 You don't need to be an 'investor' to invest in Singletrack: 6 days left: 95% of target - Find out more
No, not a what bike justification, but I found myself wondering about this but am too dim to come up with the answer;
So, given the same virtual rider, virtual bike & course & therefore amount of output & resistance, could we come with an equation of:
[b]How many pounds lighter would a 26er need to be to offset the increased "rollability" of the same frame with 29" wheels?
[/b]
There must be an "I can travel X miles for Y energy with..." style of thought...?
No
I suppose that it all depends on the terrain and conditions. My 30lb 29er climbs better than my 27lb 26er and on slippery mud it's a no contest in favour of the wagon wheeled wonder. I guess I need to do some back to back testing this weekend.
I suppose that it all depends on the terrain and conditions.
We are assuming a virtual terrain here, same across the board.
njee20 - Member
No
I'm sure there is something close, but it gave me a headache trying to get there...
No, and it would be useless without a similar equation to work out the extra energy needed to get the 29er up to speed (as and when needed) on the same course.
On paper, most would say that Vettel is a better driver than Ricciardo yet driving the same cars on the same track with the same everything else this is not the case. It seems the car, whilst the same, suits one driver's style more than the other....
It depends if the bike is on a conveyor belt....
Kryton57 - MemberWe are assuming a virtual terrain here, same across the board.
exactly, what are the conditions?
you'd have to start by establishing a 'terrain roughness' measurement, an Ra value of sorts.
(there's much more to roughness than Ra, but it's a start)
and then you'd have to quantify grip,
and then, etc.
Exactly ahwiles. Assuming the same (whatever you deem that to be)... Surely its not too difficult for the heavy hitters on here to establish a base and go from there?
My head hurts already...we'd have to agree on the type of virtual course (ie gradient, technical level, soil type, rootiness etc), but I do get the X miles for Y energy type equation which does exist but in a very simplified form and thus doesn't take into account those game-changing variables that will extend your ride like saddle comfort, geometry, tyre choice ad infinitum
I keep coming back to the notion that I should ride both of my bikes back to back around my local loop on an "ideal" summer's day (ie dry, dusty and at roughly 20 degrees) and then report back. I'll revisit this thread in June with my findings.
If it wasn't too difficult I'm pretty sure there'd be data all over the web for it (I think some manufacturers, e.g. Giant, claim that they've "done the math" but then don't present any data!)
You could probably have a stab at some ideal model that involved a bike being rolled along a course with a constant input of power, but it wouldn't be very representative of real world riding. Reckon there's just too many variables with all the rider input (weight shifts, pumping, hops, etc.) and being on and off the gas, different types of surface, roots, rocks, etc.
What colour is the 29er?
For the purpose of this thread, assume they are both pink camo.
[img] 
[/img]
The problem is that the answer will depend so much on your assumptions and model and neither will likely be that close to real life given that mtbing by definition is over different terrains - eg you could probably make the answer whatever you want just by adjusting the parameters.
On paper, most would say that Vettel is a better driver than Ricciardo
Not me - this season has shown what a useless lump Vettel really is, and how lucky he's been previously to get a car which was so good and exploited his peculiar driving style.
As for Kryton's question - absolutely no way to do this as a paper exercise without lots of real world testing.
assume they are both pink camo
Aaah, so we're riding in a ladies under garment section of M&S and being stealthy. Well, at least we know the terrain.
And of course, crucially, the only "Improvement" you get from a lighter bike is when you are accelerating or climbing. So a flat course, that you cycle around at a steady 20mph isn't going to help the 26er at all.
In fact, in my experience, on typical XC racing courses, you don't actually accelerate that much, and speed is surprisingly constant for the vast majority of the time.
seeing as 29'ers are for novice riders suspect that the 29'er would have to be the lighter bike.
And how do you factor in the exploding wheels?
You would have to start with a very simplified model and then attempt to add the various parameters later to build up a more useful model. But I haven't even seen the simplest model yet ie. flat ground 26" v 29" at constant speed. So I think we're a long way off a meaningful mathematical model, but I would have thought the big manufacturers would have the resource to develop one. But if they did it would not be in the public domain in the same way that F1 dynamic vehicle models are not freely available.
In short I'm pretty sure it could be done to a level that would be meaningful. Could make a good phd project for somebody.
I wonder if for bigger riders if a lighter bike makes less difference too as it's a smaller overall weight saving? Probably hurt the rollability of the 26er too more so than the 29er. So if you're a portly gent then you'll probably be better off with a 29er 🙂
If you're talking simple models, flat ground and constant speed then there won't be any difference between 26, 29 or any other wheel size.
Look at that bloke's (Matt?) test where he saved however many seconds over a certain loop.
Then try calculating that into a weight gain for a 26er.
That'd be a start
[img] 
[/img]
One of my favourite books (Sorry). You need some clear definitions. Level smooth surface for rolling resistance and effects of contact patch, constant power for rider (otherwise the rider can increase power to negate any difference).
I don't think weight per se will be a huge factor here for these design parameters (flat), so you'd probably want a 3, 6, 9, 12% gradient for starters.
Addition of surface terrain is probably a harder one, so you'd need some form of input for non-flat surface, say a sinusoid with period of 2.5, 5, 10, 20 cm to get some effect for wheel size.
Should be easy 😉
If you're talking simple models, flat ground and constant speed then there won't be any difference between 26, 29 or any other wheel size.
Correct. So that's your starting model.
Next step would be to add in "number of accelerations", some way of quantifying the time difference due to getting back up to your constant speed (at the start, after tight corners etc). That would tip things in the 26ers favour. For the basic model above, there are zero "accelerations" so 26=29.
After that you might add a "rollability" factor, i.e. the gain for a larger wheel over a given "roughness of terrain". Again for a smooth course 26=29, as roughness increases so does the 29er advantage.
That would get you a reasonable way towards modelling a twisty, rough course at a constant altitude. If you wanted to try and track weight, you'd need to add in a height variance over the course distance.
It's definitely do-able, but would get complicated quite quickly. Also, it would be very difficult to come up with the equation constants (e.g. by [i]how much[/i] a given 90 degree corner slows down a 29er vs a 26er). Without a real-world testing ground (unlikely) it would be quite hard I think...
Of course it's doable, it's just IMO pointless as I said - you'll get whatever answer you want to get unless you're going to put it all into a supercomputer with a huge amount of data gathering beforehand and also some actually agreement on what 'typical' riding actually is...
Imo.. a very difficult if not futile question even if you could create a complex enough model to be representative of enough reality to be of interest. Wheels and weight are inter related and lighter is rarely simply better (within normal bike ranges) it's always a trade off. Look at both parameters in isolation for benefits then figure out the balance of the two.
eg, my most efficient bike over a longer distance happens to have heavier wheels than normal, I see that as part of its advantage and I'm in no rush to change that aspect.
If you're talking simple models, flat ground and constant speed then there won't be any difference between 26, 29 or any other wheel size.
Well no. If you're going that simple, then everything else being equal the larger wheel will be faster.
agreement on what 'typical' riding actually is...
[img] 
[/img]
One of these will be fastest in that scenario
[img] 
[/img]
I'm assuming a rigid wheel (eg no tyre), aracer - simple - hence no difference.
Though come to think of it, even if you do assume a tyre and hence a contact patch, you've got the question over tyre pressure then between wheel sizes. Then maybe a question on pressure vs grip. And pressure vs pinch flat risk. Then tyre size. Then suspension travel. and it goes on and on...
The issue that throws the spanner in all the maths is human ergonomics. Cycling is much more about the rider than say the f1 driver. The human element is more dynamic and variable in riding style so the model has to be so specific as to be near impossible.
Either you find the answer yourself through trial and error or you won't know, even then the answer may be more placebo than anything.
[quote=nemesis ]I'm assuming a rigid wheel (eg no tyre)
Well that's pretty pointless and unrealistic then.
even if you do assume a tyre and hence a contact patch, you've got the question over tyre pressure then between wheel sizes.
I was assuming equal tyre pressure, which seems like a good place to start, and no obvious reason to do otherwise given the same tyre type and width.
One interesting possibility does occur to me to get some answers to this. The availability of e-bikes does provide the opportunity to get rid of one of the most variable factors involved in real world testing - though clearly it is still not that simple to keep everything the same.
To further muddy the already murky waters of the discussion, you also need to factor in - if we're talking full suss mtbs here - the pedaling efficiency of the suspension system.
My AM bike is a pound or two on the heavy side, but it climbs extremely well because the suspension design is so good, I never even need bother to lock the rear shock out.
Any flat or inclined fire trail where there are imperfections over, say 20mm size, will always IMHO favour the bigger wheels rolling ability.
My view of this was to eliminate as many variables as possible. So same "virtual" rider, same bike components with the exception of the frame & wheels, same virtual course etc....
The arguments above are introducing variables, so how about a simulation which elimates as much variability as possible...
But then you'll have an riderless ebike rolling around a velodrome in a vacuum. What's the point of that? Not really representative of the riding I do 🙂
Eh? Why? Why wouldn't you have a virtual trail, with bumps, hills and curves, not a velodrome track?
If you're talking simple models, flat ground and constant speed then there won't be any difference between 26, 29 or any other wheel size.
There would still be a difference in power output (however small) to maintain the same constant speed as there is more energy stored up in those larger wheels at constant speed. But yes, the model would need to be developed much further to be of any practical use.
on flat ground, smaller wheels are faster - they're more 'aero'
Kryton57 - MemberEh? Why? Why wouldn't you have a virtual trail, with bumps, hills and curves,
because that stuff is practically impossible to define, measure and model.
Define the virtual trail for us and once we all agree that it's representative of what mtbing is, we'll do the maths... Shouldn't be difficult to get agreement on that, should it? 🙂
Define the virtual trail for us
A topographical map of a typical trail centre should be okay e.g. Afan Skyline
Once you have your bike model sorted, the trail would be a relatively easy parameter to change. You could run multiple trail maps as they do in F1.
A topographical map of a typical trail centre should be okay e.g. Afan Skyline
That tells you very little about the actual trail though.
moshimonster - MemberDefine the virtual trail for us
A topographical map of a typical trail centre should be okay e.g. Afan Skyline
Once you have your bike model sorted, the trail would be a relatively easy parameter to change. You could run multiple trail maps as they do in F1.
just checking, you do realise that when we talk about a 'model' we're not talking about a CAD model, right?
[quote=ahwiles ]on flat ground, smaller wheels are faster - they're more 'aero'
Ah, good point. I was completely ignoring aero, as it didn't seem all that relevant - the differences will be tiny compared to other factors.
Is that typical riding? Fireroad climbs, singletrack descents?
And the surface is generally pretty smooth at trail centres, with roots/rocks then thrown in. How does that compare to mud on natural trails?
I'm not trying to be awkward, just saying that there are so many variables that aren't clearly defineable (in a way that people will agree on) meaning that the results are pretty meaningless beyond a bit of a mathematical exercise.
This is why F1 teams invest loads of money on computers but still have to do testing as reality often doesn't match.
aerodynamic drag is hugely significant, even at relatively low speeds.
on a flat smooth trail, you'd be better of with tiny brompton wheels.
Ah, good point. I was completely ignoring aero, as it didn't seem all that relevant - the differences will be tiny compared to other factors
Depends how fast the rider is and what they're riding (the fireroads for example) maybe the pros should be on 26"? 🙂
Right, as an engineer who has done modelling of things, I'll weigh in with my serious opinion. Yes I think it would be possible to do, but it would be a lot of work to do properly - back of a fag packet stuff would be pretty much worthless. You would also need to do some real world testing of rolling resistance.
At which point you may as well just do proper testing rather than the maths.
[quote=ahwiles ]aerodynamic drag is hugely significant, even at relatively low speeds.
on a flat smooth trail, you'd be better of with tiny brompton wheels.
Maybe, but unless what you're actually interested is a flat smooth trail, then you're better off looking at other factors - aero drag certainly isn't a significant factor at MTB climbing speeds, whilst wheel rolling resistance is.
aracer - Member
Right, as an engineer who has done modelling of things
And that was the best you could come up with. Having an off day? 😉
ahwiles - Memberon flat ground, smaller wheels are faster - they're more 'aero'
...
ahwiles - Memberaerodynamic drag is hugely significant, even at relatively low speeds.
At mtb speeds, with mtb tyre drag? Are xc racers using aero aids?
What about the DH and flats, aracer?
[quote=nemesis ]At which point you may as well just do proper testing rather than the maths.
No, not really, because as discussed there are too many variables which need to be eliminated in real world testing - even if you're taking my e-bike suggestion, just getting equal power outputs with that would be enough of a challenge.
Absolutely but then the same applies to trying to come up with a mathematical model of everything (or even just some stuff).
🙂
I think we're all agreeing FWIW...
[quote=nemesis ]What about the DH and flats, aracer?
Aero might be an issue, but the aero differences between different wheel sizes won't be (the wheels themselves are a small component of drag, the difference between wheel sizes even smaller).
Of course, the point isn't really that, it's that there are so many variables that in themselves are affected by other variables which come from whatever model you choose.
[quote=nemesis ]Absolutely but then the same applies to trying to come up with a mathematical model of everything (or even just some stuff).
The difference being that you can keep the variables constant when doing mathematical modelling. Sure there's lots of stuff you have to define, but there are good reasons people do mathematical modelling rather than just real world testing everything.
Are xc racers using aero aids?
Not so much on the bike, but lots of pros wearing skin suits and most seem to be adopting a more aero position on the bike these days too. Lots of aero tucking in DH too, and they'd all be wearing skin suits if they weren't banned.
Of course there are - you don't get muddy for a start 🙂 There are also situations where modelling just isn't that useful as there are too many variables. Designing a carbon fibre frame from a strength/etc perspective is great for modelling - generally the parameters are fairly well defined. Designing a frame to 'handle well' is not a great use of modelling.
Given the debate on here about what consititutes proper riding, good handling, goov bikes, whatever, it's clearly a minefield and that doesn't even start to consider the differences between people, conditions and so on.
just checking, you do realise that when we talk about a 'model' we're not talking about a CAD model, right?
No, I'm talking about a vehicle dynamics model like we use in F1 (notice the we).
The issue with this model is that in order to be accurate it will have to be complex. And that will require the modeller to make a lot of assumptions if they don't want to spend years testing and working out the individual effects of numerous small parameters first.
As with any simulation, it's output is only useful once correlated with the real world, up to that point it's just maths. For example 2+2 = 4 as everyone knows, but 1+2 = 3. Both are mathematically correct, but which formula represents best the real world? (you'd have to work that out for your particular case)
We do a lot of simulation work in the automotive/race world i work in, and it's all meaningless without correlation.
So, you'd have to test wheels for rolling resistance vs normal load, tyre type and pressure and bike geometery / set up on numerous types of terrain, then do the same for bike of different masses distributed in different places just to be able to say that "once a smooth surface gradient exceeds X, you are better off with a lighter 26" wheel" and "once a rough surface gradient exceeds Y you are better off with a 26" wheel" etc etc
Not worth the effort, just ride your bike and enjoy it!
That tells you very little about the actual trail though.
It would give you the macro terrain, which would be fairly important in assessing the effect of rider weight. I doubt that the micro "bumpiness" of a typical UK trail would make much difference when looking at the OP's question.
Isn't the micro "bumpiness" where the advantage of a 29eris assumed to lie though?
But for example is a trail with medium size roughness the same as a smooth trail with rocks of size y at x spacing?
@maxtorque - I don't think it would have to be that complicated a model to answer the OP's specific question. A lot of the variables (in fact all of them except wheel size) would be identical.
Isn't the micro "bumpiness" where the advantage of a 29eris assumed to lie though?
Not just that, they tend to ride faster on smooth trails too. But you could assess the bump rollover performance from a pretty simple square edged or ramped bump input on your trail. Again not that hard to model.
I'm with moshi here. It would be complicated, but definitely doable - they model more complex stuff.
But for example is a trail with medium size roughness the same as a smooth trail with rocks of size y at x spacing?
No, which is why you would need to develop a dynamic bike model to test different trails. The former doesn't appear to exist in the public domain, so it's a mute point.
OK, so a reminder of the actual question:
How many pounds lighter would a 26er need to be to offset the increased "rollability" of the same frame with 29" wheels?
So while we can come up with answers, there'll be as many answers as there are combinations of rider, terrain surface, elevation/twistiness profile, etc.
So yes, we can come up with an answer but does it really mean anything useful? IMO, no, not given the variability of what mtbers actually ride.
PS MOOT point 🙂
on a flat smooth trail, you'd be better of with tiny brompton wheels.
Road and Velodrome bikes tend not to have small wheels, so I reckon you are wrong there.
they got banned.
UCI reg's innit bro: 700c wheels or it doesn't count.
here you go:
[img] 
[/img]
photo taken roughly 5000 years ago, when innovation was still allowed.
although, it should be said, i think these bikes are Moultons - made in Bradford on Avon - where me Ma' was born.
(not Bromptons)
So while we can come up with answers, there'll be as many answers as there are combinations of rider, terrain surface, elevation/twistiness profile, etc.
That's true and why modelling is useful. Back to the F1 analogy, you have a vehicle model and lots of different tracks to get specific answers for specific tracks. But you also usually find common trends as you may well with 26 v 29.
I suspect 29 is theoretically faster than 26 for most real trails hence why XC racers are all on 29. When the trails get seriously gnarly it seems to swing in favour of the smaller wheels, but not sure for how long? In theory I would have thought 29ers should be killing the DH too, but maybe the practicalities e.g. wheel stiffness have not got there yet. Plus the current crop of DH riders are used to 26 of course.
photo taken roughly 5000 years ago, when innovation was still allowed.
Very interesting, but was it actually any good? Doesn't make sense to me as an engineer. If I was designing a velodrome bike without regulations I'd be looking at larger wheels, not smaller.
Edit: I guess I can see the advantage if you were sprinting from a low speed at the end
but larger wheels have a bigger frontal area, and that means more drag - and on a flat smooth track, drag is everything.
but was it actually any good?
yes, very, that's why they got banned, as opposed to simply being allowed to die out as a failed experiment.
How many pounds lighter would a 26er need to be to offset the increased "rollability" of the same frame with 29" wheels?
Obviously more than anyone (including world champions factory teams) is willing to spend to achieve.
End of.
going back to the e-bikes for a second...
This could actually give a fairly interesting opportunity for some real world testing.
26er and 29er versions of same bike (ie same material, as near as possible geometrically given the wheel sizes, ie: same intended use), standardise the tyres/pressures, if they are [i]entirely [/i]motor driven*, no pedals, and then use a large sample of test riders to go round the test loops numerous times on each bike and under different conditions.
This should give a big sample spread of rider ability, conditions, weights etc. and if no pedal input allowed you have a relatively easy way to measure and report on energy usage.
You could then start to vary individual elements, like tyre pressures, tread, biek weight, rider weight etc.
The key to it would be using a large enough sample size of riders and test tracks, but I bet you could build up a very interesting set of data.**
Would hopefully give you some indication of power output differences and energy expenditure but in close to real world riding conditions with real riders, but removing the variability of the human engine.
*might have to think carefully about things like a max instantaeous power cap, and possibly max duration at certain power to stop people literally just opening the throttle, but its all in the details.
**still might not prove anything but would be interesting either way!
but larger wheels have a bigger frontal area, and that means more drag - and on a flat smooth track, drag is everything.
I get that, but the total frontal area of the bike/rider doesn't look any different to me. I would have thought the pros of a small wheel are:-
low mass
low inertia - easier to accelerate
The smaller velodrome wheels' advantage wasn't actually the wheels. It was that you could ride much closer to the bike in front which reduced drag - so funnily enough, it was smaller wheels to allow the main aero drag generator (the riders) to generate less drag...
It should be added that in the pic, they aren't really making much use of it though!
How about we just arrange to get two 2014 Specialized Enduros, in 29er and 26er flavour and attach some gps gear to them before sending them around a short circuit. After the first lap you swap the riders around and then do two more laps with another two riders with a half time swap each?
It's scientific, innit...