Susp rates

LesiU

Ok guys, here's the thing:
I want to use springmult and dampermult to define susp rates and use just the pushrod angles, so that rF can use them for calculating wheel rates.


Now, from W_Bradshaw's description:
"
=0 means the wheel rates are adjusted for the geometry of the pushrod and the location of the lower end of the pushrod relative to the inner and outer mounts of the wishbones.

=1 means the wheel rates are adjusted only for the location the lower end of the pushrod relative to the inner and outer mounts of the wishbones.

Note that =1 does not remove all the geometric implications from the pushrod definition. Only removing the definitions of the pushrods themselves from the [CORNER] sections will do that."

It looks like none of the options will work with what I want to have.
That note to setting "1" is also confusing... rF use only location of the lower end of the pushrod.... but still, the whole pushrod geometry have some sort of influence (if I uderstand his description correctly).

What is important - I want to define susp rates by hand, in HDV.

If I understand correctly:
1. rFactor takes spring/damper mult (if they are 1.0, they won't have influence on calculations as 1.0^2 = 1.0 so they are neutral), takes geometry from pushrod and lower wishbone and calculate susp rates.
2. If [AdjustSuspRates=0] then it takes pushrod angles and calculate motion ratio.

What about pushrod angles, if AdjustSuspRates is set to "1"? rF assumes they are vertical or what?
 
Springmult and Dampermult are multipliers ......... they do not define the motion ratio ............ they simply multiply the effect of the springs and/or dampers ............... so they both should always be =1.0
 
Yes I know that and no - they don't have to always be 1.0. If you don't use pushrods and have suspension for which you have ratio (suspension ratio! Not motion ratio - don't confuse that ;-) ) different than 1.0, then you use those parameters accordingly.
 
I guess what's confusing then is the question.
So what you're trying to do then is have the suspension generate a linear effect using the multiplier.
 
So, you are assuming that rFactor is actually moving pushrods, so that their angles are changing real-time? Do you have any confirmation about that from ISI?

None the less, I have a feeling that you still miss the point of my question ;-)
 
If memory serves me correctly, Niels provided evidence of this some time ago, on the old RSC forum.
 
Do you have link to his post?
None the less, that still has nothing to do with susp ratio and I actually want to use pushrod geometry (for angles). What I don't want, is to have susp ratio calculated from susp geometry, so the main question remains.
 
Last edited:
Guys, does rFactor modify lower and upper outer susp points with camber change or the geometry always stays the same (from pm file), regardless of camber angles?
 
Guys, does rFactor modify lower and upper outer susp points with camber change or the geometry always stays the same (from pm file), regardless of camber angles?

Camber is applied at the hub/spindle, and not by adjusting the control arm lengths (as that would add a needless level of complexity).

And to your original post, the spring/damper multipliers only affect the garage screen, and can be any positive number. The multipliers are only there to mimick motion ratio (or, the leverage difference between the wheel and the "attachment point" of the spring). The actual rates are left unchanged.

And yes, the angle of the pushrods/struts change under compression. This can be quite useful, i have no idea why so many people used vertical pushrods for so long (falling rate springs, what?!). :p
 
Camber is applied at the hub/spindle, and not by adjusting the control arm lengths (as that would add a needless level of complexity).

And to your original post, the spring/damper multipliers only affect the garage screen, and can be any positive number. The multipliers are only there to mimick motion ratio (or, the leverage difference between the wheel and the "attachment point" of the spring). The actual rates are left unchanged.

And yes, the angle of the pushrods/struts change under compression. This can be quite useful, i have no idea why so many people used vertical pushrods for so long (falling rate springs, what?!). :p

I think that I must disagree with multipliers.

They don't affect garage screen, but they do cause wheel rate to be less or more depending if set below or above 1.

I have one vehicle where spring is set to 70 000, spring multiplier is 0.743 (older big block GM product) from my memory, it is a lot softer than when set multiplier to 1.0, but garage screen shows still 70 000 as that is what spring is, garage does not show wheel rates, but just springs.

Adjust suspension rates is set to 1 in this case if that has anything to do with how parameter is handled.
 
Ah i see. It has been a few years, but what i said still applies, just backwards. One is not affected and the other is, and all it is for is "fudging" the spring leverage.
 
In the mean time, I did some test spoke with some guys... as for susp rates - it works like quoted W_Bradshaw description (how "AdjustSuspRates" works when set to 0 and when set to 1).

SpringDamper multipliers are useful when implementing susp geometry using pushrods.

If we want to use those multipliers for susp ratio, then we have to put our pushrod geometry at the spindle and point vertically. Then, for spring/damper multipliers provide values already calculated as wheel rates.


Changing camber in the garage, also change lower and upper outer susp points and by that also changing lengths of A-arms.
 
Ah i see. It has been a few years, but what i said still applies, just backwards. One is not affected and the other is, and all it is for is "fudging" the spring leverage.

Yes, it is a fudge factor, but especially damper multi can be useful.

Think about suspension setup where you have spring mounted near hub at vertical, and dampers mounted to half way between spindle and pivot and they are at 30 degree angle, for example b-body mopars have something liike that, measures are probably different and of course it is live axle. Point is however that with only pushrods it would be darn hard to make that setup to work if you have real values for springs and dampers.

So there are situation where they are useful, but quite often dampers have to be calculated from wheel rate as data from specific car models is so difficult to come by.

Having car laboraory and resoures to buy any amount of suspension parts for every modder should be a rule :D


Wheelbase and track are such that if they are in HDV they rule and .pm is adjusted to match, but I think that when you set camber to certain in HDV that is camber with .pm arm positions and then garage settings change arms from there, but I can't be sure about this. With .pm file I trust carfactory as it is something I have never really got my head around with, I end always with buggy setup when I set arms way they are in reality., but measuring values and putting them into car factory seem to prduce something that works way it should so I have settled with that.
 
Camber is applied at the hub/spindle, and not by adjusting the control arm lengths (as that would add a needless level of complexity).

And to your original post, the spring/damper multipliers only affect the garage screen, and can be any positive number. The multipliers are only there to mimick motion ratio (or, the leverage difference between the wheel and the "attachment point" of the spring). The actual rates are left unchanged.

And yes, the angle of the pushrods/struts change under compression. This can be quite useful, i have no idea why so many people used vertical pushrods for so long (falling rate springs, what?!). :p

Sorry to pick back up an old thread, but i thought it was better than opening a new one, since my question is strictly related to this topic.

Something is still not clear to me reading the explanation above.

Let´s say i have a double wishbone suspension with a coilover spring and damper unit connected to a rocker that is then connected to the pushrod (like formula 1 cars, more or less), so actually something that you cannot model in rF. But the Motion Ratio (Wheel movement/spring movement) is quite costant, so it could be simulated quite well with a fixed number.

If i put the SpringMult and the DampMult to a value equal to my Motion Ratio, will the suspension finally have this motion ratio (so will the spring movement and the wheel movement be connected by this number) or still i have to do something on the pushrod?

My goal is actually to create a suspension with a linear behaviour with regard to motion ratio, how can it be done?
And would it be possible to simulate a suspension where the motion ratio is changing with wheel movement?

Thanks!
 
Odp: Susp rates

Sorry to pick back up an old thread, but i thought it was better than opening a new one, since my question is strictly related to this topic.

Something is still not clear to me reading the explanation above.

Let´s say i have a double wishbone suspension with a coilover spring and damper unit connected to a rocker that is then connected to the pushrod (like formula 1 cars, more or less), so actually something that you cannot model in rF. ...
...and stop here :)
In fact, that's exactly, how rf simulates suspension. Wondering, why you have "pushrods" defined in HDV? Now you know.
What rF does not simulate, is proper damper movement (as a separate unit) and that's why it also does not simulate McPherson strut.
DamperMult and SpringMult are responsible for ratio change on connection between pushrods and damper/springs (with coilovers, both multipliers will have the same values).

...But the Motion Ratio (Wheel movement/spring movement) is quite costant, so it could be simulated quite well with a fixed number.

If i put the SpringMult and the DampMult to a value equal to my Motion Ratio, will the suspension finally have this motion ratio (so will the spring movement and the wheel movement be connected by this number) or still i have to do something on the pushrod?

My goal is actually to create a suspension with a linear behaviour with regard to motion ratio, how can it be done?
And would it be possible to simulate a suspension where the motion ratio is changing with wheel movement?

Thanks!
The proper way, is to have motion ratio from suspension geometry itself and treat pushrods like pushrods in pushrod-type suspension (like F1, Maserati MC12 etc.) or treat pushrods as coilovers when working on non-pushrod susp type (like typical double-A arm susp found in many sports cars).

But of course, you can ignore pushrod geometry, set them vertical at the wheel centers and that way, you will have you constant motion ratio you are looking for.
 
...and stop here :)
In fact, that's exactly, how rf simulates suspension. Wondering, why you have "pushrods" defined in HDV? Now you know.
What rF does not simulate, is proper damper movement (as a separate unit) and that's why it also does not simulate McPherson strut.
DamperMult and SpringMult are responsible for ratio change on connection between pushrods and damper/springs (with coilovers, both multipliers will have the same values).


The proper way, is to have motion ratio from suspension geometry itself and treat pushrods like pushrods in pushrod-type suspension (like F1, Maserati MC12 etc.) or treat pushrods as coilovers when working on non-pushrod susp type (like typical double-A arm susp found in many sports cars).

But of course, you can ignore pushrod geometry, set them vertical at the wheel centers and that way, you will have you constant motion ratio you are looking for.


Yes, but how can you treat suspension geometry properly when no bellcrank is simulated? A big part of the suspension motion ratio comes from bellcranks, not only from the pushrod angle.

Probably here there is a word misuse: in real cars pushrods are only the beam that connect a wishbone to the rocker, not the spring axis.

So how to simulate the geometry properly? is SPRINGMULT and DAMPMULT only simulating the lever ratio of the bellcrank? But then, how the top pushord attachment movement is simulated (which movement magnitude is multiplied by this DAMPMULT or SPRINGMULT to get to the spring/damper movement)?

I still don´t understand how can we set them vertical, since also if we put the top pushrod hardpoint on a vertical line starting from the connection of the pushrod to the wishbone, then anyway the angle between pushrod and wishbone would change with suspension movement.
 
Last edited:
I don't know, what happens with pushrod (and its opt attachment), when it's beeing moved by wheel carrier. Does it move in direction it points at the moment or just like it was defined in HDV? Have no idea. Bellcrank (its ratio) is simulated by Spring/Damper Mult parameters.
When you put vertical pushrods at wheel centers, hub movement won't make much difference to ratio but yes, you are right. With very specific suspension geometry, with high camber gain, it might have some bigger impact.
 
There is no direct way to model a rocker based suspension. Using a F1 type 'pushrod' will result in a degressive wheel rate as the pushrod goes more and more horizontal with suspension travel. For a progressive wheel rate, I haven't tried.. but you could do the opposite, have the 'push rod' go *down* from the spindle to below the ground on the body. This would be a pullrod, I don't know if rFactor would cope..

I always put long ass springs in the middle of the tires, going 2 meters up. That gives a near perfect springrate = wheelrate. If I then have real data with a motion ratio, I simply adjust the spring rates rather than their position.
 
Niels, so does that mean, upper pushrod attachment (opposite to where it is connected with wheel hub) stays in the same spot all the time (like it would be a damper mounted to chassis)?
 
Thanks LesiU and Niels for this info.

So the only way to have a more or less constant wheel rate is to have a very long pushrod going exactly upper the wheel center.

I would use your knowledge for another question. What about antiroll bar rate?

I have found a topic in this forum dealing with this, but my question is related again to the motion ratio: i normally use the ARB stiffness defined in N/mm (not tube diameters and so on). Since i didn't see any multiplier for the ARB movment, will the stiffness that we define in the HDV and the in garage be seen by rF as an ARB WHEEL RATE?
 
Yes, ARBs in rF are "final" - you have to already take wheel rates into account.

EDIT: In rF 1, for some reason (a bug, miscalculation... I don't know exactly), ARB values have to be provided 2x stiffer.
In rF 2 that is fixed already, so provide values for stiffness as it is.
 
Last edited:
I never checked this but it seems by far the most logical that the ARB force is based on the difference in left and right 'pushrod position'. So if your pushrods are in the tire center and vertical, this is basically wheel rate. But placed inwards and at an angle, ARBs have little to do with wheelrates anymore.

That doubling thing got mentioned a long time ago, but it is not universal. When you isolate the anti roll bar, and measure its stiffness, I think you have to double that value in rFactor. If the rollbar is measured on the car 'at ground' as it sometimes is called in race car manuals, then you can use the value 1:1 (assuming you are comparing wheel rates and use vertical pushrods in tire center in rFactor).

edit, so it really seems to make the most sense to put the pushrods in the tire center long way up, because that gives you near perfect wheel rates for springs, dampers and ARBS. Then when you know the actual motion ratios, you adjust the springmult, dampermult and ARB values. :)
 
I never checked this but it seems by far the most logical that the ARB force is based on the difference in left and right 'pushrod position'. So if your pushrods are in the tire center and vertical, this is basically wheel rate. But placed inwards and at an angle, ARBs have little to do with wheelrates anymore.

That doubling thing got mentioned a long time ago, but it is not universal. When you isolate the anti roll bar, and measure its stiffness, I think you have to double that value in rFactor. If the rollbar is measured on the car 'at ground' as it sometimes is called in race car manuals, then you can use the value 1:1 (assuming you are comparing wheel rates and use vertical pushrods in tire center in rFactor).

edit, so it really seems to make the most sense to put the pushrods in the tire center long way up, because that gives you near perfect wheel rates for springs, dampers and ARBS. Then when you know the actual motion ratios, you adjust the springmult, dampermult and ARB values. :)

Thanks for the insight.

It makes sense, although it could be nice to have an official confirmation.

By the way, this story about the roll bar stiffness, is in my opinion really something connected to the definition of roll stiffness itself.

Many books look to this topic stating that roll stiffness must be calculated taking into accout the whole bar length, but when a car rolls, is it really the whole bar length that must be considered? or only half of it (because one wheel goes into bump and the other into reboud)?

Many vehicle dynamicists are for the second option, although it is also probably not 100% correct.
 
just a quick update.

I have just tried all the things we discussed here for pushrods and antirollbars on my F3 model and it matches very well real data now also regarding suspension movement.

Looking to the supension traces in motec now shows very similar results to what i have from my in car measured data.

regarding to the ARB, i am now using a vertical pushrod as suggested by Niels and roll values in motec seem to match quite well with real data when i use the wheel rate contribution of the antiroll bar itself, without doubling it. So it seems to me Niels is correct again, when you use ARB stiffness at ground, you don't need to double it.
 
Odp: Susp rates

I never checked this but it seems by far the most logical that the ARB force is based on the difference in left and right 'pushrod position'. So if your pushrods are in the tire center and vertical, this is basically wheel rate. But placed inwards and at an angle, ARBs have little to do with wheelrates anymore.

That doubling thing got mentioned a long time ago, but it is not universal. When you isolate the anti roll bar, and measure its stiffness, I think you have to double that value in rFactor. If the rollbar is measured on the car 'at ground' as it sometimes is called in race car manuals, then you can use the value 1:1 (assuming you are comparing wheel rates and use vertical pushrods in tire center in rFactor).

edit, so it really seems to make the most sense to put the pushrods in the tire center long way up, because that gives you near perfect wheel rates for springs, dampers and ARBS. Then when you know the actual motion ratios, you adjust the springmult, dampermult and ARB values. :)

The question is, does ARB really connected at pushrod location or are connected at wheel centers?
What I got as sort of official statement, is we have to double the values in rf1 and keep it 1:1 in rf2 but in both cases, they have to be calculated at wheel rates.

Niels, you have the tools to check that, so... can you check and confirm (or not)? :)
 
The question is, does ARB really connected at pushrod location or are connected at wheel centers?
What I got as sort of official statement, is we have to double the values in rf1 and keep it 1:1 in rf2 but in both cases, they have to be calculated at wheel rates.

Niels, you have the tools to check that, so... can you check and confirm (or not)? :)

One last thing on my side.

Can anybody cofirm that SPRINGMULT and DAMPMULT are to be considered as Wheel Movement/Spring (or damper) movement ?

Thanks again!
 
Those are multipliers for springs and dampers. It's up to your interpretation and your project for suspension geometry, how you want to use them. If you have a spring with rate 100N/mm and set SpringMult to 0.5, then that spring will become 50N/mm. Of course, if your suspension geometry have an effect on effective rates, then those will also be applied (and here remember about AdjustSuspRates parameter and its influence).
 
One way to check the arb rates is to have a car with roll centers at ground, then run it without arbs and say 50N/mm springs, at a certain speed do a 1G turn. Then add 50N/mm ARBs, see how much this changes the suspension position. I recall finding that adding front and rear 50N/mm arbs to the 50N/mm springs halved the suspension movement i.e. as if the effective rates were doubled.
 
One way to check the arb rates is to have a car with roll centers at ground, then run it without arbs and say 50N/mm springs, at a certain speed do a 1G turn. Then add 50N/mm ARBs, see how much this changes the suspension position. I recall finding that adding front and rear 50N/mm arbs to the 50N/mm springs halved the suspension movement i.e. as if the effective rates were doubled.

So it seems that the value we read in the garage is really the (single) wheel rate we are adding with the ARB to the overall wheel rate. good to know.

LesiU, from your comment and from this part of Bristow guide:

AdjustSuspRates=1 // Adjust suspension rates due to motion ratio (0 = direct measure of spring/damper rates, 1 = wheel rates)
Switch to permit the spring and damper rates at the wheels to be changed between two modes.

=0 means the wheel rates are adjusted for the geometry of the pushrod and the location of the lower end of the pushrod relative to the inner and outer mounts of the wishbones.

=1 means the wheel rates are adjusted only for the location the lower end of the pushrod relative to the inner and outer mounts of the wishbones.

Note that =1 does not remove all the geometric implications from the pushrod definition. Only removing the definitions of the pushrods themselves from the [CORNER] sections will do that.

and



SpringMult=1.0 // Take into account suspension motion if spring is not attached to spindle (affects physics but not garage display)
A multiplier-factor which is applied to the spring rate defined just below. This factor permits the spring rate to be adjusted for the fact that the spring is mounted somewhere inboard of the lower outer suspension point. By comparison, AdjustSuspRates back up in [SUSPENSION] is a switch which enables rF to make the calculation based on analysis of suspension and pushrod locations. This factor is available individually for each wheel. The switch in [SUSPENSION] is applied across the car.

So it seems that it is actually impossible to really impose a certain motion ratio to the suspension, also when you define the pushrod as Niels suggested.

That's very strange, since i have modeled my suspension using the SpringMult and DampMult as if they were motion ratios (wheel/spring) (so i have produce an overall ride stiffness higher than the spring one, since the multipliers were >1, while the real motion ratio does exactly the opposite, it makes "spring rate>wheel rate") and i got very very realistic spring movements when compared to real data.

I actually have AdjustSuspRates=1.

Still not totally clear to me what is the effect of AdjustSuspRates.
What is actually happening when it is =1 and when it is =0?
 
Last edited:
So back again here with another result to ask you guys a comment about.

Today i have tried to use both Adjustsusprates=1 and Adjustsusprates=0. I have seen no change at all in susp movement when analising data in motec.

As i said before, i have now my pushrods going vertical over wheel center and i use SpringMult and Dampmult both at the front and at the rear different than 0 (at the moment, they are equally to my real suspension motion ratios defined as wheel/spring movement).

As i said, the results seem quite ok compared to the real ones, but again i am missing the meaning of these parameters and would like to ask if anybody can shed some light.

I then tried to put all the springmult and dampmult equal to 1 and then damper movements have increased sensibly, since actually it seems as LesiU said in one of his first posts in this thread that suspension stiffness is decreased.

thanks!
 
Last edited:
Sorry for late reply.

I will explain that on an example. Let's say, we have a typical double A-arm suspension with standard coilover attached to lower A-arm on one end and to chassis on the other. Coilover is at some angle.


When you calculate wheel rate out of that, you take both: relative position of lower coilover attachment point vs whole length of A-arm and angle at which that coilover leans (usually towards chassis, you take cos of the angle) into account, right?

AdjustSusp rates can be set to:
- "0" - rFactor will calculate wheel rates taking into account those 2 things - lower attachment point of a pushrod (but rF assumes, it is connected to wheel hub, not lower A-arm) and angle of that coilover,
- "1" rF will only take lower attachment point. It doesn't care about pushrod angles.
 
Sorry for late reply.

I will explain that on an example. Let's say, we have a typical double A-arm suspension with standard coilover attached to lower A-arm on one end and to chassis on the other. Coilover is at some angle.


When you calculate wheel rate out of that, you take both: relative position of lower coilover attachment point vs whole length of A-arm and angle at which that coilover leans (usually towards chassis, you take cos of the angle) into account, right?

AdjustSusp rates can be set to:
- "0" - rFactor will calculate wheel rates taking into account those 2 things - lower attachment point of a pushrod (but rF assumes, it is connected to wheel hub, not lower A-arm) and angle of that coilover,
- "1" rF will only take lower attachment point. It doesn't care about pushrod angles.


Thanks once again for your help.

By the way, this should explain why, with my pushrod setup, switching from 0 to 1 produces no change in suspension behaviour.
Since the pushrod is vertical, it is not influenced (or not so much) by the fact that you take its angle into account or not.

I have, of course, another question, as i said in my previous post:


SpringMult=1.35 // take into account suspension motion if spring is not attached to spindle (affects physics but not garage display)

If you read the comment from ISI, it says it takes into account suspension motion. It is quite misleading. From my results, it seems to be a force multiplier, as you suggested, but it could also be a motion ratio somehow, if it is defined the opposite way i define it (so not wheel/spring movement but spring/wheel movement). Then increasing it, would make suspension stiffer.

But once again what surprises me is that my results seem realistic. and also my downforce should more ore less be, so it should not produce strange things on suspension deflection.
 
A simple test is to use SpringMult=100.0 and then divide the spring settings by 100.0

SpringMult=1.00
SpringRange=(50000.0, 5000.0, 1)
--->
SpringMult=100.0
SpringRange=(500.0, 50.0, 1)

They should give exactly the same result, but if there is any difference you should get a big effect if you use such a high multiplier. I can't remember for certain if the Mult parameters are motionratios (MR) or forcemultipliers (MR^2). So you can try both Mult=100.0 and Mult=10.0 for the settings above, but I'm pretty sure it's a force multiplier so only Mult=100.0 should work.
 
I think the springmult is just the division between the final wheel rate and the spring rate, let's say you have 100N/mm of spring rate and 1.15 of motion ratio, that would mean having a wheel rate of 76N/mm (100/1.15^2), so the springmult would be 0.76 (76/100). That's how I understand it works.
 
I think the springmult is just the division between the final wheel rate and the spring rate, let's say you have 100N/mm of spring rate and 1.15 of motion ratio, that would mean having a wheel rate of 76N/mm (100/1.15^2), so the springmult would be 0.76 (76/100). That's how I understand it works.

I just did another check and discovered that rFactor doesn't zero suspension sensors properly :D

Actually, as soon as the car is on the wheels, you can already see a suspension displacement. So when you read the overall suspension movement on track, you must subtract these values the suspension sensors shows at the very beginning of the sensors.

When i do so, my suspension became immediately too stiff and it makes much more sense to reduce the stiffnness by MR^2.

I will try it.
 
Yes, the suspension should be built with the dampers fully extended (unloaded). The telemetry will show this @ 0. When the corners are first loaded you will see compression until the spring rates equalize the load of the body. If you are getting that technically detailed remember the telemetry is also showing the vertical movement of the wheel (not the movement in the damper).
 
Yes, the suspension should be built with the dampers fully extended (unloaded). The telemetry will show this @ 0. When the corners are first loaded you will see compression until the spring rates equalize the load of the body. If you are getting that technically detailed remember the telemetry is also showing the vertical movement of the wheel (not the movement in the damper).

Thanks, you have aniticipated another question i wanted to ask.

So actually suspension movement channels in motec are showing the kynematic movement of the suspension at wheel (so not tyre deflection yet considered, it is considered in ride height measurement i guess).

Good to know, it will be useful to compare to real car data. This explains also why i was seeing somehow bigger movements, in my case.

Thanks!
 
Not quite true however! Telemetry has the suspension position which is actually the spring length. When motion ratio is 1, and each corner has 2000N of sprung load, with a 50.000N/m spring, the 'suspension position' telemetry will be 40mm with the car at rest. It is wrongly named in telemetry. I also think it is measured on the pushrod so it doesn't measure wheel displacement.
 

Back
Top