Well the tech rumors thread keeps getting derailed and I like kinematics as much as anyone else so figured I’d start a thread on it.
First off, there is some personal preference here and there certainly isn’t a one size fits all solution to bike suspension. So no one is necessarily wrong, they just might like their bike to behave differently. For example I don’t like bikes with low anti rise because I find the rear extends too much under super heavy braking or creeping into steep chutes for my liking. But I know a lot of people out there like how low anti rise behaves and that’s fine. We just want different things out of it.
The main thing everyone seems to be derailing the tech thread with right now is anti squat. There are two things that contribute to anti squat: how much the tensioned chain length is growing and force at the wheel relative to the IC (easier visualized with axle path).
The chain tension one is pretty straightforward. When under tension, your chain wants to be in the point of travel where the length of chain that is tensioned is the shortest. I would say that’s always very top of travel, but there could be a bike out there where it isn’t. How much this counter acts squat depends on how much the chain length grows. This is why bikes with high anti squat often have high pedal kick back. There is a way to have high anti squat and little to no pedal kick though which brings us to the next factor… Pedaling forces at the wheel relative to the IC.
Pedaling forces and the IC are easiest to visualize with axle path. When you pedal, not thinking about any chain stuff, your wheel just wants to move forwards relative to you and the front end of the bike. Essentially it wants to go to its furthest forward point in travel. Think about a unicycle. You pedal and the wheel wants to shoot out forwards from under you. This is how high pivot bikes can get low pedal kick and high anti squat. For the duration of travel where axle path is rearward, pedaling will make the wheel want to rise in travel regardless of whether the tensioned chain length is growing.
It's been a while since I've done suspension kinematic deep dives, and i might be totally wrong here but in the rumors thread, people were arguing anti-squat doesn't matter at 50% travel as you won't be pedaling there, but anti-squat CAN and often is a proxy for pedal kickback and chain force on the suspension. So if you have a high anti-squat at 50% it does matter as it means more pedal kickback from chain force during that portion of the travel, hence why suspension works much better without a chain on.
@CascadeComponents you would have actually figures of suspension forces, but I've often argued clutch forces aren't significant enough to affect suspension as the force of a clutch is so low compared to the forces compressing your suspension, is this a correct assumption?
Yeah clutch/spring force from the derailleur is next to nothing. It is noticeable though. However, I’ve definitely forgotten to hit the switch on a shimano derailleur after putting a wheel back on and the derailleur flying around is more noticeable than the extra bit of tug on the chain. Some of riding, especially at a higher level, is a bit of a head game so there is some argument that anything that’s annoying or distracting is worth trying to solve.
I would be intrigued to see if you could actually see the difference in telemetry from clutch on/off, or is just a placebo affect.
In one of Neko’s original videos he tested with data and using the OChain. The only difference he saw was in dynamic sag.
I found some of the distinctions that were being made in the tech thread to be a bit confusing and presumably pedantic. My understanding is that anti-squat is how much the chain tension you induce by pedalling limits suspension movement, while pedal kickback is how much chain tension caused by suspension movement feeds back into your feet. Other than the likely use of different gear ratios and body positions in each case, these are the same force, just in different contexts, and are inextricably linked. Am I misunderstanding something?
Kinda/kinda not. There are two entirely separate effects at play that can be described as AS, and bike designers account for both effects to give a total amount of AS. Only one of which having anything to do with the chain. Those are described in the op...
Kickback is basically the "equal and opposite reaction" part of the chain tension AS which is exceedingly simple: chain pulls wheel to extend suspension. Therefore wheel also pulls chain when it compresses suspension. Which it sounds like you have a solid understanding of.
The other type of AS is just inherent to the geometry of the suspension/wheel path. The higher the pivot, the more you get. It's "free", ie it doesn't come with pedal kickback, but unless you're raising your bb along with the pivot (or using a smaller rear wheel) then you're going to need an idler otherwise your chain tension AS and kickback will go insane as your main pivot moves away from your chain.
And I guess just to clarify, AS doesn't limit movement. It's working purely to extend the suspension. It's used to cancel out the natural squatting of the rear suspension under acceleration. Too much AS and your suspension will extend when you pedal and bob just like a bike with too little AS. The Norco range/sight from ~2014-2015 is a great example of that.
@Cascade,
Have you done much/are you happy to comment on any testing you've done regarding the tuning of AR on a bike, and how it interacts with suspension/shock setup?
I agree with you in that I prefer high AR bikes, but I find that a high AR bike will expose a lot of flaws in suspension setup... Slow or unbalanced damping, rampy air spring, excessive sag, all feels much crapper on a high AR bike.
I've designed a bike that I have yet to complete, but what I (think I) have learned from riding lots of different bikes is that I wanted to pair the high AR with a more progressive LR with a slight excess of travel, that I could run with a linear spring and balance the sag/dampers to sit fairly high over chop. Do you have any input here? I've noticed all of your links are adding progression and adding travel so if there's much more to add beyond "that's how I like it" then I'd be keen to see what you think.
Regarding the two modes of antisquat, the chain tension one is valid only for chain driven vehicles, so mostly bikes (excluding the ceramic speed shaft driven system) and motorcycles (again, excluding any shaft driven motorcycles, like some BMWs). Cars and the like still have antirise and antisquat geometry built into their suspension. Red Bull for example has an anti-dive front and anti-lift (rise) rear suspension geometry to ensure a stable aero platform to maximize the downforce over a larger operating window: https://www.the-race.com/formula-1/gary-anderson-the-potency-of-red-bul…
This is suspension geometry dependant only, there are no chain forces at play here.
Also worth mentioning is things like the Trust forks. They have significant anti-dive.
I spent a while riding a craftworks ENR high pivot bike with a Trust Shout on the front. The braking performance of that bike was fucking insane.
https://www.vitalmtb.com/community/AgrAde,35595/setup,40730
wow thats a fucking crazy bike, is there riding in Alabama that calls for something like that?
Gotta scare the yokels off somehow
I've found that on low anti rise bikes I like to set things up to be less supportive and sometimes, but not always, slower rebound. This comes down to getting it to not stand up under heavy braking. It's a balance, though, because too much in that direction and it feels really wallowy really quick. In my head it kind of works like a dynamic ride height shift that happens under braking. Lets say its 30% sag normally and under heavy braking becomes 15%. 15% might be too high in travel for my liking so would do some combination of backing off compression damping and/or softening the spring to get it more around 20/35. This is all terrain dependent too, though. On lower angle stuff it's not as critical.
As for progressive curves and a little more travel, you certainly can get it to pack up less. You'll find tuning easiest if you can get the leverage curve to be as consistent as possible. I find sag and rebound to be the most sensitive to small changes and then compression damping actually comes pretty easy. I think progressive bikes have gotten the reputation of being less predictable because of progressive bikes where all that progression happens quickly in one portion of travel.
Not to derail, but is there a good one source for finding the graphs of different bike's kinematics/data? Trying to compare bikes I have ridden to better understand what I'm feeling a like/don't like.
Thanks
https://linkagedesign.blogspot.com/
Google translate is your friend
Yeah thats pretty accurate, basically the amount of AS made up by the chain is what gives a bike its pedal kickback. A bike I notice this a lot in is the current Enduro - historically Specializeds FSR bikes (note: this is a spesh thing, not a horst link thing) have low anti squat & anti rise figures but with some recent models they have used the chain tension to increase AS and it results in higher pedal kickback but still has low AR numbers. High pivot bikes naturally have more anti squat so don't need much chain effect, in fact the chain effect is often reducing the anti squat under power.
The unicycle analogy is pretty good - the bulk of the AS is just weight transfer and where the mass is relative to the thing pushing it. If you are pushing below the mass it will tip backwards, (low anti squat) and if you push above it the mass will tip forwards. When a unicyclist pedals they lean forwards and in some ways you could view this as naturally settling in a 100% AS position. (I was never very good at riding a uni)
The Craftworks & trust fork is sick! the Shout was a super interesting product - I rode it for a while and the anti dive thing was super cool and really inspired confidence! I can see why it needed a super flat wheelpath but it came with too many drawbacks in terms of small bump compliance
On the derailleur clutches - I spent a while looking in to it a couple of years ago and in some cases there is a measureable force there, eg a high pivot bike in the smallest cog might have something like 0.5-2kg force at the wheel which doesn't sound like a lot but when it comes to grip, any excess friction in the system puts extra load on the tyres which unload and reduce the force at the contact patch. If you're at the edge of grip its the kind of thing that can be the difference between staying hooked up and sliding. It's something I want to go back and test on the trail but it would need accelerometers and/or load cells on the hub to get meaningful data - position sensors alone struggle to detect things like that! Shimano clutches can have quite high stiction - most of them of the have a torx head you can measure the clutch torque with, the measure the perpendicular distance from the pivot to the chain, and the distance from the chain to the main pivot/IC and you can work it out. I tried installing a spring in the chain to measure how much it stretched before activating (don't have a chain measuring strain gauge yet) which was interesting but not sure how accurate that was.
What are you considering "low" anti-rise to be? Like down around 30%, or the ~60% that's a lot more common? My last bike had AR ranging from 98%-80%, and the suspension always felt awful when you hit the brakes on repetitive square hits. O-Chain, super fast rebound, etc. - nothing made it feel good. My 2 bikes after that have a more moderate amount of AR (~60%) and feel much better in those scenarios, especially running fast rebound. And I'm actually able to run some compression damping on the fork too because the bike isn't always trying to hunker down in the rear.
I would call "low" as under 50% - 60-80% seems pretty common for real world values, its not something I've done a ton of testing with yet (working on it) though. Same way populaar anti squat figures tend to be 120-130% instead of the "ideal" 100%
I'm curious why you thought super fast rebound would help? There is a ton of factors involved with how it will feel under brakes, eg leverage ratio, type of shock and how much sag you run
Mostly because I was fighting to stop that packing up feeling - tried to get it to recover as fast as possible. Or another way of putting it, it made brake bumps feel bigger than they were. Single large hits (drops, jumps) were fine. The bike in questions LR went from ~2.82->2.1, so moderate progression. AS went from 117%-89%. I mostly ran it with an EXT Storia - tried many different spring rates, compression settings, etc. Granted the EXT rebound is always slow, so running it 1 click from full open isn't saying much.
I spent 2 years trying to sort out suspension on that bike but on both the bikes I replaced it with, I had them feeling awesome within a few rides.
Yeah I view low as below 50% as well. I’m partial to anti rise that starts on the higher end of the spectrum (100%-120%) falling to the normal range through travel. Essentially just keeps the rear from standing up past a certain point under hard braking.
Some comments:
The antisquat vs pedal kickback conversation is always interesting because I think some things get muddled when people assume that one depends on the other. The two are completely independent calculations that depend on some of the same input variables. In practice this means when you tune one the other is inevitably changed. However there are thought experiments where this is not the case (jackshaft or hub motor bike).
Antisquat:
This is not a unique to bicycles thing. Anti geometry is a vehicle dynamics term. I like the definition from Millikens Race Car Vehicle Dynamics (not my favourite dynamics textbook but a very popular one):
"...anti effect in suspensions is a term that actually described longitudinal to vertical force coupling between sprung and unsprung masses."
In plainer terms the horizontal forces at the wheel (both acceleration and braking) can have reaction forces that act in such a way as to push the sprung mass up or down. In cars the only horizontal force is the ground/tire interaction but in a bicycle (and many motorcycles) you also have a chain force (which is not technically a longitudinal force but is related to the acceleration force).
If you have a hub motor bike you still have a ground acceleration force but you do not have any chain tension. There will still be an antisquat value.
A good example of how the different forces contribute to antisquat is shown below.
Case 1: When the axle is inline with the main pivot only the "chainstay" link of the rear triangle is in compression. All horizontal forces are acted on by the pivot and there is no reaction force from the shock. The chain force is parallel to the chainstay and has influence.
Case 2: When the main pivot is raised above the axle there is now a moment (torque) in the rear triangle from the acceleration force. The shock must have a reaction force in order to balance out the force equation. The force at the axle effectively "pulls" the shock open. This reduces weight transfer induced squat.
Because of an (impossibly large) idler that is the same size as the rear cassette the chain force is once again parallel to the chainstay and has no influence. Antisquat is 16.8% in this case.
Case 3: Without the idler the chain forces now induce a moment on the rear triangle which also require a force from the shock to balance out. The antisquat value is now even higher because the chain force also "pulls" the shock open.
Fun fact (if by fun you mean so far down the nerd hole there is no coming back up): The chain force and the acceleration force have influences proportional to the total gear ratio. In this cause the sprocket radius is 64.7mm and the tire radius is 350mm giving a ratio of 5.4. The chain force is 5.4 times greater than the tire horizontal force and so increases the antisquat by 5.4 times the amount of just the tire force (16.8% from case 2). 16.8% + (5.4 * 16.8%) ~= 107%. This tire to sprocket ratio is why the chain influence on antisquat is SO much higher than the tire contact force. In smaller gears this difference becomes even more pronounced (but the chain angle influence starts to dominate).
With respect to Cascades otherwise great info I have to disagree that it is the chain length change that matters for antisquat. The length change is the driver of pedal kickback.
It is the chain angle that has an influence on antisquat (because it is the chain angle that dictates the force direction). If the chain angle is sloped down more (bigger rear cog) then generally the antisquat increases a bit. For high pivot bikes the main pivot sits above the chainline which inverts this relationship. This leads to excessively high antisquat in the lower gears (as well as very high anti-squat values overall) making an idler essentially mandatory.
Pedal Kickback
To cut my essay short I will just say pedal kickback is purely a function of tension-side chain length changes. There are three factors that are generally considered:
- Tangent length changes
- Wrap length differences around sprockets of different sizes as they move through space
- Rearward axle path causing the wheel to "wind" up the chain a bit (very small influence generally).
The wrap length and tangent length differences are highly dependent on chain angle. You see a strong correlation between antisquat and pedal kickback because antisquat is ALSO highly dependent on chain angle.
Antisquat does not drive pedal kickback and pedal kickback does not drive antisquat. They both are just highly dependent on chain angle.
Chain length change is just another way of saying if the chain is always wrapped around a point concentric with your pivot then the chain force is more or less reacted entirely at the pivot as opposed to creating a moment. Of course that is assuming you are in gears with equal teeth, which is only really third or fourth gear or something like that.
I'm not from Alabama lol. I'm not even from North America. I'm in New Zealand and there is definitely riding here that calls for the weirdo machine.
Both products were interesting in their own respects. The Craftworks was an incredible show of engineering and packaging. The bike's kinematics were on an absolute knife edge. The bike feels perfect, and balanced, and predictable, and amazing (though distinctly very-high-pivot), even if the kinematic graphs that linkage spits out look quite interesting. They all balance out quite well through the envelope of travel that you want the performance, but move a pivot around just a tiny bit in linkage and the whole stack of cards falls over. And it looked like such a normal bike - with a normal fork on it nobody would take a second glance at it. It was a really good looking basic two-triangle design. And I really loved the flex response of the frame. If Craftworks can come back from the whole fucked up hellscape of the last 5 years and come out with an ENR that uses a 29" front wheel and has improved seatpost insertion depth then it'll be one of the fastest things on two wheels.
The trust did some things well and some things very poorly. If you were the sort of person that got on well with it, It was the best thing ever. But it hated braking bumps and it hated low speed tech.
Thanks. Yeah I think we're on a pretty similar page.
I'd prefer the AR to fall at a higher rate than it does, but I think that it should be passable as is.
The high pivot VPP (or HPV Peepee linkage as i call it lol) moves the IC forward and slightly down as it compresses to get the falling AR and predictable AS but I couldn't really lower the AR without fucking up the axle path I wanted and the AS behavior - i wanted a flat 130 in the climbing gears and a slightly lower (at sag/dynamic ride height) but rising AS in the descending gears.
The LR ended up slightly more progressive than I wanted but it's a nice straight line that'll work well with linear air springs or coils.
PK i wanted under 5 degrees for the descending gears and ideally with a lower gradient in the beginning/mid stroke but really didn't care about the climbing gears. I like the feel of near-zero PK under braking when the wheel is starting to lose traction. Once the wheel locks, PK becomes major because the freehub engages. So on repeated bumps the rear wheel will break traction and lock, then on the next bump when it gains traction again, you get a spike through the pedals before the brake begins to slip. It feels rough and it messes with brake feel. I reckon that the "harsh under braking" feel that some might attribute to brake squat/high AR is actually just PK as the wheel breaks and gains traction.
Hopefully the graphs aren't shrunken and unreadable. Here goes...
Genuine question: Couldn't a freewheel chainring reduce or eliminate pedal kickback?
On cars, the force are from the axle or upright, trying to rotate the car, and the weight of the car, center of mass, limiting that rotating force. With grip and without gravity, the car would land on it's back.
You can see the anti-squat forces on offroad rigs, where they project the aftermarket suspension to rise the vehicle upon acceleration, giving it more traction and clearance.
High anti-rise makes the bike squat under braking forces. That puts you in quite a force on the spring, where it could feel harsh.
Low anti-rise buckles you to the front, where the suspension maintain it's ability to move with little force imputs. But you are being thrown to, or over, the handlebar.
Personally, I prefer high anti-squat, like single pivots.
Google OChain.
Google WRP chainring as well
yes in some scenario’s where is ratchet is free to move, but if it’s not in the perfect part of the freewheel stage, you might only have 1 or 2 degree of movement. Also if you are locking up the brakes there is no freewheeling.
With low pedal kick or idler design, it’s consistent. Or an Ochain is always helping you.
Genuine Question 2: Has anyone ridden an O-chain and a freewheel crank (WRP/Intend) back to back?
Genuine Question 3: If you've ridden either, on a WC style track (i.e. fast & no slow jank), is the effect of those products felt on the whole run or only felt in braking?
For me, I have a hard time believing those products actually produce a significant kinematic change to the bike. (Important: I'm not say they don't change the feel of the bike.) In my view they simply isolate the chain slap tugging on the pedals which intern makes the bike feel smoother. If there's less noise at the contact point, the rider can have a better feel for the grip at the tires. FYI, I do have an O-chain waiting to install and try, so I'm not knocking these products just curious of the group's perspective.
Post a reply to: Kinematics