Talking Telemetry and Downhill Bike Setup with Giant's Dave Garland

Jun 13, 2017 at 15:24
by Mike Levy  
Talking Telemetry with Dave Garland.


While it can look like controlled chaos from the outside, World Cup downhill racing is often a game of millimeters, tenths of seconds, and the non-stop search for the fastest setup. Using telemetry, an electronic automated system that collects data, to get as close as possible to that elusive perfect setup isn't a new thing, but Giant team mechanic Dave Garland has put together his own compact system to help Elliot Jackson tune his Giant Glory Advanced. More specifically, to dial in Elliot's bike to provide as much grip as possible.

Pinkbike's Ross Bell sat down with Garland at the Leogang World Cup to get a closer look at Elliot's bike, and to find out exactly what all those wires and that black box actually do.

''This project really started in my mind as far back as 2008 when I began looking closely at componentry, what each component on the bike does, and its relation to the next part that does a job close to it,'' Garland explained while emphasizing that each part of the bike affects other parts. Understanding how those components relate to one another, and what to change to eek out those few tenths, requires more than just a deft touch and good rider feedback. ''This weekend is the perfect example. Don't quote me on it, but I think you've got fifteen people on the same second,'' he said of the competition at Leogang. ''There's no way on earth that you can guess an adjustment and make that closer. You have to go to something that tells you.''

And that something is a computer.


Talking Telemetry with Dave Garland.
Talking Telemetry with Dave Garland.

bigquotesThere's no way on earth that you can guess an adjustment and make that closer. You have to go to something that tells you.

Garland has mounted a host of compact sensors on Elliot's Glory; there's a sensor that measures rear axle position (a more time efficient approach than having it on the shock, he says), a wheel lock-up sensor that can also measure wheel speed, sensors for fore/aft and lateral movement, and also brake pressure sensors. All of these are bespoke units manufactured to Garland's specs specifically for downhill racing

''I've been lucky to be involved in pretty high-level motorsport awhile back where some type of telemetry was used, but it was impossible for bikes because it was far too big and had a lot of information that you don't really need. So, skip forward to two and a half, almost three years ago now, and I started trying to find ways to measure what I needed to measure in order to make a rider go fast. What you see is pretty much the finished article. There are a few updates, and the great thing is, with this system I can, with my software engineers, actually update it on an almost weekly basis based on what I've learned. That's a great advantage.''


Talking Telemetry with Dave Garland.


When someone mentions telemetry, you probably think of suspension setup and massive boxes strapped to downtubes, probably with a rat's nest of wires coming out of it. Garland's system is quite clean, though, and it measures much more than fork and shock action. ''You might see a few other systems being used right now, and they're largely suspension-based, which is great because suspension has to work properly. But downhill is such a specific sport; there are no constants, and everything is changing all the time. I've found in recent years that people always seem to point the finger at suspension as the cause of characteristic problems like flat tires or over-working components that go to failure. Yes, suspension has to do its job, but it's really just one part of the picture in getting a downhill bike to work properly.''

The end goal is to maximize grip, Garland believes, and to do that he needs to know much more than just what the fork and shock are doing. The fore/aft and lateral sensors, which are fixed to the frame rather than the bike's suspension, tell him if Elliot's machine is moving around too much, a behavior that might indicate too high of a spring rate or excessive compression damping. The wheel lock-up sensor tells Garland if Elliot is on the binders too often or too hard, which can have a massive affect on how the bike's suspension performs. ''This helps the rider understand where they need to let the bike do the work a little more, get their braking done in the right braking zone, and allow the bike to get through the corner with its correct geometry,'' explains Garland when asked about the connection between braking and geometry.
Talking Telemetry with Dave Garland.

''Because, as everyone knows, if you're braking hard through a corner, the bike isn't in its most efficient geometry to get you through a corner.'' Defined braking zones aren't just for motorsport, it seems.

bigquotesThe single most important job you have to produce to make a rider confident is grip; everything is to give the rider as much grip as possible and to make sure that grip is very equal front to rear


Talking Telemetry with Dave Garland.
Talking Telemetry with Dave Garland.


And speaking of braking, brake pressure sensors don't lie when it comes to when Elliot is pulling the levers. ''During testing in the winter, what we found, because a downhill rider needs to ride a downhill bike as much as he can, is that when your fingers are covering the brake levers, without you even knowing it, you're pulling the brakes sometimes. It's due to muscle memory.'' This testing revealed values of 20, 30, 40, 50 bar (290 - 725 PSI) of pressure when riders were saying they weren't braking, simply because their fingers were on the levers and it was a subconscious action. More braking equals less speed, of course, and vice versa, so have no doubts that the team has been working on this one.

There is also a rear axle sensor that measures the wheel's position, something that Garland says is a more time effective way of evaluating suspension action than mounting a data acquisition unit directly on the Glory's shock. Garland pays special attention to front and rear ride height - where the fork and shock are sitting in their stroke. ''We can tell whether that's equal, and it does need to be equal all the time in order for the bike to be balanced,'' he says, with an unbalanced bike asking its pilot to shift their weight distribution around too much.

''Downhill bikes are at the point where if it's correct and everything else is correct with your body position on the bike, you're steering a ship. You're no longer very animated on the bike. You can see the riders who've got that worked out because they're very still, and they don't have to make a bodily correction very often to get the bike to do what they want it to.''


Talking Telemetry with Dave Garland.


Garland believes that equal balance produces grip, and grip is the key to going fast: ''The single most important job you have to produce to make a rider confident is grip; everything is to give the rider as much grip as possible and to make sure that grip is very equal front to rear.''

A World Cup race weekend is a fairly condensed event with limited time available for practice and testing, so one of Garland's main focus with his telemetry setup is to have an effective system that produces easy to understand metrics. A racer is likely to only get eight or nine runs, at most, during practice, which means that the data has to be simple and clear to understand. ''This is about taking the guesswork out of it; it's not going to solve every problem, but it helps you go in the right direction very, very quickly,'' he says.

What shows up on the computer screen are easy to read metrics with relatable stats laid over other relevant data so that a rider and mechanic can easily see how things interact and the knock-on effect. ''A lot of it is down to flex in each component. I mean, the wheel is really important for producing corner speed; if that thing is really, really stiff, then you're expecting the tire to do all of the work. Well, it can't do it all; it has to have its partner to help it,'' Garland asserts to highlight the connection between components. ''Lateral and vertical movement; I've been doing this kind of tuning since the late-90s. [Spoke tension] is one of the most important adjustments you can make.''


Talking Telemetry with Dave Garland.
Talking Telemetry with Dave Garland.


But Garland doesn't blindly trust the computer, however, as what the rider is feeling being just as important: ''We have camera footage to go by as well, and the most important value is what the rider tells you. As long as you can draw a picture between all these different parameters, then you can quickly and effectively work out where you need to go.''

Despite much of the field turning to 29ers, Giant's downhill bikes all roll on 27.5'' wheels, something that probably won't change during this season. Garland is still looking forward to getting the most out of the current platform using his telemetry system, though, and he believes that there's still plenty of room for improvement: ''These bikes, 650s, in my mind they're being sold a little short because we're really only in the second evolution of these bikes. You know, 26ers probably went through fifteen evolutions to get to where they were four or five years ago. Now we have 650B bikes that are well capable of doing what a 29er does. Obviously, 29ers makes it slightly easier but, for me, these bikes are still, and will be in the future, very competitive bikes. We've seen that this weekend,'' he summed up, citing Gwin's winning run on 27.5'' wheels.


Talking Telemetry with Dave Garland.


Must Read This Week

113 Comments

  • + 121
 "Don't quote me on it, but...."

PB: *literally quotes him on it*
  • + 7
 I am actually so surprised its taken this long for all the teams to start dabbling in using it after all the Sunn Chippie team were using it back in the late 90's with Nico and Anne-Caroline.
  • + 1
 @heavyp: data acquisition has been used on many riders' bikes ever since then. Nico/Sunn were just the most prominent ones using it.
  • + 71
 Telemetry: sexy sounding word that is often used incorrectly to describe what is really called "data acquisition". Telemetry means you are wirelessly transmitting the data in real time and monitoring it remotely. Data acquisition is recording data to later analyze, which is what is being described in this piece.
  • + 13
 Agreed, this is definitely data acquisition and NOT telemetry. It is interesting that some people so close to this technology don't know the difference.
  • + 10
 Double agree. Grand Tourismo taught me something.
  • + 23
 Pffft, you say telemetry I say tomato.
  • + 6
 @panaphonic: But it wasn't how to spell..
  • + 0
 @deadmeat25: I can spell English words just fine thanks.
  • + 6
 @panaphonic: its Gran Turismo i think is deadmeats point...
  • + 2
 Someone owns a DJI
  • + 1
 @m-t-g you are correct. My husband has spent years not to mention £££££ on this project and it is unbelievable
the things it measures and is capable of.
  • + 7
 my thought EXACTLY.... leave it up to the bike industry to throw around sleek sounding terms that don't actually correctly describe what they are labeling.
  • + 16
 "[Spoke tension] is one of the most important adjustments you can make.''
So, Mike Levy, given the brackets around spoke tension, you inserted those words in place of some other words or gesture from Dave Garland. I'm curious what Dave *actually* said, since its been shown both theoretically and empirically that spoke tension (assuming it remains above zero) does not influence lateral tension in a spoked bicycle wheel.
So is Dave actually tweaking something else, and the words inserted by Mike are inaccurate?
Or is Dave tweaking spoke tensions for some other purpose? Do WC downhill mechanics actually WANT the spoke tension to hit zero and permit significant rim deformation?
Would love to hear more about this. But don't like the direct implication that higher/lower spoke tension = higher/lower lateral stiffness of the wheel, contrary to physics.
  • + 14
 it was a gesture. he was dancing and waving his hands wildly in the air. like whacky waving inflatable arm flailing tube man.
  • + 6
 Yep. E is E, spoke tension does not affect spoke stiffness, other than the mentioned situation of a slackened spoke. Sheldon Brown empirically busted this myth many years ago but it continues to live on somehow.
  • + 0
 Ok spoke tension doesn't affect lateral stiffness.

But I guess the system has to be considered as a whole : more tension in the spoked puts more stress on the rim etc.

And what about vibration transmission in a highly stresses system?
  • + 1
 But it does affect the side stiffness of the rim.... real stif rims accept no sideways deformation, and there you need more than a tire deformation only. Should that be the reason why offroad motorbikes use spokes most of times in stead of alu or steel rims?
  • + 0
 And there is somthing between lateral and side force, sorry for my Englisch but the wiggling of the rim perhaps? Spokes are in the mid mostly and supporting there right?
  • + 11
 Spoke tension doesn't affect lateral tension, it affects total wheel compliance, holding the axle fixed and deflecting the rim. I assume you've seen the ubiquitous slow-mo footage of riders cornering with the rear wheel locked (i.e., skidding through a corner) and the rear wheel deflecting and springing back each time it contacts and leaves the ground? That's the movement spoke tension importantly affects: for a given type of spokes, tighter spokes means less deflection or more stiffness. This is because the spokes act as a spring, pulling the rim back onto the axle's plane, and more tension effectively preloads the spokes, giving a higher initial spring rate. Besides, Garland isn't the only guy doing this - Manon Carpenter's mechanic laces her wheels at lower-than-normal tension to make her wheels more compliant. It's a thing.
  • + 1
 Jesus, finally! So many DH mechanics pass this old wive's tale down to their riders -it's 100% BS.
  • + 4
 @Jubbylinseed: What you said is absolutely not true. Spoke tension does not change wheel stiffness, in any plane, as long as they are tensioned such that the spoke tension remains above zero.

It does change the resonant frequency of the spokes themselves (between the hub and rim), which can have an affect on perceived ride feel, but it does not affect the tires interaction with the ground.
  • + 1
 OK, now we need a PB article that shows some test results between wheels with different spoke tension. Show me the numbers.
  • + 2
 @bcmrider: I'd love to see this. Stick some telemetry on the rims and empirically show me the supposed compliance that these mechanics can "tune" by adjustments to spoke tension.
  • + 1
 @Jubbylinseed: The videos you describe simply illustrate deformation of the wheel/rim. They offer no evidence that differences in spoke tension produce more/less of that deformation.
Bicycle wheels absolutely DO deform / exhibit flex, no doubt. Smaller rim cross sections (narrower, shallower), lighter rims (less material), fewer spokes or skinnier spokes will all produce a more flexible wheel. Less spoke tension will not. Perhaps this is why we see most DH bike staying with (relatively) narrow rim profiles to retain some (apparently) desirable lateral compliance. But if WC mechanics really are reducing spoke tension in belief they are increasing compliance, it would be a shame since it is a physical falsehood. Although it may explain the occasional exploding wheel.
  • + 1
 @Inertiaman: Yet the sheldon brown article that everyone refers to shows approx 10% variation in stiffness (deflection) from 2 full turns of the nipple. www.sheldonbrown.com/rinard/wheel_index.html
  • + 0
 if a spoke is a spring, and it has more preload(higher tension), that would not increase the lateral flex your saying? If you have a wheel with way low tension it flexes laterally quite bit no?
  • + 2
 @m-t-g: Sheldon Brown busted this myth using radial-laced wheels. When you lace a wheel with a 2 or 3-cross lacing pattern along with j-bend spokes, your spokes are not in a straight line, which makes them act like springs. The higher the tension, the straighter this line is, the stiffer the wheel.
  • + 1
 @Cordall: This is true, but if you read the full text, the author explains that those 2-turns-loosened spokes are approaching zero tension and the wheel is "almost sloppy." I doubt any DH riders are heading down the course with spokes at 5 or 10kgF tension. Across the normal ranges of tension, say 80kgf to 120khF (for the steeper spokes aka drive side rear or rotor side front), or even substantially below that to 40kgF or less, there is no change in lateral stiffness.
  • + 2
 @Inertiaman: No, look at the results again. There is a trend appearing as the spoke tension is changed. The first set of results show 10% change over the first 7 measurements, the second set of results show 5%. This is ignoring the last 3 measurements as zero tension is approached.
  • + 1
 @Cordall: But the results show that the wheel gets ~10% stiffer as the spokes are loosened, not tightened.
  • + 1
 @Cordall: As @toast2266 points out, the "trend" you are highlighting is actually an *increase* in tension, not decrease, so it contradicts your argument. But more practically speaking, that "trend" is thru the measurements during the first 8 turns of loosening, where all are within (or nearly within) his range of error of measurement, which is .0002". Essentially all the measurements are .069 +/- .002". So the apparently (and VERY slight) increase in tension can easily be explained by measurement difficulty at such small distances.
More importantly, these measurements are really just an example to support the mathematical/engineering "on paper" expectation that the tension will not influence stiffness.
  • + 1
 @AntoineCaron: spokes behave as springs no matter what lacing pattern is used. I've read a couple research papers on wheel stiffness vs lacing pattern, and the conclusion there was that lacing pattern had little effect on wheel stiffness overall. There were some differences, but pretty small.
  • + 1
 @Inertiaman: @toast2266: I never said that spoke tension did, or didn't effect wheel stiffness. Just that the only evidence that we have at hand shows directly that there is a correlation. You cannot just arbitrarily apply a +\- tolerance so that the figures fit YOUR argument. As much as you don't want to admit it, look at the numbers again. The first set show a 10% change. 10% is not a tiny insignificant error that you can just choose to ignore. The second set show a 5% change, low and behold, in the same direction as the first. Just what kind of engineer or scientist do you think would ignore those results? If you are so set on stiffness not being related to spoke tension, please point me in the direction of some experiments that show it.
  • + 2
 @UtahBrent: There's no reason to assume spoke tension always remains above zero in a DH context. This is where Sheldon Brown's argument falls over. I'm suggesting that the forces involved in DH riding can be enough to slacken the spokes at the bottom of the wheel (for a vertical impact) or on one side of the wheel (for lateral loads). Spoke tension affects how much the rider can load up the wheel before the spokes opposite the load go slack. If each spoke has 800N of force on it, that's 800N the rider can apply in the opposite direction ("squashing" that spoke) before that spoke goes slack. Most spokes won't directly opposite the force applied, so 3200N or a bit more would be enough to slacken the spokes. Spoke tension will directly affect how much force the wheel can take before the spokes opposite the wheel goes slack, and the wheel starts to flex.
  • + 1
 @m-t-g @Inertiaman
@Jubbylinseed is correct, and it illustrates the problem with Sheldon Brown's analysis. Looks like many people read the graphs very accurately, but missed an obvious detail. Check the load value he used (115N) and the reasons given for this choice. This value does not represent anywhere near the dynamic loading applied in a real world scenario, which throws the entire analysis out the window, not to mention measuring tension in "number of turns back". Like Jubby explains, if you use realistic loading at lower tension, then the scenario of zero tension suddenly becomes *very* plausible - and thus - spoke tension *does* impact wheel stiffness. Of course you only have to ride a tight and loosely tensioned wheel yourself and shred some grippy corners to experience this. Just because the *static* spoke tensions are above zero, does not mean they will stay so in a dynamic scenario. "E is E" just suggests that the spoke material will have linear elastic behavior, you cannot extrapolate this to the wheel itself! A wheel is not solid nor linear elastic. Wheel stiffness will indeed vary with spoke tension, and peak loading will vary depending on rider mass and usage scenario. For those touting physics and engineering as an explanation, the fault is rarely with science (as you all correctly suggest), but with the person applying it. But hey, let's not forget that time Albert corrected Isaac either - we should all think for ourselves. Smile
  • + 1
 @Cordall: Regarding "you cannot arbitrarily apply a +/- so that figures fit (MY) argument.".
The +/- figure I am applying is NOT arbitrary. It is the measurement accuracy of the jig/instrument that Damon used for his analysis. He specifically states that his setup was accurate to .002". The supposed changes you are citing all fall within that amount (with a single exception, all of the figures you are referencing = .069" +/- .002"). In other words, they are all equal if you take the measuring accuracy into account.
Even if we accepted your interpretation of 5-10% TOTAL change over 7-8 incremental tension reductions, you need to consider that in the context of a ~ 100% change in deflection on the 10th increment.
So measurable deflection went virtually unchanged despite an 80%+ reduction in tension. You don't consider that supportive evidence?
(note: the 80% mentioned above is my conservative description based on my own measurements bringing wheels to tension; spoke tensions rise from 10-20kgF to >100kgF in the final 1.5-2 turns of nipples).
So to answer your question "what kind of scientist or engineer would ignore a 5% change" . . . well, ALL scientists or engineers will ignore 5% changes if the measuring accuracy of the instrument is also 5%. That's what margin of error is all about.
  • + 1
 @uuuu: Several comments to your post:
a) the analysis referenced above is not Sheldon Brown's work. It was done by Damon Rinard. It just happens to be archived in the Sheldon section of his (former employer) Harris Cyclery website.
b) Damon's test, while interesting and useful, is hardly the only "evidence" to support the physical fact that lateral stiffness in a spoked wheel is essentially unaffected by spoke tension. Go read the "Wheel Building" section of the Nox Composites website if you want a more contemporary take on wheel physics (including a shout-out to Damon)
c) Damon's work is not categorically irrelevant (as you suggest) because his loading value is less than what you claim real world values to be. What would you consider a good real world lateral load example? Why? What is your hypothesis to support the idea that Damon's data would look any different with a 300N load or 400N load? There is no physical evidence to suggest that his outcome would be different w/ a 300N load (the starting deflection would be greater, of course, but it would be largely constant until full de-tensioning ocurred).
d) You correctly say that zero tension can influence wheel tension, but then you incorrectly leap to "thus, spoke tension *does* impact wheel stiffness." More accurately, you should say "the absence of tension influences wheel tension." This in no way proves wrong (as you suggest) the assertion that changes in positive tension do not have any corresponding change in wheel stiffness. No one anywhere on this thread has identified evidence, or theory, to support the idea that changes in positive tension cause changes in stiffness (probably because no such data exist).
e) yes, a wheel isn't a static object. So what. That doesn't mean that static analysis of it is irrelevant. You need to submit a rationale for why and how its dynamic behavior may change substantially from its static behavior. Damon't test was only on silver alloy wheels. Does this mean that the results are irrelevant for wheels that are anodized black? Of course not.
f) Suggesting that bike wheel spokes are routinely reaching zero tension as an explanation for how tension changes might influence ride feel is absurd. Spoke tensions in a properly built wheel remain above zero in nearly all use conditions. When spokes do become de-tensioned, wheel failure becomes a very real possibility, because the load must be supported solely by the rim "arch" in the region of de-tensioned spokes. Bicycle wheel reliability is largely dependent upon maintaining substantial tension in the spokes: the pre-tensioned structure is the "magic" that makes bicycle wheels such an extreme example of high strength-to-weight design. Perhaps WC downhill mechanics, where a wheel "life" need only be one 3 minute run, are knowingly sending wheels down the hill with the expectation to de-tension them. I'm skeptical that's the case, but I'll listen if its the actual mechanic telling me, despite the fact such an idea is directly opposite of virtually every commercial wheel builder on the planet.
g) As an example of (f), roughly speaking, a single wheel can support a 400kg radial load without any spokes reaching zero tension. So two wheels = 800kg. In other words, you'd need to generate more than 8G's in a bermed turn to de-tension even one spoke due to radial loading. Needless to say, the best dual slalom riders aren't generating anywhere near 8G's.
h) Lateral forces to a wheel are essentially zero in an "equilibrium" turn, such as a berm. Radial forces from rider weight, even under G-outs, are relatively modest. Large lateral forces are most likely to occur due to substantial oblique impacts to the wheel from trail obstacles. Large radial forces are also most likely during severe impacts, of such severity that it bottoms the tire to the rim. The antidote to such impacts/forces is optimally high spoke tension, to minimize the risk of wheel failure, and not some "low" tension that is mythically going to provide a "compliant" ride.
  • + 1
 @Inertiaman: Read @Jubbylinseed's most recent post once again. The higher the spoke tension, the greater the load that can be applied (via rider mass and riding forces combined) before any single spoke in the wheel hits zero tension. The inverse is also true. This means that your (claimed) 400kg peak radial load value will be DIFFERENT depending on the spoke tension. This (from what I can see) leaves your only argument as one that suggests this scenario is somehow not very likely. I mean no offence, but most data analysis on this topic is from a road and best-case XC background. As Jubby hints, the peak forces in DH riding are *much* higher. These are not insignificant differences and you only need to watch both sports to see this. Forget the lateral G in a bermed corner, watch a rider going off a drop or smashing through a rockgarden. It's the sharp impulses where a moderate to large force is applied over a very small dt that will cause tensions to drop.
I never said that a static analysis was irrelevant, but what I am saying is that you cannot extrapolate purely static analysis to a dynamic scenario and expect correct results. That is exactly what has been shown here by many people. As I pointed out, the static analysis' mentioned here as "evidence" are heavily flawed anyway.
Feel free to link an article that uses DAQ on a WC level downhill bike to record peak radial and axial forces at the rim if you wish to discuss further. Honestly I gain nothing from convincing you, and am not out for an argument, but I can tell you that if you find data that is actually relevant to this scenario, you will find that you are wrong in your initial (very blunt) hypothesis.
"Spoke tensions in a properly built wheel remain above zero in nearly all use conditions." - a quote by you without any data to back it up. If you want to prove your point, you need to find this data for specifically the application you are discussing - WC DH.
FYI, reading your later points, I agree that using lower spoke tension to encourage "compliance" is silly - just like using an inverted fork to achieve the same thing. Flex (or "compliance") of these varieties act as an undamped spring, which *generally* decreases vehicle predictability and response. Optimally high spoke tension is best IF wheel strength and stiffness are primary goals. But two points: a) I believe you are making incorrect claims based on incorrect data, in particular: inadequate peak force values (so the mechanic technically isn't wrong in the cause and effect of his claim - at least innocent until proven guilty in my book) - and b) unless you have done double-blind timed testing on a WC level track with a WC level rider, you cannot discredit the possibility that more compliance in the wheel may be allowing faster times (so the mechanic may actually be doing what's best, even though for the average rider it'll probably just result in wasted money and faster-destroyed wheels). I certainly don't agree with it in principle, but as you hint yourself, it's entirely possible that it's better - and when the one thing that matters is the time on the clock (not mechanic labour or parts cost) it's not an unreasonable choice.
Basically I think you've had a crack at something here that isn't actually wrong, on any count. But here I was hoping to get by with a Family Guy joke and have instead been roped into an internet argument. Happy trails!
  • + 1
 @uuuu: These discussions are quickly becoming pointless in this venue. Some final thoughts:
- just because DH use conditions are extremely different than road or even XC use conditions does not exempt the wheels from the physics of a pre-stressed spoked wheel structure. These physics have been thoroughly described as far back as Pippard's definitive work in the early 1930's. Additional engineering work has since been done in several venues to support and expand that work. DH is "different" but its still a pneumatic tire on a spoked wheel. Ricard's work noted here is only an informal illustration of more fundamental physics equations that had already determined what Ricard's experiments would show. Attacking that as not relevant because it is not DH ignores the underlying physics.
- you say "if you want to prove your point, you need to find this data for specifically the the application you are discussing.". Why is the burden on me, and not you? Several of the claims here, and one of the statements in the article (potentially mis-quoted) are asserting things that contradict long-determined physical and engineering fundamentals. Shouldn't the burden be on them to explain exactly why and how those fundamentals don't apply in DH? Show me an equation for radial or lateral stiffness -- static, dynamic or otherwise -- that includes spoke tension as a significant contributing variable.
- much of your argument seems to pivot on the assumption that zero tension is easily/often reached in DH scenarios. I think you are correctly identifying the scenarios that would produce zero tension (very brief, high force impacts), but I think your extrapolations about what is made possible by those events are deeply flawed. Consider:
--- the de-tensioning events you are highlighting only de-tension one or perhaps two spokes. Its been repeatedly shown that the LAZ (load effected zone) of a bicycle wheel is centered at the bottom 4-5 spokes, typically distributing loads by roughly 5-20-50-20-5 percent across 5 spokes. The tension in the other 27 spokes in a 32 spoke wheel actually RISE (slightly).
--- related to the above, when I (and others on this thread) have noted the "> 0 tension" caveat, we are referencing tensions on ALL spokes. Experiments (like Rinard's) that show lateral deformation at zero tension for all spokes are not representative of the case of localized de-tensioning.
--- asking what would happen w/ lateral stiffness in the case of localized de-tensioning is a valid question. To "guess" an answer we can observe that adjacent spokes are still tensioned, so an upper limit for lateral deformation would likely be determined by the rim structure's ability to flex laterally in the region of 2 spokes or so, = 3 inches -ish of the circumference. I think its obvious that the lateral deformations in that 3 inch region aren't going to be very large.
-
I stand by my statement that spoke tensions remain > zero in nearly all use conditions. That fact is the fundamental basis for the integrity of pre-tensioned wire spoked wheels since their invention. Even on a DH bikes, the % of the use time when de-tensioning occurs is extremely small as a percentage of use time (how many severe impacts per run that totally bottom out the tire to the rim?) and that figure must then be multiplied X the percent of spokes it occurs to (~5%) which gives fractions of 1% of use time in DH where zero tension is occurring.
-
With that, I'm done trying to condense 40 page research papers into comment section posts. What I'd really like to see, and all I was trying to solicit with my original post, is this: I would like to hear/see more specific explanations from WC DH mechanics on exactly what they are doing w/ spoke tensions, and what they expect/believe/feel that is providing in performance. If a mechanic can say "at 25 psi tire pressures and atypically low 30kgF spoke tensions, the wheel does such-and-such which gives benefit so-and-so, but we only do this on race bikes because spoke and rim fatigue would preclude it as a longterm strategy" then I would be fascinated to listen. I'm not holding my breath for such an article to appear.
  • + 1
 @Inertiaman: I see no valid data, just a lot of rambling to try and back up a statement you made, which you're quickly realising is incorrect. Why is the burden on you to find something to back up your claims? Because you're the one who (very bluntly) stated that the mechanic (and/or the writer of this article), his claims, and all similar claims are incorrect. If you're going to call someone out then IMO you should have *relevant* proof on hand to back yourself up, and it's clear you don't.
No engineering or fundamentals in physics have been breached here. If you had a thorough understanding of the concepts you would realise this. The input loading is what defines the magnitude of positive tension required to prevent spoke tension hitting zero. We know that static spoke tension has a realistic limit (we can only apply so much tension to a spoke), and we also know (or should know) that the input loading is essentially unrestricted (how hard can you smash into something?). While the input cannot be infinite, it is a *very* common real-world scenario in WC DH to impact the rim to the point of permanent deformation (eg. a flatspot). Given the facts that a) flatspots are a common scenario (to a plastic deformation magnitude well above what would cause tension to hit zero - eg. 20mm rather than 5mm), and b) rim materials allow elastic deformation before plastic deformation (with even greater elastic potential with some carbon rims, now commonplace), it only takes a little logic on top to realise that zeroing spoke tension in DH is actually an incredibly likely scenario, even without permanent rim damage occurring. The exact same scenario applies to lateral forces and knocking a rim out of true, there are plenty of real-world scenarios in DH where rims have been pushed well beyond the zero tension point. Again, the amount of elastic deformation possible in rim materials is easily more than what is required to shorten spokes by the 5mm (say 10mm max) required to zero the tension.
The amount of actual engineering knowledge required here is first-year university level at best. The part you've missed is logic and application. Do you ride DH? If so, these concepts should not be foreign, in fact the engineering concepts are optional - a lay person should be able to understand this. I hope you now can too. I will politely eject from this conversation, a good day to you. Smile
  • + 1
 @uuuu: your words: "The input loading is what defines the magnitude of positive tension required to prevent spoke tension hitting zero. We know that static spoke tension has a realistic limit (we can only apply so much tension to a spoke), and we also know (or should know) that the input loading is essentially unrestricted (how hard can you smash into something?). While the input cannot be infinite, it is a *very* common real-world scenario in WC DH to impact the rim to the point of permanent deformation (eg. a flatspot). Given the facts that a) flatspots are a common scenario (to a plastic deformation magnitude well above what would cause tension to hit zero - eg. 20mm rather than 5mm), and b) rim materials allow elastic deformation before plastic deformation (with even greater elastic potential with some carbon rims, now commonplace), it only takes a little logic on top to realise that zeroing spoke tension in DH is actually an incredibly likely scenario, even without permanent rim damage occurring."
I agree with all of that. Oddly, I don't think it contradicts anything I've said. So a very simple question for you: what do you think might be a representative number for the # of times, on a typical DH run, that the rim experiences an elastic deformation event as you describe? Gross generalization of course, but toss out a number. 10? 20? 100?
  • + 10
 I'm sure all this stuff is top secret, but it'd be really interesting to see that screen that he's looking at. some kinda graph to make sense of what they really are measuring and how they analyze it? Are they using a power meter on this bike also- be interesting to see who rider input changes these measured items. One thing I always thought they are missing out on for measuring is brake force: On/Off + intensity - Are you dragging your brakes, braking early/late etc. Combine all that data together and i think you have a good combo for analysis...
  • + 4
 One thing I've learnt from mountain biking over the years. It doesn't matter about the sizing or shape of the bike but how much time you've had getting used to the bike. (No shit, I'm faster on a 15" Hardtail on any track, pump track or trail and I'm 6'3" now). But, that's difficult for modern DH riders as all the new tech changes every single year! Hear me out tho'. This type of technology will become irrelevant unless the UCI changes it's DH race courses. Why? They are becoming less about long travel and sensitivity, weight shift and brake power. It's all about how fast is the top speed and how big are the crowd pleasers.
  • + 3
 I'm guessing they use accelerometers. The x axis is time and the y axis is a positive and negative "accerlation" measurement along that axis. Where a negative and positive acceleration are defined by the axis. They have multiple graphs to get an idea of how the bike acts along different axis (a more 3-dimensional picture).
  • + 4
 That's what I love about this sport. So much depth to it. The old Oxford Brit was one smart cookie with his laws of motion! So much passion and so much fun. It's like baseball in that way. It's just as much fun for the genius types who want to geek out on it as it is for the beer guzzling party crowd on the side of the track. The athleticism required to pull it off is second to no other. Equal perhaps but certainly not less. And the DH crowd knows how to let loose and have fun! This is one of the greatest sports in the world as far as I'm concerned, as it can be enjoyed by all. Unless your a stick in the mud.
  • + 0
 "Oxford Brit"?? Who's laws of motion are you referring to? Isaac Newton was a Cambridge College man.
  • + 7
 what like to here more about spoke tension and it's effects.
  • + 10
 Indeed. Tell me again why I need the stiffest possible wheel, which can only be achieved with a carbon rim at 8x the price of aluminum?
  • + 3
 @slyfink: And if we even needed boost axles in the first place?
  • + 2
 Agreed, that's a bit of arcane knowledge that is hard earned and kept close to the breast my friend. Everything is a system and should work together. That's the beauty of a well calibrated instrument which is what a well tuned bike is.
  • + 3
 Jared Graves mentioned this during an interview about his EWS setup about a year ago - softer wheels.
  • + 3
 ENVE currently working out a marketing campaign for "ultra premium laterally compliant composite rims"
  • + 2
 Fancy doing this yourself? Have an android phone? Know how to read numbers and graph? Download this;
play.google.com/store/apps/details?id=com.nowinstruments.vibsensor&hl=en

it allows you to read and record vibrations. Sellotape it to your bike in different places to mess with tuning. Or orientate it on your person near your bodies center of mass and secure you can measure with back to back runs if the ride is smoother or not.

It really boils down to acceleration of a wheel with respect to time. Message me if you want help or info.
  • + 1
 You can add shock displacement using an action camera and free tracking software. Search for free mtb telemetry on youtube.
  • + 1
 @piesforyou: Definetely, Would you base it off shaft displacement of wheel displacement based on a pivot?
  • + 6
 One of the most lucid telemetry articles for me, very nicely done.
  • + 7
 You mean data aquisition?
  • - 2
 @karoliusz: Naw, I meant telemetry ; )
  • + 1
 Dave talks about grip. Gwin himself mentioned he chose his tires for grip not for lower rolling resistance. As far as spoke tension goes it applies to grip. If the wheel is too stiff it has the tendency to bounce off line. If it has some compliance it allows the tire to find or maintain grip. Think about how a slow rebounding tire maintains grip better than a hard or fast rebounding compound tire. In the moto-cross world some riders have a selection of crowns available. Ones that allow more or less lateral flex of they fork. Think direction of forces. The forces are not all in line with the direction suspension movement.
  • + 5
 But spoke tension (again, with the caveat it remains > zero) does not influence wheel stiffness, either radially or laterally. It seems intuitive to many that it should, but that intuition/myth is physically demonstrable to be false. Spoke angle/hub flange width, rim material, rim cross section and even spoke diameter contribute to wheel stiffness, but spoke tension does not.
Radial deformation of the wheel/rim is on the order of 1/1000th of the deformation of the tire . . . how exactly is that contributing to traction?
This is why the quote in the article is so interesting to me . . . I have a hard time imagining that a WC mechanic isn't aware of the physics of a spoked wheel. Something is surely getting lost in translation.
  • + 1
 @Inertiaman: well a tire can roll... so there is also latteral&side impact together.. rim only haves spokes in the middle...
  • + 1
 @Inertiaman: i mean rolling in corners as sideways rolling from force from bike and riders weight
  • + 1
 @Inertiaman: Can I ask a dumb question here? I'm a maths person, not an engineer. What does 'stiffness' mean? I assume in some sense that it means the amount by which the spoke bends under an applied force.

If that's the definition, then I guess I can see how tension and stiffness are unrelated.

However I did not know this about spoke tension / wheel stiffness.
  • + 1
 @WaterBear: stiffness is resistance to flex, or resistance to deformation. So in a wheel, lateral stiffness is resistance to deformation from "sideways" forces and radial stiffness is resistance to deformation from "vertical" forces (radial forces may not be vertical; for example when leaning/turning).
Note that I was describing stiffness of the wheel, not of the spoke. What you described above is spoke stiffness, which does contribute somewhat to wheel stiffness. Straight 14g gauge spokes will build a measurably stiffer wheel than butted 14/15g spokes, all else being equal (rim, hub, spoke count), but it will be heavier and (due to metallurgical properties of butted spokes) arguably weaker.
Tension is a very crucial factor in the STRENGTH of a wheel, but does not influence stiffness.
  • + 1
 @Inertiaman: OK. Let me see if I understand this.

Wheel stiffness is unrelated to spoke tension (provided S.T is >0).

I...guess that makes sense? An applied force of X newtons is going to deflect the rim by a fixed distance no matter what other forces are acting on the rim, as long as those other forces are in balance with themselves (which the forces due to spoke tension are, balanced presumably by some internal stresses in the rim).

Am I even close?
  • + 2
 @WaterBear: the rim itself is not relevant to this consideration. What @Inertiaman is (correctly) saying is that your spokes can be considered to effectively be springs with a linear spring rate (stiffness and spring rate both have the same definition - unit force per unit displacement). When they are preloaded against each other (left vs right when we're considering lateral loads) the net force sums to zero while there is no lateral load on the rim. When you push the rim one way, you reduce tension on spokes on one side and increase tension on the spokes on the other side such that the sum of their forces adds to the amount of load you are applying to the rim by pushing it sideways. Until you release tension completely (this is where @Inertiaman mentioned tension having to remain above zero), it will simply behave as a linear spring in terms of deflection (though the rim's own stiffness may have nonlinearities due to the way its profile deforms, but that part isn't affected by the spokes either) because whatever amount of preload (spoke tension) you apply to the spokes on one side is simply opposed by the equivalent preload on the other side.

The maths and the evidence both demonstrate that within realistic limits on a properly functioning wheel (ie spoke tension always remaining above zero) spoke tension is irrelevant to lateral stiffness. However, if your spokes do come loose to the point where they lose tension completely, then things change.
  • + 1
 @Socket: Your explanation is very clear, thanks. I think you are saying, in effect, that the spring constant of the spokes doesn't change with their tension.

Is that correct?

(Thanks for everyone's patience with all the derpy questions.)
  • + 2
 @WaterBear: bang on.
  • + 1
 Great write up! It's always baffled me how little they use data in riding and Motocross. Compare it to four wheels, where EVERYTHING is measured, all the time. Even in junior motorsport, they overlay data from each car to see who braked where, what they can do differently, GPS positioning, throttle, Gs, the whole lot.
  • + 1
 There are a number of differences between MTB and any of those sports though:
1. The rider is the vast majority of the sprung mass on a MTB, and usually about half the sprung mass on a MX bike.
2. Most motorsports where that kind of data logging takes place are on closed circuits, not single-run courses.
3. A lot of maths that is relevant to vehicle handling can be thrown out the door with bikes, because the rider is actively able to participate in very large motions of the centre of mass, and be an active component of the suspension at lower frequencies.
4. There's way more money in all of those sports than DH. Consider the fact that this is basically one guy's project, not the project of a large team of engineers.
5. A lot of stuff with bikes is really hard to measure accurately
6. Even the stuff you can measure accurately isn't necessarily meaningful. Measuring lateral accelerations on a bike for example (as mentioned in this article) is pretty hard because it's hugely affected by the lean angle of the bike. Unlike a road circuit, those lateral accelerations which only exist for fractions of a second could be caused by the bike skipping or bouncing around, or they could be caused by the rider deliberately moving the bike in that way.
  • + 1
 Very interesting. However, chasing fractions of a second with data acquisition but dismissing the 29er advantage with "29ers makes it slightly easier but, for me, these bikes are still, and will be in the future, very competitive bikes".
  • + 1
 I now finally understand why these forums are so full of anger and frustration...

Here in this fantastically interesting piece, we have a picture of the pinnacle of racing and bike development; engineers and racers at the top of the sport working with bike and component manufacturers to shave tenths of seconds off 3 to 5 minutes of a race run; and as Dave Garland says, and has been saying for a decade (anyone remember the CRC tech videos? They were insightful), the attention to the smallest details can mean the difference between being on or off the podium.

Investment into all those small details costs $$$$$, and a hell of a lot of time and effort, and the only way the companies will get that back is if we, the budding amateurs we are, get on board and buy the tech as it trickles down, initially on the dentist's bikes, though the middle-aged middle managers and their kids' bikes, down to the bottom feeders on the look out for bargains.

Sure, if you have the $$$$, you can continue to participate in your chosen "hobby" / "lifestyle" (delete as appropriate) as technology and the latest "development" change apace. Oh, and allow yourself to tell anyone in the group described immediately below this to "shut up and ride your bike" if they express any form of disillusionment with the pace of change.

If you don't have the $$$$ to follow the changes however, due to the increasingly diminishing returns on increasingly expensive additions / changes to ones' bike(s), it leaves you with increasing difficulty participating in the "lifestyle". If you don't have the coin, you have no choice but to be left with distant memories of being "part of the gang" on your increasingly outdated steed(s). Oh, and allow yourself to criticise the industry, and anyone lucky enough to have the cash to keep up with the changes.

These two groups of users are, and will forever be, at each other's throats, and present here.

Which one are you?
  • + 2
 im pretty sure R&D for new components costs way more than some data acquisition tech
  • + 2
 @Asmodai: What R&D is there?
From what I have seen: The bike industry is getting its ideas from other industries and changing it to get patents to fund qualifying R&D. For 80% of bike components its just a New Product Development cycle. Actual R&D comes into valving and possible fluid dynamics. Carbon and composites are old news now and engineers just have to read journal papers to understand the technology behind it to apply it to bikes. Is this R&D? I don't know. I just wanted to ask the question.
  • + 1
 @gaffney92:

how much do you think it cost to buy a couple of data acquisition boxes? 5k -10k at best?
  • + 1
 @gaffney92: R&D = Research and Development. Reading science journals and papers being part of the 'Research' part.
So is data acquisition. At least that's my understanding of it.

It's a messy process I'd imagine, I'm in a different field, but if there's something in common it would be the semi-chaotic nature of the process.
  • + 4
 @orientdave

What I got from it, is that setup far outweighs component choice. I am more than sure that if you have a 5000$ worth DH bike from 2 years ago, paying a guy like Dave 2000$ to come to the hill with you and set your bike up will make your riding experience and lap times better than investing 2k in some components of arguable real-life value. The big steps in bike evolution are over, possibly since some time already. A 2020 DH bike will not give you the same thing over 2016 bike as quality setup for 2014 bike would. You can be certain of it. Meanwhile every user on Pinkbike has a perfectly setup bike. Tell them they may get somethign by paying 2,5k for ENVE rims and they may shrug their shoulders, tell them they may run wrong sag and rebound and they will tell you to go fk yourself. Tell them they may need to ride better and they will gladly hit you in the face at the first opportunity. But tell it to them in the other way like pay for ShockWiz and they may listen. Tell them in PM about some skills site or coaching and they may be interested.

Purchase of new sht will always go over spending time to work on the sht you have and definitely over working on yourself. That's the way humans work in marketing filled "airwaves". And that includes people with limited budgets. I heard a few times how someone wants a lighter fork or bars expecting that it will be easier to lift the front wheel. This is fkng insane.
  • + 3
 @WAKIdesigns: as ever Waki, it is a pleasure to read your thoughts (when not in troll mode!!). Cheers!

I have been quietly riding my 9 speed 26 inch bikes whilst I watch the rest of the field disappear over the hill, and have been left here trying to reconcile my feelings about all things bikes, and my relationship to the whole damned circus.

I am now, finally, finally, gladly "working on myself" in any number of areas (something which I have been neglecting for years under the pressure of running a business and a family). Here's hoping you too have been coming up with answers to life's questions and keeping everything sane and manageable!!
  • + 4
 I wonder if the take the weight of the telemetry components into consideration...
  • + 1
 I wondered this too!!
  • + 1
 of course son
  • + 2
 It makes you wonder just how much further it will go. Will be interesting to see tech continue to develop over the next 5-10 years and what we will be riding/talking about then.
  • + 1
 36er hover bikes.....obviously =P
  • + 5
 Gripping
  • + 1
 Does he have one bike set up permanently (or semi-permanently, for the race weekend) as a telemetry rig and then another set up as a race bike?
  • + 1
 I find it interesting that some pros actually use the "Shockwiz" for adjustments. Or would you say it's rather a "sponsor-thing" that it gets more attention??
  • - 1
 ''This project really started in my mind as far back as 2008 when I began looking closely at componentry, what each component on the bike does, and its relation to the next part that does a job close to it,'' Garland explained while emphasizing that each part of the bike affects other parts. Understanding how those components relate to one another, and what to change to eek out those few tenths, requires more than just a deft touch and good rider feedback. Dave Garland has an expert understanding of what I call FMEA - Failure Mode Effects Analysis........
  • + 2
 Didn't really watch the last round but their bikes looked to be working really well in Lordes.
  • + 2
 I'll be baggin' trophies on my 36er..till somebody comes down the hill on a 39er.

dirtysixer.com
  • + 2
 Best and most exciting article i've ever found / read on PB. Cheers Mr Garland and many thanks !
  • + 1
 This was very insightful as to how teams use telemetry...I always wondered about this type of tech.
  • + 1
 Now thats a tool box and not just a Knipex?Abbey showcase (PS I love German tools)
  • + 1
 I'm working on a prototype to get data based on cheap components. More info to come at www.bitsfromitaly.It
  • + 2
 27.5 is not dead yet!!! at least for another week!
  • + 6
 It's like a reality dating show, this week 27.5 got a rose to remain in the game. Wonder what they'll do for sweeps week??
  • + 3
 27.5 won't be dead as long as Gwin is cozy smoking the competition on one
  • + 1
 very nice article!! great to get such a close look! Wink
  • + 1
 So are those sensors just accelerometers?
  • + 1
 Probably a combination of accelerometers, gyroscope sensors, and strain gagues, undoubtedly rather sensitive and expensive ones. It'd pretty easy to build one up if you've got some coding knowledge and a FPGA microcontroller to hand. If you've got the time and patience it'd be possible to build up a system from a RasPi and sensors, but the accuracy and sensitivity comes from expensive sensors and a finessed package
  • + 1
 @ROOTminus1: I have been thinking about building a data acquisition system using linear pots. Accelerometers would be much easier and cheaper, but I don't really understand how you'd use accelerator data to do anything useful.

I guess you could ride over the same hit several times with different settings and try to reduce acceleration, but it seems like actual suspension metrics would be much more useful.
  • + 1
 @dthomp325: hint, you can get velocity from acceleration.
  • + 0
 @bogey: Velocity is the integral of acceleration, which means it will be difficult to get an accurate velocity value without very frequent re-calibration, but even if you had a correct value of velocity I don't see how that would help you tune your suspension. How does "the bottom bracket is moving upwards at 5 ft/sec" inform your suspension setup?

A traditional linear pot setup tells you how much of your travel you are using and how fast the suspension is compressing and rebounding. I don't think there is any realistic way to derive that data from pure accelerometer data, unless you are riding on a completely flat, completely smooth surface.
  • + 1
 @dthomp325: Using more than one accelerometer will give you the info that you need but takes some serious knowledge, daq and software to pull off successfully.

A LVDT is how I would personally do it to keep it simple but that's not the only way to do it. I linear pot will not give you the accuracy and resolution that you need and may even get wonky with the vibration and shock from riding the bike.
  • + 1
 @bogey: 200mm LVDTs look pretty pricey...
  • + 0
 if you want to win you should use all the tools you have to do it. I like telemetry would love to do one day with my bikes
  • - 2
 Not meaning to be negative but all this "telemetry" as they like to call it helping to shave time and find the quickest setup but I haven't seen a giant on the podium for quite a while.
  • + 5
 look closely, i think i spotted one in lourdes and fort william Wink
  • + 1
 What a fantastic article!
  • + 0
 Can I hook this up to my ebike motor?
  • + 1
 hello asymmetric grips
  • - 1
 technology as is best

Post a Comment



You must login to Pinkbike.
Don't have an account? Sign up

Join Pinkbike  Login
Copyright © 2000 - 2017. Pinkbike.com. All rights reserved.
dv56 0.116841
Mobile Version of Website