Vertical Frame Compliance Is A LIE??
HTML-код
- Опубликовано: 24 авг 2024
- A semi-rigorous study on vertical vs lateral compliance in the rear triangle of a hardtail mountain bike frame. This is not conclusive evidence, but a compelling finding that suggests it's the lateral flex that we interpret as vertical compliance in a frame.
Keep it constructive in the comments.
Support the channel!
www.thebikesau...
![[DOKKAN BATTLE] Worldwide Campaign Announcement Video](http://i.ytimg.com/vi/zxfEiMksZQs/mqdefault.jpg)
The most techncical review on a steel hardtail! This video tickles an exciting intersection between art and engineering. You can do all the experiments you want, but if the bike does not feel good, you won't enjoy the ride. Conversely, you may love how a bike rides, but none of your analysis can tell you why. In this instance, I am glad the subjective feeling, empirical data, and theory align!
Thx for making kick-a$$ frames, Daniel!
DY for MTB Pres
🎉
My nose bled after all the data, it was fascinating. Better than any GCN tech vids 😎💯👍 AWESOME! 💯
😂😂😂 apologies
@@TheBikeSauce Hello,
I have a questions on the data:
Were the accelerations measured absolute?
Can they measure in relation to one another or be substracted post processing?
Wouldn't it be more representative of frame compliance to measure rear axle movements in relation to BB?
So for marketing, a frame is vertically compliant. But it real life it might actually be horizontal compliance at a specific frequency depending on your setup, speed, and terrain. What the data shows vs what we perceive while riding is amazing
I'm totally here for that type of content !
Working on the Mechanical vibrations and acoustic field, your analysis is top notch and well "vulgarized" for newbies, well done buddy !
Matlab is where I spend my time at work and this video speaks to me! As more data comes to light, I wonder if the square taper bottom bracket might get renewed attention as a source of riding comfort given it's supposed higher flex. Analog devices makes mems accelerometers with PDM output (for all purposes, analog) that I've used in some past experiments, including as a very crude bandlimited acoustic guitar pickup. They are good to around 6k . I imagine making a test rig using a handful of these with a multi channel audio recorder like those zoom to facilitate gathering raw accel data for multiple locations in the same run. Anyway, know that the science heavy work is most appreciated even if it doesn't have high appeal to a broad audience. Cheers!
Haha I’m surprised at the engagement so far, but yea. Super niche video
Agreed. It doesn't matter what level of ride comfort your BB shell is experiencing. What matters is what the rider perceives and most MTB riders are on super stiff aluminum or carbon cranks with hugely oversized, hollow spindles and have been switching to shorter length cranks en mass which are even stiffer.
Riding BMX, no suspension obviously, the impact difference in feel between tubular chromoly cranks VS forged aluminum cranks is obvious. I mean just look at the cross section of a round steel tube VS a solid aluminum block. If everyone wants their rides to be "buttery" maybe ditch the ultra-stiff cranks?
I appreciate the scientific approach. It's frustrating to hear people say that one frame/handlebar/wheelset is more compliant, when they have totally different suspensions, tires, etc.
The language used to describe bikes is imo one of the most BS ridden ones of alle time. It is practically esoteric without any self reflection. I think, you might find more hardcore belief than in church service.
@@thomasspotzl4240 "this suspension fork is so buttery!" without providing any data from a shock dyno
It would be interesting to plot points at the bottom bracket, top of the seat tube, and at the saddle clamp to see if there's a big difference between the 2 frames as well for that. Vertical compliance while seated is more a function of how much your seatpost can bend backwards rather than through the frame, right?
Though, since you were on your mtb, I'm guessing you were out of the saddle more, so you would be feeling compliance through the pedals. So the lateral flex might make sense for that since cranks and pedals are more like levers on a frame, right? I'm not an engineer at all, so I'm probably way off base on everything. : )
Your feet being off center from the fore aft axis of the bike, I think the compliance you feel at the pedals might actually be torsional compliance about the bottom bracket. Might want to look at the rotational data about the fore aft axis. OR just measure vertical accel at the end of the pedals.
Also, looking into the frequency domain, my feeling is that a "compliant" frame should look more like a low pass filter (more low frequency vibration, and less high frequency).
Definitely. Aside from the Dc gains, the plots do reflect a LPF or maybe more appropriately a bandpass filter
Huge shout-out for your concise explanation on your method of data analysis. Clearly a complex topic I can barely comprehend, but you did an excellent job of making the topic approachable and digestible. You are providing a style of content that is completely unmatched in the scene of outdoor gear reviews and many science focused channels could stand to catch up to your quality. Chapeau
This is the kind of video I want to see! This type of data could be used, for example, to test how much more compliant some wheels are versus others by mounting the sensor at the rear dropout like you mentioned. For example, I’d be really interested in testing a bird spoked wheel versus a standard steel spoked one.
For the analysis and visualization of the data, it would be really useful to compare one run of each bike against the other run from that same bike. That would give us an analysis of how much standard variation there is in the data or if we’re getting consistent results. If the results are consistent, then for visualization it would be useful to average the two runs together so that we’re only looking at two lines instead of four.
Thank you so much for publishing this video and great work! I know how time consuming the testing and component swap is, not even to mention the time of condensing it all into a video and editing it as well as you did.
Such a beautiful video. I didn't understand a word, but enough for me to subscribe to the channel. You'll never know, I might even learn something...You are one of the people, who make the world a better place to live. Thank you!
Hey Nolan, as a fellow Mech, in the absence of building a test stand to directly + repeatedly measure the actual deflection, I wonder what kind of results you would get by adhering strain gauges to the back and chain stays and rerun the same test. I’ve not had to deal with as noisy a set of data as that would be first hand, but would love to see you try😂. (Maybe some student labor can help figure it out)
Anyway, I’m the kind of nerd who would love to see more PSDs in my bike content, so thank you for going all in on the actual engineering experimentation!
I’ve been toying with the idea of mixing the channel here and my cycling research group on campus, but have to work through some details
Very interesting methodology. What I think is missing is the measurement performed at the top end of the seat tube. That's where people try to maximize compliance. The bottom brackets are always designed to have very little vertical compliance so that your power when climbing isn't wasted on the compliant frame. So these results do not surprise me. They even always market themselves as having stiffer bottom brackets.
To me, ride quality is the exact opposite. The biggest impacts never go through your saddle, it all goes through the feet. If a bike is too stiff a the BB, it will corner poorly and ride harshly. This is why you see world cup downhill racers trying to reduce the lateral stiffness of their frames and wheels. Companies always say "laterally stiffer, vertically complaint" because that is what consumers expect them to say. In reality, those companies have de-tuned the stiffness of their bikes the last 5 years.
Perhaps on a bike where I’m seated during the test this would make sense. In this video, data was only collected while descending and standing on the pedals
This video rules. I appreciate the rigor you put this experiment through and the effort made to translate it into layperson's terms. Kudos!
I greatly appreciate your work here. It is so important to show what is actually going on. Would be neat if you'd contrast level of variance between frames an 2-5 psi tire pressure difference to show flow important tires are compared to frame choice.
Also your results This track with an experiment we did at the university years back but we had to use students in our group and mask the frame to do some level of blinding. We tested road , trekking and MTB frame with same wheels etc and we basically found it was all a wash they were not cyclists but even I failed to identify what was more compliant but it wasn't the ~33% spread as with others... But it was still sobering how much expectations played a role in perception of comfort/ compliance. Runs were on cracked pavement and cobbles in 2000s uni didn't want to give us the expensive accelerometer dataloggers and the were being used in seat dampening tuning for trucks.
That makes sense because we don’t often hit bumps and rocks straight on so that lateral compliance is what we feel. Tube shapes and wall thickness is probably more important than material or other geometric aspects. And when seated you get a lot of flex from the seat post and saddle. Plus tire pressures have a huge effect.
Yea true.
Rigidity is a combination of tubing diameter, wall thickness, geometry and material. There are significant differences in rigidity between aluminum, titanium and steel. Aluminum is about 1/3 of the density of steel but 1/3 of the rigidity as well. Titanium is 55% of the density and 55% of the rigidity. Different alloys of steel, aluminum and titanium have between zero and negligible differences in rigidity of dissimilar alloys of the same base metal. An interesting study was conducted and published on rigid fork compliance in "Bike Tech" volume 3, number 4, August 1984 comparing forks made with 4 different types of fork legs 3 types of Reynolds 531 and another in Columbus SL. They compared different head tube angles (71°- 75° H.T.A.), total fork rake (38-62mm), fork rake radius and constant trail for vertical compliance. A 71° H.T.A. had nearly double the vertical compliance of a 75° H.T.A. - The results were surprising. I don't know of any other specific studies which tested a variety of different designs to compare vertical compliance but am convinced that a number of others are likely to exist. It would be interesting to see the results of a variety of different frames tested for lateral rigidity and comfort, geometry, features such as curved seatstays... for comfort according to a testing machine measuring vertical compliance as well as actual riding (which is subjective).
It would appear to agree with the experiences of both Jan Heine and Grant Petersen, whose bikes have a fair amount of what I think of as "torsional compliance" around the BB. One can see, e.g. a Rivendell Rambouillet wind up several degrees clockwise then counter-clockwise as viewed from behind, and then spring back to the central position as one pedals hard. The feel is one of getting in sync with the frame, and it is not a slow bike despite having the traits so many people associate with being 'slow'. Also, the give makes it a lot less hard on the knees, and one doesn't feel like a sharp bottoming out of the pedals. It's kind of hard to put into words that people are used to, so it tends to get discounted.
This is my thinking as well. Most of the movement in the rear end would happen around the axis approximately formed by the center of the BB and center of the rear hub. And the rear triangle will rotate depending on which side of the rear triangle in imparted with more force. So relative to the rear hub, the BB does twist.
This video is awesome, Nolan! So great to see your data and hear you explain it all so well. 👏🏼
9:40 wouldn't you expect lower vibration at the bottom bracket simply due to the fact that it's midway between the two wheels? When you hit a bump, the vertical travel measured at the axle of the wheel is going to be about as much as the size of the bump, but midway between the two wheels it's only going to be about half as much.
Right. Agreed
@@TheBikeSauce YAY! Not too bad for a software engineer.
Really well done. Your rigorous approach and transparency on the methodology are great!
Great video. Fantastic to see you applying this kind of rigor to bicycles. I first learned of fast fourier transforms 30 years ago, in a signal analysis class at college. The concept blew my mind then, and still does. 'if you make more videos like this, I for one will definitely be watching. Thanks.
As someone who studied art and cultural history and works with fiction writers for a living, the details were nearly impenetrable to me. As someone with a layman’s interest in science and engineering, the broad stroke observations and conclusions were fascinating. Thanks!
This was a very fun and informative video
Im glad to have someone actually take measurements to validate claims of compliance - i have always heard conflicting arguments about stiffness and material choice being essentially nullified by structural design, which had always felt logical. Its great to start taking the data to investigate why we think we feel what we do. Id love to see this data centric mindset grow in the community
Wow Nolan, fantastic video! I heard you discuss it with Russ the other day and was hoping it would drop soon. I'm no engineer, but it seems doubtful that there is much vertical compliance (as you found, and for the reasons you already stated). Somewhat similar type of confusion re: what we feel vs the cause w/tires. Generations of riders were certain that skinny tires are faster because they 'feel' faster because people felt the road buzz. But the evidence continues to mount that this feeling is wrong.
Excellent analysis. I’m riding the same bike on technical trails in reunion island. Tested with both 29’’ wheels and 27.5 ‘’ plus wheels, and a long experience in hardtails. With 27.5 plus wheels (3’’ at 0,8 bar in front and 2.8 ‘’ at 1 bar in rear), i can’t return to 29’’ wheels. Only if you’re a chrono hunter, the plus format on hardtails is more fun and definitely more comfy;
As always , another great video. Very interesting results , looking forward to more of these type of videos from you. You are one very intelligent guy!!
Hi Nolan! Im no engineer, but props to dumbing down all the information, and doing all the experiments and lastly moving all the kona parts to the neuhaus!
Arm chair here: This is great. I'd love to see a bigger replicate set with runs from people who don't have a pre-formed opinion on the bikes. Maybe you could have students do it. And as an engineer, how could you try to find frame properties that correspond to your result? Is it a steel v aluminum question? Or the specific steel and aluminum alloys used? The tube properties? The welds? How would we know. Given that the wisdom on bike tires has gone 180 since I started riding , vertical compliance being a myth would not surprise me. But why would lateral compliance be perceived as more forgiving? Is it that energy is out of the frame before it gets to your hands and rear? I hope this video sparks a good conversation.
Fantastic analysis, nicely done! Fits with some observations published by Alee Denham of Cyclingabout. For further nerdy analyses, seek out "Why It's Impossible For Steel Frames To Be More Comfortable Than Aluminium".
Great work! One thing to say about the vertical vibration being lower at the bottom bracket than the rear axle for both bikes is that this doesn't mean both bikes have some measurable vertical compliance. This could be due to the BB accelerometer being closer to the fork and further from the rear axle. The amplitude of vibration will be less for purely geometric reasons (like sitting closer to the middle of a seesaw) even if both frames are effectively rigid in the vertical plane.
Totally agree. Also there’s a giant damper on the pedals (a person)
Thanks for the effort to share your results. I too am a keen hardtail rider. I recently discovered the torsional stiffness of the crankset had a very noticeable effect on how jarring a small landing was at my ankles. I became so interested in perceived differences that I developed a way to measure twisting give of all my crankset sets. I now use an older deore crank on my Ibis DV9. The turbine cranks with a 30mm spindle were least compliant. I now will check the side to side stiffness at the bb of a carbon DV9 and a heavy steel frame!
Fascinating stuff, really enjoy the science videos. Riding a steel rigid bike with 270 lbs between me and my child on a MacRide I can see their head bobbing up and down and feel the flex standing pedaling up a hill. Not really sure which way the frame is actually flexing, but is is significant.
I would think measuring the vibrations at the rear axle and TOP of the seat tube would show more of what you feel (since you were presumably sitting on the seat). I have an old Cervelo RS from 2010 and the seat stays are pencil-thin and not even straight tubes - they are purposely arc-shaped to allow them to flex and minimize the vibration felt in the seat. The chain stays are stiff and straight to maintain their shape when power is applied to the cranks and keep the bike frame from deforming just from pedal forces. Curious if you ran those trails again and took the data from the top of the seat tube what differences you might see.
Also, there is likely a lot of variation in each run because your wheel didn’t travel over the exact same path each time. To help better understand the “input” vs “output” vibration, you should consider dividing the data (point by point - easy with Marla) to reveal the transfer function. You may need to offset the data in the time domain a bit to account for propagation delay?
Definitely for a road or gravel bike, but descending on the MTB, I’m usually standing on the pedals
The argument about different lines on each run is valid, but again that’s why I took a frequency domain approach. As long as the PSDs show the same frequency content, the exact lines don’t matter as much.
@@TheBikeSauceah, ok. Then I understand why you measured at the bottom bracket. Great video, and nice job as always. 👍🏻
@@TheBikeSauceI get it regarding the value of PSD, although you have so few runs and there is likely pretty large variability in each run. If you did a statistical analysis on the variability of PSD in the two runs on one bike, and figure out 2 or 3 sigma variation, you might find that it overlaps with the other bike. So is there really a statistically significant difference?
Don’t get me wrong - PSD is interesting to look at, but in my experience, I think you could make an even stronger conclusion if you calculated velocities and displacements from your accel data, and compared how each impact event at the rear axle manifests itself at the bottom bracket throughout your ride. I come from a field where we were always very limited in the number of runs/test we could make, too (airbag systems for passengers cars - each test destroys a $250k prototype vehicle). Our systems essentially calculate how much the vehicle is deforming using a similar setup: accelerometers at the vehicle periphery, compared to accelerometers in the cabin. The systems estimate the displacement at each location (double integration of the accelerometer signal), for each data sample, versus time, and compare the displacement curves against each other. That was more meaningful than PSD in our case; we could actually see the deformation of the vehicle happening as the displacement curves separated while the impact event progressed. Granted, we are looking for gross/plastic deformation and you are in the elastic region, but those integrations also really smooth out the data like PSD, and would let you compare input versus response more directly IMO.
Nicely done! Going in, I had reservations on the data because you first only show the accelerometer at the axle, and I was like... what??? But you explain this and the outcome of your experiment is very interesting. Thanks for doing this and your well articulated explanation of the data! I have designed a LOT of bikes in my 35+ year career in the bike industry, and we generally quantify vertical compliance at the saddle. Because with road and gravel bikes, this is the place where vertical compliance matters most, this is where it's done. For a lot of obvious reasons, this is not as relevant to a mountain bike, but it's still not wholly irrelevant either. When testing a frame for full ISO testing, there is a vertical fatigue test, but this test can of course be run with a single force load into the 'saddle' with a very measurable amount of deflection. The reason this is still somewhat relevant to a hardtail mountain bike is that the seatpost is typically an 'infinitely stiff' steel bar (nothing is infinitely stiff, of course, but for the purposes of the test, the frame is doing all of the flexing and this tester post is isolating the frame vs. allowing flex in the seatpost as well). This denotes some amount of vertical compliance through the frame, not at the bottom bracket, but rather through the whole frame itself, from axle to axle. I think compliance is more a factor of how the BB hangs between the two axles, like a hammock or skateboard. When you're on a long board, your damping is coming from the flex in the deck, not the wheels or trucks. I wonder about a couple of things: what would the data set look like if you did the same exact test but tried to isolate the frame even more by running 30 psi in the tires instead of 20? I also wonder how some strain gauges running along the top of the chain stays and/or seat stays would look in comparison. Anyway, great stuff!
It would be pretty cool to see all hardtails spec'd out with a "tested" lateral frame compliance number. Though it would be a challenge to standardize the test
Great video thanks for not dialing it back!
Really cool video! That accelerometer setup is awesome! It would be super interesting if you got a bolt on seat stay brace made up. Then out of curiosity see what the runs on the new house look like. I remember when I made a hard tail back in school the rear end stiffened up a lot after welding the chain stay and seat stay braces in. Ive has allot of questions in this realm recently trying to explain why certain bikes I have ride way better than others. Keep this type of content coming🤘
That was bad ass, Nolan. Super engaging video. Thanks!
You sir would be overqualified for the dilettante of the year. Kudos.
I see some parallels to work done on stiffness of wheels. Guy at northwestern showed that it was rim torsional stiffness that was the main contributor to wheel lateral stiffness. Basically, box section rims deform locally at the contact point because the rims have low torsional stiffness and it has very little to do with rim lateral or radial stiffness. Make the rim a bit deeper and the wheel starts to get rapidly stiffer laterally. ❤I certainly have a set of tubular box section rims that are my first choice for roubaix type races. I find them to be very surefooted and comfortable. They appear to deform to the ruts by deforming laterally and finding grip rather than bouncing off or deforming vertically. I would say that this is the same for frames. They are more compliant because thye deform side to side because of torsional stiffness reasons. You can get too much not his though, I had a cross frame with a lot of rear wheel wag that was an amazing descender on rough surfaces but horrible when sprinting or climbing as it would wag all over the place under power and was very unpredictable when cornering and would lose traction randomly.
Very cool,
Thanks,
You should share this with Steve from hard tail party 👍
This is great. I`d love to see more of such rigorous experiments. Subscribed.
Absolutely loved this video, definitely here for more of it. Peak Torque might have some thoughts on this too--he's made some videos in the past outlying the engineering behind "vertical compliance" and why it's fake.
Putting the accelerometer around the midpoint of the seat tube and underneath the saddle would be very interesting too--maybe the Neuhaus reduces vibrations even further by those points.
This might also explain rear chainstay shapes as well. As most of the time they are ovalized vertically. In part for tire and chainring clearance, but this also should help with torsional flex instead instead of horizontal flex.
Wishbone vs regular seat-stays?
Longer seat/ chain stays on a laterally rigid frame.
Great work, btw👍👍👍
As I see it, frame compliance is far more significant in a road bike than a gravel or mountain bike.
If a frame has a few mm of compliance that would be a much bigger proportion of the compliance in a road tyre compared to a mtb tyre.
It would be really interesting to see a similar exercise comparing a stiff road frame to a compliant road frame.
Good video you deserve more subs, constructively I’d suggest bonding any accelerometer to the surface area of measurement. For example directly to head of the rear axle using an epoxy not so easy to remove post test, but would guarantee you don’t isolate some of the harmonics like the cable tie setup. Oh and give that cap to Dr ‘Ouchy he needs it more than you!
I feel like non-marketing engineers have been saying for years that vertical compliance just isn't a thing. There's certainly a subjective difference between bikes, but I've never seen any evidence of vertical compliance being meaningfully different between unsuspended bike frames. Wheels and tyres, sure. (And even with wheels, there's been talk over the last few years that maybe a laterally stiff wheel isn't the best for control OR comfort, particularly while mountain biking.)
It’s great to see some actual data driven analysis on here! I’m an ee working on different kinds of vibrations, the thing that I would be interested in comparing would be data for a smaller section, ie a step change from one height to another, to see how that was propagated through the frame. Because this data is an accumulation of a longer period, and compliance is something you feel at particular points, ie a stiff frame and a compliant one would feel relatively the same on a smooth section, you’d only feel the compliance at particular instantaneous points. I think this steady state data may be masking more interesting things that occur instantaneously when you hit something.
This makes a lot of sense given the structure of a typical diamond-shaped bicycle frame. I wonder if there'd be any difference for bikes that have the seattube more-or-less decoupled from the toptube/seatstay junction, like the Trek isospeed frames.
Really interesting experiment. Looking forward to what comes next.
Others may have asked already, but I would be curious to see data taken from the rear axle and seat collar. While one is generally only occasionally in the saddle on a mtb the data would be of interest for gravel/road frames. Cheers, great video
Nice experiment. I always wondered about compliance but I never really could tell differences. Thanks for taking the time to help us understand this a bit better.
Great vid!
Excellent video. Now do a carbon frame vs. aluminum! Just want to see how much they differ.
This was awesome Nolan! It would be interesting to see what differences one might find in material, e.g. steel vs. titanium. Might be able to help with that soon if you're interested.
BTW, love the idea of mixing your bicycle engineering program work into the channel.
Super cool expt. Appreciate the rear axle control. I might have placed the sensors at the contact points (saddle, handlebar/stem, cranks) instead of the BB, but i understand your reasoning. The other thing that comes to mind is you mentioned the test runs taking place on a downhill stretch - I assume you've corrected for that (or that it's insignificant enough to ignore) when looking at the Z axis data.
I have so many questions for wheel makers now.
Zipp Zero Moto rim noises
In an earlier life I worked on defense electronics. We used a pretty burly vibration table as part of the qualification for things like vibration sensitivity of VCO’s, and I have really wished something like it could be used to show the actual frequency response of a bike frame. The drive was a pseudo white noise (calibrated via the on-platten accelerometer), but it could be setup for other stimuli. An accelerometer on the platten surface would provide the denominator, while a frame mounted accelerometer on the three touch points could provide the numerators so that dampening or amplification of vibration vs. frequency could be plotted. It would make very light work of comparing tires, frames, seat posts and handlebars.
That’s the holy grail! If only we had the funding..
@@TheBikeSauce For all the supposed R&D spending in the bike industry it is pretty shocking how much sciency sounding anecdote is passed off as data. I am guessing that a vibration table is dirt cheap compared to a wind tunnel, but clearly frame comfort is not worth quantifying or optimizing in the industry. Your approach here is laudable and scrappy, good work.
Yes yes yes finally a voice of reason!
Try it out with the cheapest crankset with the most lateral flex you can find to see how it feels.
(or modify/make a pair with a flat flexy section in the middle of the arm)
You could actually take this one step further to prove the exact same thing for Steel vs Alloy vs Titanium.
You'll likely return with the same exact results as the triangle/truss structure is not designed to flex.
This may hurt the egos of all the "Steel is real" or "Titanium is the best" crowd.
Anecdotally, I had the Honzo frame for a short time and sold it very shortly after due to the harshness.
I replaced it with a Santa Cruz Chameleon in alloy which rode a bit less harsh, and was lighter.
CyclingAbout also did a YT video on this content, and included the difference between metals, but his work was based on the research of others.
🤯 Outstanding! I like the way you explained everything. The science hat was perfect. Thank you. Keep the great content coming. Ride safe and have fun. I think I know how, when & why my last frame cracked.
Hey Nolan, great video. As a bike rider myself and a mechanical engineer here goes a question for you. Do you think that the chain stay and seat stay flexing outwards from the frame's vertical plane of symmetry would result in the compliance felt by the riders of a hardtail? Maybe a FEM model of the two frames with their particular geometrical features would give this answer. Thanks a lot and keep posting videos like that!
I’d like to the data when you place the accelerometers closer to the touchpoints of each bike, e.g., at the top of the seat tube and headset (so the flex of the different seatposts and stems/handlebars aren’t captured).
Compliance means bend or flex, right?
Do we want to measure flex at the handlebars, pedals or seat?
Do the tires provide a lot more flex than a frame?
the tires, vertically, 100% yes. Horizontally.... hmm..... the rims also have to be considered, those flex laterally too...
Yea everything flexes to some degree. It’s why this type of vid is tough to make. Im making no sweeping claims here 😀
Great video! I could wish that my ME profs were as interesting as you, but in fact, they were! (ME, UC Davis).
Add me to those who'd really like to see similar tests comparing rear axle to seat post and front axle to handlebars.
Would be interesting to see this experiment conducted on a technical, chunky, hard hitting trail to put the rear triangles under much higher stress and force.
I will say your explanation of lateral vs. vertical compliance makes a lot of sense given the general construction and layout of tubes on the rear triangle (Also noticed the Neuhaus has no bridges on either stays, which perhaps helps?). I've stretched rear chain stays of older bikes in the past to fit slightly wider hubs in the past so definitely more give on the lateral plane.
Very interesting indeed. Nice one.
I thought about, my the data would likely just rail / saturate the whole time rendering it useless
the frame flex in the front has a signifikant impact to the ridefeeling - if you mount a rigid fork and measure the equivalent to the break stiffness - there may be e certain difference between the two
This is great stuff! It sounds like your observations support the historical description of frames "planing".
I would love to see this type of setup measure other things--seatpost flex, handlebar/stem flex etc.
Outside of bikes I'd be really interested to see if this type of equipment can answer whether wood or fiberglass tool handles transmit vibration to users differently.
Great work! Impressive how you were able to tease out data to reflect that difference in perceived comfort. I would love to hear a followup on how that feeling of greater BB compliance impacts subjective feeling when pedaling hard like on a climb.
I would be interested in a comparison with a well designed CFRP frame. It may be that the isotropic nature of alloys is very limiting in the differential stiffness that can be achieved; IE you can't uncouple the vertical and horizontal deflections very much with that material selection. Increased stiffness in one area will likely come with an increase elsewhere, and vice versa.
Wow! That was so interesting! Would be cool to know how much would some of the other parts affect this results to have something to relate.
Thanks for doing this, great job on figuring out a way to test. Given the constraints of the rear triangle not offering vertical moment, I kinda assumed when I think I can feel a ride quality difference in a bike there were several possibilities. Once is as you tested, compliance in another direction. I have also though that there might be flex at various major junctures, bb, headtube that would diffuse the trail impacts to feet/hands/ass. I don't know how easy this is the to test. The last thought is that various tubes absorb vibration differently. In this case, you could measure at the drop out, then zip tie 2/3rds the way up the seat stay and see if there was a difference. Anywho, great stuff, thanks for taking the time to run through this widespread concept. I know want to know how the bike industry measures this or if consumers just get the buzz words.
I loved the experiment design, especially adding the control at the rear axel. It made the data so much more definitive.
Long seatpost length, seat/chain stay thickness, tire volume help with vertical compliance.
Really cool experiment but "compliance" AKA flex in a frame is only "complying" to the forces being put through the tires and wheels and then those reduced forces are also being attenuated, a little or a lot, by your crank arms, spindle, suspension fork, bars and grips. And easy next step would be to change tires which will move the frequency of the vibration. Or change the PSI. What is the amplitude of a 10 PSI change VS buying a new frame?
I wonder if you would get more "signal" intensity if the front end was not damped. Possibly try 2 bikes with forks with same spec.s (length, rake, A2C) but use a 3rd, but different brand, fork with same spec.s and then compare the 2 frames.
Thanks for being the consumate scientist in stating that you could be wrong. But also welcome test results from other independent experimenters. This is science, and science welcomes criticism- positive and negative.
I love this video! ❤❤❤I wish my teachers would have used examples like this, maybe I would have retained more info from school. Your students are truly lucky.
Was there no difference in the fore-aft frequency?
You’re giving me bad memories of college Fourier analysis…. I survived it…and still don’t fully understand it.
Awesome video! So what about bridge tubes? I had come to the conclusion they had no purpose other than to mount fenders or rim brakes, but they might affect the lateral rigidity of the rear triangle.
A frame is built like a truss, which by design is made to resist flex or comliance.
You really meant it and didn't hold back when you put on that science cap on
Doesn't this only count of you ride “exactly” the same lines on both bikes? There are to much variables to take into account. Like rider fatigue (line choice), state of the soil (wet or etc.). Fun experiment though. I think frame compliance can be better measured in a lab, they have tools to measure material fatigue for instance. Those tests could also give an inside in compliance/flex.
To some extent, multiple runs helps alleviate those issues, but the soil condition could be a large factor, I think. This is way more rigorous than most RUclips experiments, though I do think a few more runs each would be worthwhile given the time commitment to switch parts over. It's difficult to conclude anything in detail, but I think there's enough good data here to raise good questions worth investigating further.
I don't think lab tests would replicate the frequency inputs of real world trails, though. Ideally, he'd be doing a lot more back and forth swapping but there are time and budget limits, a key feature of engineering.
Given these kinds of interesting and useful analyses isn’t just varying tire pressure going to have a massive effect on the feeling of comfort and compliance (and traction)? I’ve even noticed a big difference between TPU tubes and butyl.
You’re giving me bad memories of college Fourier analysis…. I survived it…and still don’t fully understand it. I’d be interested to see an analysis of compliance due to a carbon seatpost on an Alu gravel bike .. mine rides a bit harsh.
it would be interesting for other people to make measurements and expand the dataset. we know characteristics of the tubes used for the frame, but i wonder how different rider weight and riding style could affect the feeling of compliance. A 160lb rider definitely is going to have different feel than someone 210lb.
Awesome! That's one of the many reasons i am subbed. Also, as a Chemist, I am a big fan of data in general :) - Very cool!
would be cool to record the same data comparing wheels/tires. really cool and informative video, thanks!
Hi, can the vertical compliance in a frame come from the top tube bending? I have noticed on my slo-motions that my hardtail’s top tube slightly flexes on big impacts (handlebars also flex). Does this explain the ‘vertical flex’ thing or is the rear the only part of the frame that should be flexing. Thank you
World be great to see a comparison between steel, carbon and alu sa well!
I'd be really interested in data from the sensor placed at the top of the seat tube.
Very interesting look at the “feel” of steel that is thrown around in conversations. With that in mind perhaps the “contact points” are where this feel can be analyzed if possible. The grips,the pedals and the saddle.
I agree that lateral movement is probably more significant but I would speculate that thin wall seat stays and chain stays would have some vertical movement. Just a thought
I would think that its a fairly uncontroversial statement that after tires, the overwhelming majority of vertical deflection on a rigid bike comes from seatpost and handlebars. The frame material and design are probably nearly irrelevant.
@@cjohnson3836 I have stayed with 31.8mm diameter bars on my mountain bikes and honestly the additional stiffness of 35mm was not necessary for me. So many variables into the feel of a bike and I don’t know it can be isolated to just one or a few frame materials. But I do think that the design and layout of the tubing would have noticeable impact on the overall feel.
@@hardmtnbiker I think handlebars get into a different world than the frame. Now you're dealing with sheering forces loading onto a lever arm. Material properties now do matter because you're trying to thin out a tube as much as possible while maintaining safety. Steel and carbon can really take it further than aluminum, in this application. When it comes to frames though, I think it (material) is not particularly relevant. We can talk all day about the on paper differences between materials but the rider is the same. And so that material still has to conform to the same end, safely support a rider of approximately 160-230lbs across a particular type of environment (road, trail, whatever). So, regardless of material, it has to hold up that 230lbs rider, safely. Therefore, yeah, tube dimensions are going to be manipulated to provide that, however the material requires. And so the differences between material as a contributor to "feel" I think disappear.
Out of all (steel, alloy and carbon) frames I have owned the honzo was probably the most uncomfortable and least compliant. Quite a shame as it did handle great, just felt like riding a brick on wheels.
I love those nerd stuff! Do you fell the comfort increase on the saddle or riding on the BB? The seat post joint on the second frame seems to have way more freedom to flex and make seat post to work like a spring. If is this the case, would be interesting to put the sensor in the top of the seat post.
Interesting experiment, thanks for doing it! I hope you don't mind me asking: what is your day job?
College engineering professor
We have a brand-new suspension system that more than compensates for the lack of vertical compliance in triangular shaped frames. TYRES!
Would have been interesting to see a measurement at the top of the seat tube. Is there a reason I'm missing as to why this wouldn't have been a good point of measurement?
I ride 2 hardtails, a Ragley Big Al and a steel Esker Japhy and think a lot on this subject. I love both bikes but they are different in how they feel through the pedals…and I ride clipped in candy pedals with stiff xc shoes! If I extend the Japhy chainstay so it is matched, or as close as possible in length to the aluminum Big Al, there is quite a difference in feel on these two bikes. Well, even on the Japhy’s shortest chainstay length there is a difference in feel. I say “different” because the details just remain elusive as what the hell is really going on when each bike is ridden hard. But it is definitely different! Well, I do have a favorite and it is the steel bike. Long rides or really going for it on rough terrain, a good steel frame just has that feel in pedals and also how bike kinda moves with you. I don’t know why but the bumps, the rough, the vibrations…all seem just right on steel.
It makes sense to me. Energy needs to be transferred and dampened and dissipated, whether by allowing vertical or lateral frame flexing. It doesn't matter the direction as long as there is a way to transfer and dissipate the energy.
Data is fun. I wonder if a motioncapture video of your position and movement on the bike would be able to give some insight into your ride feel. I know it is close to impossible, but nevertheless that data would be fun to dive into.
Do you have any knowledge of the material/construction differences between the bikes (steel isn't just steel) and tube shapes and welding technics. The rabbithole goes on and on.
Well great videos. Love them.