea of having rear ride frequencies higher than front, as I understand, is a ride quality concern.. not necessarily handling.

I'm still not convinced either from whoever has been saying it, that you generally match roll stiffness inversely to weight distro. In fact, I'd expect the opposite. For most consumer cars where you're going to be front heavy.. even the sportier ones where distro is going to be between 55F/45R or 50F/50R I'd expect roll stiffness distro between say 60F/40R and 55F/45R ( assuming same tire all around ). Stiffness distro further forward to promote understeer, decreasing as you bring it back towards the weight split, and then getting into twitchy nasty oversteer handling as you go inverse.

But that's just what I think. And I believe Claude's whole "magic number 5% than the static front weight percentage as a baseline" thing goes along with that.

[murpia]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by exFSAE:
I still ask... what happens when you put pitch (or roll) centers outside the chassis.

When the pitch axis is say well ahead of the front axle, does that mean under braking that the whole chassis will lift and all camber will migrate to positive? Conversely with the pitch axis well rear of the rear axle would, under braking, all camber gain be negative (with the chassis squatting). </div></BLOCKQUOTE>

In my view, expecting the car to really rotate around a single geometric pitch centre is optimistic. Read all the posts about roll 'centres' and you'll get the idea. In the longitudinal case, the brake balance will usually dictate unequal contact patch forces front and rear (as well as the weight transfer effects).

One you get into ride frequency responses, the key point about locating a 'pitch centre' outside the wheelbase is to avoid a 'porpoising' effect. Pitch-heave mode coupling is a real problem on stiffly sprung, aerodynamically ride-height sensitive cars. Even with no aero issues to worry about, any vertical oscillation hurts grip.

Regards, Ian

[exFSAE]

Well.. the chassis for damn sure rotates about SOME point, I guarantee you that. I'm well aware kinematic centers are an approximation at best.. but still.. its better than not havin a damn clue.

And my question is still unanswered http://fsae.com/groupee_common/emoticons/icon_smile.gif

[flavorPacket]

no, it rotates about an axis.

examine relative wheel recession front and rear, add weight transfer, add braking and driving torques, tire rolling radius changes, and you'll be able to find what you need.

But, if you ask me, I'd say just make the arms parallel, build it, and tune the damn thing. And when Claude asks you where your pitch center is, tell him, "over home plate."

[murpia]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by exFSAE:
Well.. the chassis for damn sure rotates about SOME point, I guarantee you that. I'm well aware kinematic centers are an approximation at best.. but still.. its better than not havin a damn clue.

And my question is still unanswered http://fsae.com/groupee_common/emoticons/icon_smile.gif </div></BLOCKQUOTE>
I'm sorry if it seems like we're evading answering the question...

If we go back to your original post you ask about two independent concepts that both could be termed as 'pitch centre'. The first is the intersection in side view of an axle's suspension linkage planes. The relationship between that point (usually termed an 'instant centre'), the CG position and the contact patch or wheel centre (depending on inboard / outboard brakes etc.) is well explained in a number of sources (e.g. RCVD section 17.3, Gillespie Chapter 7) and anyway you seem to understand it correctly as it forms part of the common definition of a percentage anti-effect.

Perhaps the key point here is that no source I'm aware of takes the step of combining the front and rear side view instant centres into a single point, the way most sources treat the equivalent front view analysis of 'roll centre'. That suggests to me that a single side view point defined in that way has no utility, and that one can't assume the vehicle will rotate around it under braking or acceleration...

The other definition of 'pitch centre' is a dynamic one, as analysed in Gillespie Chapter 5. In fact there are two 'motion centres' one of which, if certain conditions are satisfied, could be termed the 'pitch centre'. The position of these motion centres has nothing to do with side view geometry, but instead relates to the natural frequencies of the axles and the pitch moment of inertia of the vehicle. Therefore a vehicle won't rotate about this 'pitch centre' either, under braking or acceleration.

That still leaves you with the problem of calculating what happens to your car under braking or acceleration... Well, one typical approach is quasi-static analysis, where you model your vehicle and use d'Alembert assumptions of a force applied at the CG. You can include non-linear features such as tyres or suspension motion ratios if you want, and usually your solver will have to iterate to converge on an equilibrium solution for a given CG force. Forces from inertias and dampers will therefore be neglected.

Note that such an analysis is equally effective for pure lateral behaviour. In fact if you did the longitudinal analysis first you'd realise that calculating the classic kinematic 'roll centre' is unecessary using this method.

If you can't live without damper and inertia effects, you'd better use a true multibody simulation like ADAMS or SIMPACK, in which case you can easily handle combined lateral and longitudinal forces.

I hope that all helps...

Regards, Ian

[scott_rfr]

So in the end I should go read Gillespie Chapter 5 on dynamic pitch centers to get a better handle on that.

Geometric pitch centers are not worth noting much about. However when doing the kinematics most programs ( Im using Optk) calcualte pitching motions about this pitch axis, I assume this is a reasonable assumption for doing pure kinematics?

Lastly the anti features Pat you say you dont like antis. However how best else do you recomend controlling the motion of the chassis in the longitudal plane? You could run very stiff springs but then you loose grip, run a third spring could be heavy. Also this notion of binding to echo some more of a previous post cant you take most of this out by the linkage geomerty? I have heard that argument quite a few times.

Scott

[murpia]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by scott_rfr:
Geometric pitch centers are not worth noting much about. However when doing the kinematics most programs ( Im using Optk) calcualte pitching motions about this pitch axis, I assume this is a reasonable assumption for doing pure kinematics? </div></BLOCKQUOTE>
Can you elaborate a bit on what OptimumK is calculating for you? Kinematics in general is converned with relative motions. The relative motion of the sprung and unsprung masses will be around the 'instant centres' (there's one in side view and one in front view). Typically, those 'instant centres' move around a lot. The 'instant' position of the 'instant centre' determines the 'instant' amount of anti effect (!).

I have a feeling OptimumK isn't calculating a point the sprung mass rotates around due to longitudinal forces. However, I've not used it, so I may be wrong.

The kinematic packages I have used calculate a table of values for parameters such as anti-effects for a range of unsprung heave positions (defined as wheel centre to reference plane height). The reference plane is usually defined parallel to the bottom of the chassis, as many racecars have flat bottoms. The unsprung heave positions to axle rideheight relationship is usually assumed linear, most racecars don't pitch enough for the small angle approximation to fall down.

Is this how OptimumK works?

Regards, Ian

[scott_rfr]

Optimum K genterates a pitch center by connecting lines from tire contact patches to side view instant centers the intersection of these lines front and back is the pitch center. Optk does this for left and right sides and the left and right pitch centers are connected to form the "pitch axis". When performing pitching motions it pitchs the car around this axis. It does all the usual as far as keep recalculting postion of IC and redoing all the math all over again.

It does not calculate anti's however. At the moment there in no input for cg height, masses, spring rates etc. Hopefully in future versions. Murpia what software you use?

Scott

[murpia]

If OptimumK is not asking for CG position, spring rates, masses, brake balance, roll couple distribution, tyre forces etc. then in my opinion that 'pitch axis' is pretty meaningless.

In terms of commercially available software I've used Ansible AeroSusp. I can recommend it as it seems to be accurate, but I don't know the cost. It handles most variations of double-wishbones in terms of pull or pushrods, anti-roll bar types, 3rd spring types etc. It only handles the motion of one corner at a time, but it can calculate link forces against contact patch forces which is a handy feature.

Othewise it's all been proprietry stuff whether coded in MATLAB / C++ / VB or in the form of an extension to a CAD package. Personally, I would always correlate the results of any kinematic package with at least one other source. Usually I build an Excel sheet that can 'drive' the kinematic model and the CAD model with a little cut and paste or a macro or two. For example I have a sheet that takes suspension geometry co-ordinates and generates an AeroSusp file, links into a CATIA assembly and also internally calculates the static heave position of the vehicle and the suspension link forces.

Here's a question back: Given some track data of suspension positions, how would you define and calculate the pitch and roll axes of the vehicle?

Regards, Ian

[scott_rfr]

for roll if I knew the suspension positions I could calculate axis, dont know how I would do this for pitch.

Scott

[murpia]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by scott_rfr:
for roll if I knew the suspension positions I could calculate axis, dont know how I would do this for pitch.

Scott </div></BLOCKQUOTE>
Can you elaborate a bit? I'm interested in the formulae people might choose to attempt this, and the assumptions they might make...

Regards, Ian

[R. Alexander]

Thought I'd bring this good conversation back.

I'm working in OptimumK right now with the kinematic pitch centre.

I know it's just kinematic and doesn't look at forces, but I've been using it as an estimation to figure out the anti-effects of my suspension config.

I've heard that having the pitch centre behind the car might be a good thing, but it looks like if I put the pitch centre back there, I'd have negative antisquat and anti lift.

Would that make sense?

Ryan

[murpia]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by R. Alexander:
Thought I'd bring this good conversation back.

I'm working in OptimumK right now with the kinematic pitch centre.

I know it's just kinematic and doesn't look at forces, but I've been using it as an estimation to figure out the anti-effects of my suspension config.

I've heard that having the pitch centre behind the car might be a good thing, but it looks like if I put the pitch centre back there, I'd have negative antisquat and anti lift.

Would that make sense?

Ryan </div></BLOCKQUOTE>

You need to read this months Ortiz column in Racecar Engineering...

But anyway, don't confuse the ride motion pitch centre with the side view version of a roll centre which might also be termed pitch centre.

Regards, Ian

[The Stigg]

Hi guys,
if you are trying to make complicated relationships (multi-body-dynamics) too easy than you might go totally wrong!

I never believed in the theory of "the suspended mass rolls about a axis" and I don't believe in this theory for heave, pitch and warp either.

A simple way to disprove this theory is given by tyre & steering mechanics.
- What do you think will happen to your roll center or axis if you steer a front axle with a lot of caster, jacking up one side of the front end?
- What do you think will happen with your roll axis when your contact patches are moving lateraly?
- What do you think will happen with your roll axis when your vertical tyre stiffness is changing in every single end of the car due to different tyre pressures &lt;-&gt; temperatures?

Regards, Andy

[Pennyman]

Andy's right, kinematic analysis is worthless and we should design all our suspensions with zero caster and choose tires with infinite sidewall stiffness. http://fsae.com/groupee_common/emoticons/icon_wink.gif

Really now, there's no harm in investigating this stuff, but obviously there are additional variables out there. At the end of the day, pick a happy medium and get the car done soon enough to sort it out.

[blister]

hey andy,

So we have a kinematic roll centre, which is what too many of us think to be the centre, where the suspended mass pivots around.

Then we have the force based roll centre, which is quite often seen as an alternative/opposite to the kinematic roll centre.

What is really the relationship between those two? Doesn`t the relative distance between the FBRC and the KRC determine the split in Weight transfer between elastic and geometric weight transfer?

Also, do you include the effect of caster jacking into the FBRC calculations?

ciao, urs

[The Stigg]

Hi Guys and Urs,

What I've tried to say is that a lot of people are trying to defend their designs based on false assumptions.

I agree with pennyman that the kinematic- or FBRC is simple to understand and can! result in a good suspension design. But don't be surprised when it's not! and ask yourself why.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">What is really the relationship between those two? </div></BLOCKQUOTE>
To be honest I'm not sure if there is a possible relationship but wouldn't be unhappy if there is one:-)

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Doesn`t the relative distance between the FBRC and the KRC determine the split in Weight transfer between elastic and geometric weight transfer? </div></BLOCKQUOTE>
I don't think so! From what I've measured with the DAQ the time lag with FBRCs includes the time the tyre needs to build up the slip ange and lateral force. Geometric and elastic weight transfer tend to happen prior in time.
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Also, do you include the effect of caster jacking into the FBRC calculations? </div></BLOCKQUOTE>
Yes I do include the caster jacking effect in my FBRC and WT-calculations.

Regards Andy

[Chris Allbee]

If you really want to know that kinematic and forced based roll centers are dodgy at best, then try back calculating the "roll axis" and degree of roll about said axis using measured points on the car during either a computer simulation or from video of the car running.

If there were truly a single roll axis, then you should be able to find it fairly easily after you account for the translations of the vehicle.

Nevermind the fact that it gets very complicated very quickly once you add any pitch or yaw to the picture.

If it were really as simple as any of the "Roll Center" theories would suggest, I doubt companies would waste millions on developing mathematical models to find accurate centers and axis of rotation in the fields of robotics and bioengineering.

It truly is a complicated subject and understanding the simplified theories is fine and dandy to begin to grasp the complexities of the system, but they are not an answer unto themselves.