Working with a suspension spectrum analyser [Archive]

Honey

19th December 2005, 15:25

if you did an fft over the whole track, the graph is (as it seems to me) meaningless, because you are outside of "stationariety" of the signal...well i am assuming that the suspension response it is not over a whole track, i think is better you consider many fft over small time windows, enough small for the suspension response to be considered "stationary"...also i think the update time of the phisic model should be important for the aliasing issue, but as long we assume (as i think) the lfs phisic model is good, the i believe it is fast enough to avoid aliasing.

so my hint is to try to fft only a small time...maybe the crossing of the chicane or much better the test wich mercedes class A failed: a sudden steer and countersteer on a plain straight. hope it helps ;)

Bogey Jammer

19th December 2005, 17:17

So, FFT analysis is only for weight transfer watching? :worried:

I'll try to FFT the curves only

Bogey Jammer

19th December 2005, 17:36

I think hard but I can't see the relation between weight transfer and frequency

It sounds very strange for me, I'm afraid I'll dream about this question this night lol

Bob Smith

19th December 2005, 17:48

Well as you are only analysing suspension frequencies under g, then as I understand it there should be a peak either side of the "natural frequency". For cornering, this will be due to the left/right weight transfer, changing the mass rest on the spring, and therefore the freqency will change. Being that cornering is usually done at the limit of the tyres, there weight transfer will often be the same amount, and depending on it being a left or right corner, those peaks make sense.

I still have a few questions though:
What is the y scale?
What data is it you've actually plotted.

If you changed the scale to be track time (x) vs. frequency (y) you should see the left frequency going high and the right low during cornering (or vice versa), and going high together under braking (since you have been analysing the rear wheels).

Edit: you posted while I was browsing, to answer your question:
The relationship is that the spring stiffness is fixed (the value you see in the garage), while the frequency depends on the mass on the spring. Weight transfer is taking off load from one spring, and putting more on another. So while the spring stiffness never changes, the frequency should drop on the additionally-loaded spring, and increase on the less-loaded spring (during cornering, this is the inside wheel).

Honey

19th December 2005, 17:52

no i only meant short period of time, if u go straight it helps finding bump frequency of the track, i only intended that if suspensions are not solicitated, the spectrum of the output will be (teoretically) zero, because it has zero input...only this ;)

the "merceds test" should help finding compression and rebound frequencies in a single step.

Bogey Jammer

19th December 2005, 18:12

Thanks!

I understand and the subject becomes very interesting.

What is the y scale?
What data is it you've actually plotted.

y scale is in dB
One lap of the track is plotted, but all data between -0.40G and 0.40G is not included.

StewartFisher

19th December 2005, 18:19

The relationship is that the spring stiffness is fixed (the value you see in the garage), while the frequency depends on the mass on the spring. Weight transfer is taking off load from one spring, and putting more on another. So while the spring stiffness never changes, the frequency should drop on the additionally-loaded spring, and increase on the less-loaded spring (during cornering, this is the inside wheel).
I disagree. Following the discussion about suspension frequency change with downforce loading I went right back to the differential equations for a spring-mass-damper system and confirmed that spring load does not affect the suspension frequency. Only the mass supported by the spring and the spring rate (and, if underdamped, the damping constant) change the suspension frequency. During cornering the weight distribution changes but the mass distribution doesn't. This is ignoring change of CoG with chassis roll obviously, but that should be fairly small.

5th Earth

19th December 2005, 19:38

I disagree. Following the discussion about suspension frequency change with downforce loading I went right back to the differential equations for a spring-mass-damper system and confirmed that spring load does not affect the suspension frequency. Only the mass supported by the spring and the spring rate (and, if underdamped, the damping constant) change the suspension frequency. During cornering the weight distribution changes but the mass distribution doesn't. This is ignoring change of CoG with chassis roll obviously, but that should be fairly small.

Yah, he's right, this is the problem I ran into back when I was making my suspension spreadsheet. I initially thought that downforce would have an effect on spring frequencies, but eventually we determined downforce had no effect at all.

But then that begs the question, why do the graphs have multiple peaks, above and below the natural frequency?

Further fuel on the fire--don't forget that the tires, being flexible, also have a frequency, usually much higher than the suspension itself (anywhere from 5-13 Hz, given a quick look at Bob's GRC). With critical damping (generally not actually used), the flex of the tires over bumps will be pretty significant.

colcob

19th December 2005, 20:10

Just a few quick contributions from me, its past my bedtime.

Firstly, Spring frequency doesnt change with load transfer as I understand it, its certainly not the case that weight transfer changes the MASS resting on a spring, it just changes its current load, therefore should not influence frequency.

Also, I presume your input data was suspension travel. Now what we call 'suspension frequency' is actually the frequency of the sprung mass resting on the suspension under gravity.
But a measurement of suspension travel doesnt just contain data about the movement of the sprung mass, it also contains data for the movement of the unsprung masses, and without incorporating data about the absolute z-position of the two masses, its impossible to separate the two in the suspension travel data alone.

So your graphs are a bastard child of spring mass frequency and unsprung mass frequency, and probably tell us as much about the bumps on the track as they do about the suspension. Although even that data is distorted by the fact that all the bumps will have been run over at different speeds.

Bogey Jammer

19th December 2005, 20:17

About tires you said that their frequency range is 5-13Hz, I think it doesn't interfere in the case of the LX4

edit: your post frozen me colcob :melting:

Honey

19th December 2005, 20:17

But then that begs the question, why do the graphs have multiple peaks, above and below the natural frequency?
as i said before i believe that the suspension output, as a signal, cannot be considered stationary (statistically speaking), so it's "meaningless".
what should be done IMO is calculating fft over short periods of time...short enough that the suspension signal can be considered stationary within it.
then plotting all the fft results in a graph of the spectrum against time, this helps finding the track-bumps frequency...

using math is cool, but it should be applied under the theorem's hypotesis to be meaningful, which is not this case IMO.

Bob Smith

19th December 2005, 20:50

I disagree. Following the discussion about suspension frequency change with downforce loading I went right back to the differential equations for a spring-mass-damper system and confirmed that spring load does not affect the suspension frequency. Only the mass supported by the spring and the spring rate (and, if underdamped, the damping constant) change the suspension frequency. During cornering the weight distribution changes but the mass distribution doesn't. This is ignoring change of CoG with chassis roll obviously, but that should be fairly small.
Yeah OK then, cornering is more like changing the value of g for different sides of the car than actually moving the mass. But that makes it much harder to understand the data. :(

So, thinking about the sprung/unsprung masses bit that Colcob brought up, wouldn't we then technically have two frequencies, which could be seen with a spectral view? Actually we'd have two moving bodies, both of which could have more than one frequency. I'm just thinking about sound, since when a speaker cone moves it's moving one path but can be moving at many different frequencies.

More thinking: if it is suspension travel is what was used to create the graphs, then even though weight transfer won't affect frequencies, it will still cause the suspension to compress on one side and extend on another (i.e. roll). So this would still affect the graph even without mass transfer occuring. Yes/No?

Oh and finally as for the tyre frequencies that are present in GRC, do not take them to be overly accurate as I've very little data to base my tyre deformation formula on (that and it's also very simplistic).

Shotglass

19th December 2005, 21:50

hoiw exactly did you calculate those graphs anyway ?

Bogey Jammer

20th December 2005, 05:16

hoiw exactly did you calculate those graphs anyway ?
Open RAF with F1perfview, export it to a csv file, import it into excel, calculate suspension velocity, export it into a FFT program, and (in my case) import FFT data in excel.

OK I think now that working FFT with LFS is useless, I can't see what I thought I could see, and as I was afraid, suspension are mainly setup with feelings and many tests, not with calculation. I'm a bit disappointed since I spend very long time for it but I'm cool now because I don't need to break my head for things I can simply do.

colcob

20th December 2005, 07:28

edit: your post frozen me colcob :melting:

Um, I'm not sure what you mean, but I'll try and explain it a bit better, I was in a rush last night.

Firstly, I dont think its a waste of time what you've done, I'm sure that this kind of analysis can be useful, but filtering and interpreting the data is probably something that only gets taught in third or fourth year of a auto engineering degree course, so its going to take us a while to work it out from scratch.

To try and explain a bit better what I meant, imagine the following:
A car is driving a course on a perfectly flat level surface, so the unsprung mass, ie the wheels and some suspension components, do not move vertically at all.
The Sprung mass, ie the chassis, moves around on the suspension and induces changes in the suspension travel.

In this situation, an FFT analysis of the suspension velocity will show you something about the natural frequency of the suspension and any harmonics.

But in the real world, the road surface has small high frequency bumps, large low frequency changes in elevation and everything in between, which all induce changes in the suspension travel, even if the car is driving in a straight line at constant speed. So all of those oscillations feed into the suspension travel data, and add frequencies to the resulting FFT plot.

But thats not necessarily a bad thing, because we only really care about how the car drives in the real world, and handling those bumps and elevation changes is one of the main problems behind suspension setup.

It would be interesting to see plots from a short section of straight from 3 different tracks, in the same car with the same setup, to see if this picks up noticeable differences in the dominant frequency of bumps on the different tracks.

This could actually be useful data for car setup, as say you know that South City has predominant bumps of about 5hz in a braking zone, it might be a bad idea to have a front suspension frequency of 2.5hz for example.

Thats all getting pretty hypothetical, particularly as the bump frequency depends entirely on how fast you drive over them, but you get the idea.

Shotglass

20th December 2005, 08:46

Open RAF with F1perfview, export it to a csv file, import it into excel, calculate suspension velocity, export it into a FFT program, and (in my case) import FFT data in excel.

did you scale those plots ? normally they should go up to 100 Hz

EDIT:

calculate suspension velocity

oh and this step is just plain wrong you have to do the fft over the deflection of the suspension and not its time-derivative

and another thing from whar youve said ive gathered that you did cut out blocks from the data sets put those together and did the fft over that
is that right ? cause if you did youll get false fft results

EDIT2:

oh and another thing ... the derivation to get suspension velocity explains the spectrum you get as what you did is a prediction so the overall signal should become more or less white which is exactly what those plots show

Bogey Jammer

20th December 2005, 14:42

Um, I'm not sure what you mean, but I'll try and explain it a bit better, I was in a rush last night.
Sorry to be not clear, I understood that you wrote and it proves that my FFT graph are misused and not accurate. Anyway I'll keep my tool for later if needed.

did you scale those plots ? normally they should go up to 100 Hz hum, I don't understand what you mean, but I can say that the scale was automatic and linear, I only display a range from 0Hz to 5Hz.

oh and this step is just plain wrong you have to do the fft over the deflection of the suspension and not its time-derivative

and another thing from whar youve said ive gathered that you did cut out blocks from the data sets put those together and did the fft over that
is that right ? cause if you did youll get false fft results
Yep, you must be in the thruth. However, I'm not sure about FFT the deflection. I tried once and the results were weird.
What do you mean by "white"?

Shotglass

20th December 2005, 16:02

hum, I don't understand what you mean, but I can say that the scale was automatic and linear, I only display a range from 0Hz to 5Hz.

raf files sample data at 100 Hz therefore the fft should produce points between 0 and 100 Hz

However, I'm not sure about FFT the deflection. I tried once and the results were weird.

then there must be something wrong with the way you calculated it
you want the spectrum of the spring oscillations so youll have to analyze the amplitude of the springs deflection over time

What do you mean by "white"?

white as in same magnitude for all frequencies

But in the real world, the road surface has small high frequency bumps, large low frequency changes in elevation and everything in between, which all induce changes in the suspension travel, even if the car is driving in a straight line at constant speed. So all of those oscillations feed into the suspension travel data, and add frequencies to the resulting FFT plot.

which is exactly what you want

for a valuable spectral analysis the ideal road would be one that has bumps at all frequencies (or at least at all frequencies up to half the sample rate at which you measure spring deflection) in essence a white noise road surface

StewartFisher

20th December 2005, 20:13

Yeah OK then, cornering is more like changing the value of g for different sides of the car than actually moving the mass. But that makes it much harder to understand the data. :(

So, thinking about the sprung/unsprung masses bit that Colcob brought up, wouldn't we then technically have two frequencies, which could be seen with a spectral view? Actually we'd have two moving bodies, both of which could have more than one frequency. I'm just thinking about sound, since when a speaker cone moves it's moving one path but can be moving at many different frequencies.

More thinking: if it is suspension travel is what was used to create the graphs, then even though weight transfer won't affect frequencies, it will still cause the suspension to compress on one side and extend on another (i.e. roll). So this would still affect the graph even without mass transfer occuring. Yes/No?
Since the 'suspension travel' is measuring the extension of the spring, you'd have one frequency for the vertical motion of the unsprung mass and one frequency for the vertical motion of the sprung mass. You'd also probably get frequencies corresponding to the roll and pitch frequencies as both of these will affect the ride height on an individual wheel.

colcob

21st December 2005, 05:23

for a valuable spectral analysis the ideal road would be one that has bumps at all frequencies (or at least at all frequencies up to half the sample rate at which you measure spring deflection) in essence a white noise road surface

I dont necessarily agree with that, as a real road almost certainly doesnt have uniformly distributed bump frequencies up to 50hz, but probably has certain characteristic bump frequencies that might be worth taking account of. Only guessing though.

Shotglass

21st December 2005, 11:24

I dont necessarily agree with that, as a real road almost certainly doesnt have uniformly distributed bump frequencies up to 50hz, but probably has certain characteristic bump frequencies that might be worth taking account of. Only guessing though.

i think the problem here is a fundamental misunderstanding of the relation between a fft analysis and what happens on a real road on your side

let us assume for the moment that the wheel always stays in contact with the road surface and that the tyres are solid (or at leat that their effect is negligable) so all the bumps in the road directly affect the spring with the cars mass attached to it

think of the road (or rather the bumps it has ie the changes is surface height and the frequencies at which they occur) your using for the analysis as an input signal (at this point the solid tyres are important so the bumps in the road directly excite the suspension without prior filtering through the tyres ... generally this assumption isnt neccesary but the relation between bumps and the suspensions reaction wouldnt be as tighly as it is under said assumption)
by driving across that road your feeding this signal into your suspension which is the system you want to analyze (here the assumption of a non lifting wheel comes into play as we dont want to analize two interacting inseparable systems)
the output from the system ie the reaction of your suspension to those bumps is the spring deflection
this is basically a system identification problem

for system identification (especially if all you care about is magnitudes which is the case for suspension analysis as phase rarely matters) a white input signal is a good choice since it excites the system at all frequencies so youll get the systems frequency response for all frequencies
the main idea here is that the spectrum of the excitation relates to the nodes where you are able to gather data about the fequency response
the spectum you get at the output ie the spectrum of the springs deflection is actually the multiplication of the excitations and the systems sprectra (and to be exhaustive also the spectrum of the window you use for the fft)
therefore unless your excitation is truly white (ie same magnitude for all frequencies) the outputs spectrum will be distorted by the inputs spectrum

after you have the response to all those frequencies you can evaluate any road surface by multipliing the spectrum of the road surface with the spectrum of the suspension ... the result of this will be the spectrum of the suspensions response so you can tell at which frequencies the suspension will swing and who lage the amplutide of those oscillations will be (unfortunately a fourier analysis isnt exactly useful for transient behaviour)

another way to do this would be sweping through the fequencies ... unfortuatelly the only way i can think of to do this is by wiggling the steeringwheel left to right at increasing frequency while driving on a perfectly flat road surface ... unfortuantely that way its practically impossible to ensure that the amplitudes at which you excite the suspension are the same for all frequencies

colcob

21st December 2005, 13:07

Okay. I think I get you. You obviously know substantially more about this kind of thing than me.

Honey

21st December 2005, 17:50

the spectrum of the springs deflection is actually the multiplication of the excitations and the systems sprectra (and to be exhaustive also the spectrum of the window you use for the fft)
you are correct about all points, except that IIRC the spectrum of the output is equal to the product of the spectrum of the "transfer function" and the spectrum of the input signal if and only if the system is stationary and i think it is not.

IMO fft should be applied to small periods of time to be meaningful, don't you think? ...i'm asking it to you because you seem to have fresh in mind the theory :)

ps as the lfs suspension model is linear, shouldn't be an impulse the spectrum fo the suspension transfer function?

Shotglass

21st December 2005, 20:04

you are correct about all points, except that IIRC the spectrum of the output is equal to the product of the spectrum of the "transfer function" and the spectrum of the input signal if and only if the system is stationary and i think it is not.

imho its stationary under the asumption that the wheel doesnt lift from the ground
oh and if im not mistaken the transfer function is the systems spectrum (ie the fourier transform of the systems impulse response)

IMO fft should be applied to small periods of time to be meaningful, don't you think? ...i'm asking it to you because you seem to have fresh in mind the theory :)

depends on what you want to do with it
for system identification (of time invariant systems) you want a large number of nodes along the frequency axis therefore youll use a large number of samples
for analysis of instationary proceses however (like used for adaptive prediction in (speech) coders or for utilizing psychoaccustic effects like in mp3 coding) you want shorter sets of samples (during which the process can be considered stationary) to get a better resolution in the time domain (this has to be paid with a loss in spectral resolution though ... basicaly spectral and time-domain resolution are interchangeable and have a joint upper bound much like the minimal uncertainty (to be considered the inverse of resolution) in impulse and position measurements have a joint lower bound (heisenberg))

ps as the lfs suspension model is linear, shouldn't be an impulse the spectrum fo the suspension transfer function?

im not sure what you mean with that question

Honey

22nd December 2005, 17:49

imho its stationary under the asumption that the wheel doesnt lift from the ground
oh and if im not mistaken the transfer function is the systems spectrum (ie the fourier transform of the systems impulse response)

IMHO to be stationary it would be required to lock the steer at 0° (and thus a straight road), about wheel that lift from ground, i'm not sure if ti could be the reason for non-stationary, but sounds pretty reasonable.
about the definition of "transfer function" you are right, i remembered wrong: the time domain function is the "impulsive response", my mistake :)

depends on what you want to do with it
for system identification (of time invariant systems) you want a large number of nodes along the frequency axis therefore youll use a large number of samples

since the question that originated this thread was basically "how to find track bumps frequencies", i think only small time windows can reveal that.
it's like the human speech (which is clearly non-stationary), if you want to know the spectrum of the letter "i" inside the word "stationary" you surely don't calculate the spectrum of the word but you will apply a proper time window that surronds the "i" on the time axis and then calculate the spectrum only on that time window

for analysis of instationary proceses however (like used for adaptive prediction in (speech) coders or for utilizing psychoaccustic effects like in mp3 coding) you want shorter sets of samples (during which the process can be considered stationary) to get a better resolution in the time domain (this has to be paid with a loss in spectral resolution though ...

basically is what i said before too, but i think specral resolution should not be an issue for suspension and track analisys

basicaly spectral and time-domain resolution are interchangeable and have a joint upper bound much like the minimal uncertainty (to be considered the inverse of resolution) in impulse and position measurements have a joint lower bound (heisenberg))

this is totally different, the spectral estimation is caused by implied limits of mathematical definitions/functions against reality. the minimum uncertainity (the heisenberg principle) is a conseguence of the fact that measuring "something" implies perturbing the system and thus it changes the system itself, so the measured value is actually the measure of the whole system "original system to be measured" + "measuring instrument".

im not sure what you mean with that question
basically i was wondering if someone (except the devs) knows more about the lfs suspension model...

Shotglass

23rd December 2005, 01:04

IMHO to be stationary it would be required to lock the steer at 0° (and thus a straight road)

steering angle ... interesting point ... but it doesnt actually change the suspension geometry (or at least not in ways which have a effect on suspension frequencies)

about wheel that lift from ground, i'm not sure if ti could be the reason for non-stationary, but sounds pretty reasonable.

actually stationarity is the wrong word since it is a property of random processes ... for systems the right word would be time variant and a suspension that basically floating in mid air is obviously different from one with gorund contact
matter of fact the suspension is inherently time variant though as it has different damping depending on the direction of travel

since the question that originated this thread was basically "how to find track bumps frequencies", i think only small time windows can reveal that.

hmmm i thought the question was "how to analyse suspension transfer functions

this is totally different, the spectral estimation is caused by implied limits of mathematical definitions/functions against reality. the minimum uncertainity (the heisenberg principle) is a conseguence of the fact that measuring "something" implies perturbing the system and thus it changes the system itself, so the measured value is actually the measure of the whole system "original system to be measured" + "measuring instrument".

actually this isnt what heisenberg is about its a very common false explaination though

Bogey Jammer

23rd December 2005, 09:58

Very impressive :cry:

FFT is more hardcore than I though lol

Honey

23rd December 2005, 11:39

Very impressive :cry:

FFT is more hardcore than I though lol
happy to impress you, but...you could also clarify our "uncertainity": do you want to find track bump frequencies or the suspension transfer function (ie the spectrum), or both?

if you don't stop us by answering that, we can continue for years... :P

Honey

23rd December 2005, 12:40

steering angle ... interesting point ... but it doesnt actually change the suspension geometry (or at least not in ways which have a effect on suspension frequencies)

i cannot find a better way to explain wat i mean, than the following example (i hope): think of steering enugh to bring suspension near saturation...now if you hit a bump, the suspension saturates/clips an thus you have as output a spectrum with infinite bandwidth caused by non linearity. well this is extreme, but the concept is that you are introducing an "unknown" force/signal that is biasing the system and such bias is time variant, so also the system will be time variant. imo.

actually stationarity is the wrong word since it is a property of random processes ... for systems the right word would be time variant and a suspension that basically floating in mid air is obviously different from one with gorund contact
matter of fact the suspension is inherently time variant though as it has different damping depending on the direction of travel

yes it is clearly so, the fact is that in the last years (lol sounds like i'm an old mummy :D ) i work on a totally different field, so i always feel there is something important that i don't remember about this topic

hmmm i thought the question was "how to analyse suspension transfer functions

let's hope the topic starter clarify us wat he wanted to know

actually this isnt what heisenberg is about its a very common false explaination though
well my example was the problem that brought to the heisenberg relationship and may seem not clear how the example fit into that, he found the relation because during his studies about quantistic theory he faced the problem of measuring at the same time speed and position of electrons. on the other hand the problem of spectral estimation is given basically by removing any "infinite" value from the formulas given by theory such as integral/series and mostly by approximating algorithms to calculate it, so it's mostrly "numeric" precision...i understand you point of view, but i don't agree and one thing that i think "prooves" it is that error forumulas and relationships about spectrum estimation are expressed in terms of part of integrals you give away or terms of series you truncated and not by the heisenberg formula, another thing: one problem is "measurement", another is "estimation/approximation" the two problems may be considered "dual" perhaps but they are generated from different causes anyway all of this may be considered one of the mathematical/phisical things that have different toughts, i believe none of us will change point of view about that, but surely it isn't a problem ;)

Shotglass

23rd December 2005, 15:44

i cannot find a better way to explain wat i mean, than the following example (i hope): think of steering enugh to bring suspension near saturation...now if you hit a bump, the suspension saturates/clips an thus you have as output a spectrum with infinite bandwidth caused by non linearity. well this is extreme, but the concept is that you are introducing an "unknown" force/signal that is biasing the system and such bias is time variant, so also the system will be time variant. imo.

im not entirely sure about this but my gut feeling says yes and no
on the topic of saturation ... yes youre right this sort of comes down to my point that the wheel should always stay in contact with the ground only in the other direction which i forgot to consider as it should never happen if the setup is fit for the track
on the bias point id say no since lfs doesnt simulate progressive springs (yet) therefore the spectrum shouldnt change with biasing

ive just thought of another thing i hadnt considered properly yet which is the time variance induced by the 2-way dampers ... obviously the different damping constants create 2 different transfer functions (tf)
to get rid of this youd either have to
1) use no dampers at all therefore youll get the pure tf of the spring and can then derive the 2 damped tfs from that
2) use the same values for both compression and bump damping and after 2 test runs with your desired values for bump and compression youll have both tfs of the time variant system
3) only analyze blocks of data where the suspension is traveling in one direction ... this will of course lead to poor spectral resolution

another thing: one problem is "measurement", another is "estimation/approximation"

i could argue that the spectrum estimation problem is a measurement problem though since as you cant improve spectral resolution by higher sampling frequencies the only way to do so is by longer measurements and therefore by sacrificing temporal resolution

the two problems may be considered "dual" perhaps

frankly i dont know enough about quantum physics to see if i could actually prove that both problems are the same so lets settle this with calling them dual

Honey

23rd December 2005, 16:25

im not entirely sure about this but my gut feeling says yes and no
on the topic of saturation ... yes youre right this sort of comes down to my point that the wheel should always stay in contact with the ground only in the other direction which i forgot to consider as it should never happen if the setup is fit for the track
on the bias point id say no since lfs doesnt simulate progressive springs (yet) therefore the spectrum shouldnt change with biasing

ive just thought of another thing i hadnt considered properly yet which is the time variance induced by the 2-way dampers ... obviously the different damping constants create 2 different transfer functions (tf)
to get rid of this youd either have to
1) use no dampers at all therefore youll get the pure tf of the spring and can then derive the 2 damped tfs from that
2) use the same values for both compression and bump damping and after 2 test runs with your desired values for bump and compression youll have both tfs of the time variant system
3) only analyze blocks of data where the suspension is traveling in one direction ... this will of course lead to poor spectral resolution

point (2) is the only completely correct, but is insufficient because it is a simplified sysem which is rarely used by lfs, to better model we should write down equations, but i surely don't want to do this :D

i could argue that the spectrum estimation problem is a measurement problem though since as you cant improve spectral resolution by higher sampling frequencies the only way to do so is by longer measurements and therefore by sacrificing temporal resolution

i think we can argue for years about that, but for the forum sake we better consider it OT... :D

frankly i dont know enough about quantum physics to see if i could actually prove that both problems are the same so lets settle this with calling them dual
well i'm not an expert too, but i had to study enough (too much for my taste :D ) of it in several courses, but i had to study much about measurement (i'm an electronic engineer) where heisenberg was a key point...but obviously could be that my teachings were biased and so am i.

what comes in my mind now is a phrase that one of my professors was used to say:"the art of mathematics is giving same names to different things", well i never agreed with that, but i think it fits to our discussion :)

so i will follow your suggestion: let's end calling it "dual" :D

PS i'm still awaiting that the thread starter will answer to our doubt..."suspension" or "bumps"?

Bogey Jammer

23rd December 2005, 19:12

Honey

23rd December 2005, 20:14

I have not finished to study all the last posts yet but to answer to the question: FFT is quite useless for me now. I wanted to use it to help me to setup my suspensions. I thought I would adapt suspension frequency to track frequency but it was wrong since track frequency is not clearly defined, and suspension frequency depends on the car.

However, FFT is useful for something, but I don't know what, it seems that it's not very helpful...
well my suggestion is always the same: don't apply fft over th whole track, but divide the track into "interesting" segments i.e. each corner, segments with bumps, etc. then examine both time and frequency graphs the reason in doing this apart from possible mathematical reasons, is that you will deal with less information (divide et impera), so it may be easier to find caracteristics/behaviour on segments of track, if so then it will be up to your decision how to find the compromise that satisfies you.

as i said before, i'm not able to interpret the graphs you showed (doing fft over the whole track), so maybe you could at least give a try at doing a segmented analisys, it shouldn't hurt ;)

Shotglass

24th December 2005, 00:42

point (2) is the only completely correct, but is insufficient because it is a simplified sysem which is rarely used by lfs, to better model we should write down equations, but i surely don't want to do this :D

well the way i see it a suspension with 2 way dampers is switching between 2 systems all the time ... so the obvious way to predict behaviour of the whole system is by first analyzing both systems seperatelly as good as possible

so i will follow your suggestion: let's end calling it "dual" :D

actually i rethought on that and came to the conclusion that dual doesnt mean basicalle the same but rather the same thing but done exactly the other way round ... therefore dual means the opposite of what we were both trying to say with it

I thought I would adapt suspension frequency to track frequency

what exactly do you mean with adapt ?