Absolutely. For as much money as everyone spends on these bikes, to run them with only guesswork as far as suspension setup goes is a little ridiculous. I was always of the opinion that a lot of the damping adjustments just aren't that noticeable to most of us, but was having all kinds of problems with sequences of smaller, higher frequency impacts. Decreased the amount of rebound damping a bit front and rear and it's like a completely different bike. Amazing what a setup change can do for your confidence.
Is their a way that really fast guys set up their suspension,in general .
LOL! Great post!
Today I did my very technically advanced suspension analysis and adjustment.
First run(6:30 am)had my usual stiffer suspension.I wasn't very awake at all and thus rode like crap and thought it was a good idea to blame the suspension.
Second run I thought I'd smooth out the initial stroke compression and did a much rockier trail...Got hung up on a rock on the front side,which shifted my weight forward,compressing the suspension,rolling over the rock,suspension compressed,full endo,bike on my head,broken deraileur,bent seat,road rash on my back,elbow,side,calf.
My graph says stiffer is more better....my wallet is agreeing.
First run(6:30 am)had my usual stiffer suspension.I wasn't very awake at all and thus rode like crap and thought it was a good idea to blame the suspension.
Second run I thought I'd smooth out the initial stroke compression and did a much rockier trail...Got hung up on a rock on the front side,which shifted my weight forward,compressing the suspension,rolling over the rock,suspension compressed,full endo,bike on my head,broken deraileur,bent seat,road rash on my back,elbow,side,calf.
My graph says stiffer is more better....my wallet is agreeing.
haha righto. FFTs of wheel displacement only give you very implicit data, as in you can't look at a PSD of wheel travel and go "well then I need more rebound and less HSC" or whatever. What it does is help you understand the natural frequencies of the bike itself (natural frequency of the wheel, natural frequency of the sprung mass, how the rider moves, pedalling inputs etc) much moreso than the relatively random inputs. If you rode down a set of stairs or on a course with unbelievable amounts of braking bumps (Thredbo for example) you might start to see discernable input frequencies, but for the most part nothing much shows up there, all it really does is show you the vehicle's modes of response. It's not easy stuff to understand, and I'm still learning a lot about it. However, relatively speaking, I'm only at the beginning of my research
Yeah, that's kind of what I expected. I'm not really familiar with this technique for analysing chassis dynamics. In my field, we nearly always just want to push resonant frequencies as high as possible so we can drive harder at lower frequencies without getting oscillations. Seems that's not the case here. Typically bicycles seem like they're slightly underdamped, so I imagine increasing damping within the range of a resonant mode would bring that mode lower in frequency, no? Although it's not immediately obvious to me the difference between adding the damping to the rebound or the compression stroke would make. I'd have to run some numbers. As a random thought, the restoring forces due to the spring vary far less than the compression forces due to impacts etc...
I'm still struggling to see the justification for this technique though. I can't say I've ever felt my bike resonate, per se...
You gonna PM me a link to your thesis or what? I have educational access, so a journal reference is fine.
A bike is also underweighted or overweighted, depending on the rider's movements and intentions...?
Also completely different to my (old) field. 15g RMS @ 2000-20kHz? Sure, no problem - here you go.... but 5-15 Hz inputs? Variable loadings? Eh? Sounds like magic to me.
Fvckin suspension, how does that work?
ICP rulz!
What PSDs do is let you compare the various frequencies between bikes; kind of like comparing damper data. When you know what damping coefficients you need for LSC, HSC, LSR, HSR etc then you can compare that between one damper/bike and another. When you know the resonant frequencies of the bike, you can start to understand how different adjustments affect each one, and what effect that will have on the ride, but like I said, it's very implicit. This is an area I am still working on, you might start to see some results in 12 months time or soYeah, that's kind of what I expected. I'm not really familiar with this technique for analysing chassis dynamics. In my field, we nearly always just want to push resonant frequencies as high as possible so we can drive harder at lower frequencies without getting oscillations. Seems that's not the case here. Typically bicycles seem like they're slightly underdamped, so I imagine increasing damping within the range of a resonant mode would bring that mode lower in frequency, no? Although it's not immediately obvious to me the difference between adding the damping to the rebound or the compression stroke would make. I'd have to run some numbers. As a random thought, the restoring forces due to the spring vary far less than the compression forces due to impacts etc...
I'm still struggling to see the justification for this technique though. I can't say I've ever felt my bike resonate, per se...
You gonna PM me a link to your thesis or what? I have educational access, so a journal reference is fine.
I would also like to start working on some stability criteria stuff but that's a while away yet, waiting on a few bits and pieces first.One thing worth considering when it comes to the difference (in resonance) between adding compression and adding rebound damping is the concept of "effective" frequency. When you say ride your bike down the road and hit a kerb, the bike goes from no motion, to very high acceleration of the wheel, to high velocity of the wheel, to a lower acceleration of the wheel in the opposite direction, to a moderate (rebound) velocity of the wheel, to a relatively slow acceleration of the wheel in the original direction until it comes to rest again (ignoring overshoot due to underdamping at the moment). Now the most basic theory would say you've only had one random input and that there is no real frequency to be found there bar the single natural frequency of the suspension.
However, if you consider each segment of that single motion separately, you could calculate a few effective frequencies of each of the wheel and the sprung mass based on the different damping ratios generated by the different speeds the damper is seeing in each direction. What this means is that you effectively get a high damped natural frequency (HSC has the lowest damping coefficient of any of the 4 parts of the damper curve) when you first hit the kerb, but a lower natural frequency response in say the rebound part of the stroke where the damping coefficient is significantly higher. Of course this is a simplification ignoring the motion of the rider etc, that goes without saying.
As for the thesis, PM me your email address, I don't have a clue where (or if?) it's published but I can send you a copy.
Last edited:
What PSDs do is let you compare the various frequencies between bikes; kind of like comparing damper data. When you know what damping coefficients you need for LSC, HSC, LSR, HSR etc then you can compare that between one damper/bike and another. When you know the resonant frequencies of the bike, you can start to understand how different adjustments affect each one, and what effect that will have on the ride, but like I said, it's very implicit. This is an area I am still working on, you might start to see some results in 12 months time or so
One thing worth considering when it comes to the difference (in resonance) between adding compression and adding rebound damping is the concept of "effective" frequency. When you say ride your bike down the road and hit a kerb, the bike goes from no motion, to very high acceleration of the wheel, to high velocity of the wheel, to a lower acceleration of the wheel in the opposite direction, to a moderate (rebound) velocity of the wheel, to a relatively slow acceleration of the wheel in the original direction until it comes to rest again (ignoring overshoot due to underdamping at the moment). If you consider each segment of that single motion separately, you could calculate a few effective frequencies of each of the wheel and the sprung mass based on the different damping ratios generated by the different speeds the damper is seeing in each direction. What this means is that you effectively get a high damped natural frequency (HSC has the lowest damping coefficient of any of the 4 parts of the damper curve) when you first hit the kerb, but a lower natural frequency response in say the rebound part of the stroke where the damping coefficient is significantly higher. Of course this is a simplification ignoring the motion of the rider etc, that goes without saying.
As for the thesis, PM me your email address, I don't have a clue where (or if?) it's published but I can send you a copy.
I'm so sorry I brought this thread up.
haha don't be. I find it interesting stuff to discuss, it lets me put some of my ideas out there for people to think about and criticise/discuss, which often helps me clarify my own mental understandings of this stuff, which is and hopefully always will be a work in progress. There are a lot of smart guys on this site with a lot of useful/interesting things to say, and I love hearing their input.
And for what it's worth, for those guys out there who find faults in my logic/claims or arguments, please don't be afraid to call me (or anyone else for that matter) out on it just cos I act like I know what I'm on about
Nobody knows it all, particularly in this field, and there's always something new to be learnt - so if something pops into your head that seems contradictory to the status quo or what someone else is saying, spit it out so we can consider another point of view.Edit: btw for those interested in reading my thesis, I uploaded it here: http://www.filedropper.com/thesisdraft01backup071008
Let me know if the link doesn't work or whatever, I've never used file hosting websites like that.
Last edited:
106 pages of 'Final Year Project' ? 
Or is it how you call your PhD thesis ?
Anyway, it'll be interesting reading this evening.
Or is it how you call your PhD thesis ?Anyway, it'll be interesting reading this evening.
It's an undergraduate thesis, I didn't do a PhD (I'd still be doing it!). I had to cut it down pretty heavily to fit in all the relevant stuff along with all the irrelevant stuff that my lecturers wanted to see.106 pages of 'Final Year Project' ?
Anyway, it'll be interesting reading this evening.
Last edited:
I wish I could read through all this... we need an executive summary.
"pros ride firmer springs and damper settings than amateurs, but prefer soft setups for dicking around"
...I think...