Friction *regen brake* sanity check

[BalorNG]

I'm talking about your general conclusion that 'adding a motor to a bike is useless for long trips'. It is most emphatically not - even a simple a motor with an overrunning clutch plus a decent battery will be able to compensate much more than it's weight over VERY long distance, you just need to use it sparingly on the hills, where application of additional power is critical for time and anaerobic effort saved.

I don't understand why people here think physics doesn't apply to them. Please run the numbers below and let me know if you either come to the same conclusion, or what physical constraint you are ignoring.

Flat:
Let's rough your weight to 100kg and your contraption with 2kWh pack to 70kg. Your CdA 0.092 is in the range of recordholding non-electrified HPVs and just not realistic. It will be a miracle to get that to 0.138.
Source: BROL

Yields 6.8Wh/km * 600km = 4080Wh.

Compare to a non-electrified velomobile around 30kg = 130kg total, and easier to achieve CdA 0.092 = 5.2Wh/km * 600km = 3120Wh.
4080Wh - 3120Wh = 960Wh. It takes you 960Wh extra to lug around all that nonsense for your whole trip.

Wait what? Now you add air drag to into equation?! When it comes to motor/battery? Are you serious?
Plus, 'recordholding' practical HPVs is 0.04 - I've seen strava posts with close to 60 kmh (absolute flat of course) on 220w.
Recordholding 'nonpractical' HPVs is less than 0.01, but now that is indeed an impossible goal.
Of course, you might as well be right, that would depend on how I manage to pull off making of the shell, but battery emphatically is NOT a factor in air drag.
Plus, by splitting the weight in a very different manner from what is real (not that I am proud of being obese, but that's one of critical factors in decision making process!) you further stack the 'cards' in your favor. Are you really that desperate to 'win' this argument?

Hills:
170kg * 9.8m/s^2 * 4000m = 6664000J = 1850Wh. If you're able to regen at 50% you get 925Wh back. And that's only fully usable if your destination is at the top of a hill.

Oh, the tired old 'regen braking is useless on e-bike'. Conventional e-bike with aerodynamics of a brick and 'stupid' controllers and motors that lack efficiency - maybe. I expect regen efficiency around 85%, that's what actual measured no-load torque and motor resistance tell me, though how much extra rolling resistance drag will cost me needs to be determined.
Frankly, I see regen argument is kind of 'sour grapes' one due to complexity of implementing a manual clutch - once you have it, it is an absolute no-brainer.

Non-electrified velomobile:
130kg * 9.8m/s^2 * 4000m = 5096000J = 1415Wh.
1415Wh - 925Wh = 490Wh. You save 490Wh going uphill.

925Wh - 490Wh = 435Wh. It still takes your contraption 435Wh extra.

Now what the hell was THAT? Which orifice you've pulled that conclusion from? It takes 'less', not 'extra'. Desperate, indeed.

Your argument about needing to burn 1000W while riding uphill is specious. Cyclists have been downshifting to climb hills for 100 years now to keep cranking away at FT. If your single-track streamliner is unstable at that speed, build it as a trike like most other solutions.

Obviously, these are back-of-napkin, zeroth-order approximations, but they illustrate the point.

Sir, are you an engineer or a politician? You pretend the former, but quality of your (motivated) reasoning suggest the latter.

 

[BalorNG]

- and I'll be using it intermittenly, that also solves that overheating issue

"intermittent" isn't what I would call doing downhill braking. Intermittent would be braking a few seconds here, a few there, with minutes or more in between.

Downhill braking of the type I *think* you want to do is what I would call a continuous-rating type of application.

(plus I'll be using in a most efficient manner, hence - it should only get 'warm' even in continuous operation... ~1kw max on a motor rated to 4kw).

Remember that with RC motors (if that's what you'll be using) that power rating is normally *only* at a very high RPM, with a propeller forcing air over and thru them. If the motor you have is different, then make sure that the power it is designed to handle is under the specific conditions you'll actually be using it under.

You may well have already run all these numbers...but I prefer to bring up stuff that was obvious to someone than not to bring up what someone may not even be aware of, and have their project fail because they didn't know.

It is *highly* unlikely that I'll be able to charge more into the battery than it's max capacity - I'll have to start on a steep mountain for that...

I just wanted to point out the potential for the problem, as people have had EVs damaged by this phenomenon (the first one I heard of was a DIY car conversion, maybe a decade or more ago). That way you know it exists, and can work out ways to avoid ever encountering it.

If you start at the top of a slope with a full battery, then go downhill braking without having used the battery, you *can* overcharge it, or at the least you can cause the BMS (if any) to turn off the input for HVC, which can potentially blow up the controller of the braking motor, due to the current suddenly having nowhere to go, which causes voltage on the controller to sharply and rapidly rise (nearly instant), and that voltage can be higher than the FETs and LVPS in the controller can handle.

Good points, but indeed I've run the numbers (I^2R losses, etc) and it seems quite manageable. Plus, modulating speed on descends and bring the vehicle to a complete and abrupt stop are different tasks and latter will have to be relegated to conventional brakes.

Unlike inrunners, outrunners serve as their own cooling fans a bit even w/o dedicated ones - for same reason Magnus effect operates, and drone motors in particular are made to utilise that effect. Plus, when it comes to friction brake, I wonder if a tire will act as a heatsink?
After all, extra losses, unless your 'roller' slips, happen not in the motor, but in the tire, and hot can will be cooled the be tire, that in turn will be cooled by the air/road.

 

[Chalo]

I expect regen efficiency around 85%, that's what actual measured no-load torque and motor resistance tell me,

Expect around half that efficiency if you use a friction roller. The more braking power you can apply, the more of it you'll lose to friction and hysteresis.

 

[BalorNG]

I expect regen efficiency around 85%, that's what actual measured no-load torque and motor resistance tell me,

Expect around half that efficiency if you use a friction roller. The more braking power you can apply, the more of it you'll lose to friction and hysteresis.

Yea, that's pretty major concern. I've read about 100w of additional drag - while not *small*, it is still 10% of 1000w, hence cannot account for *this drastic* loss of efficiency.
I think I may need to set up a test rig and ascertain exact efficiency figure of a friction drive under full load.

 

[fatty]

Now what the hell was THAT? Which orifice you've pulled that conclusion from? It takes 'less', not 'extra'. Desperate, indeed.

What part of subtraction are you not following?
4080Wh - 3120Wh = 960Wh. It takes you 960Wh extra to lug around 40kg of electrification over 600km (aero).
1415Wh - 925Wh = 490Wh. You save 490Wh on hills with 40kg of electrification.
925Wh (extra) - 490Wh (saved) = 435Wh. It still takes your 70kg contraption 435Wh versus an equivalent 30kg bike.

Your argument about needing to burn 1000W while riding uphill is specious. Cyclists have been downshifting to climb hills for 100 years now to keep cranking away at FT. If your single-track streamliner is unstable at that speed, build it as a trike like most other solutions.

Obviously, these are back-of-napkin, zeroth-order approximations, but they illustrate the point.

Sir, are you an engineer or a politician? You pretend the former, but quality of your (motivated) reasoning suggest the latter.

I'm studying to be an engineer -- I have no horse in this race. Show me objective math that demonstrates carrying 40kg up 4000m over 600km is more advantageous than 100 year-old derailleur.

 

[fatty]

Of course, you might as well be right, that would depend on how I manage to pull off making of the shell, but battery emphatically is NOT a factor in air drag.
Plus, by splitting the weight in a very different manner from what is real (not that I am proud of being obese, but that's one of critical factors in decision making process!) you further stack the 'cards' in your favor. Are you really that desperate to 'win' this argument?

It isn't an argument because you haven't presented a cogent response.

As I already explained, I'm not directly deriving CdA from your batteries. I'm deriving CdA from the complexity and inefficiency of your construction resulting from the decision to electrify (tube clamps? bonding tube joints with carbon fiber?). That is, you claim 0.046 theoretical and allow for practical implementation to be double that, 0.092. I think a simpler, non-electrified build without the 40kg of infrastructure necessary for 2kWh of electrification and multiple motors might get you to 0.092, and that you'll be lucky to hit triple your theoretical, 0.138, with a 70kg clamped-tube frame. I think that's a generous estimate for a first build without fabrication ability.

Post up your model in Flow Simulation showing 0.046 at 11m/s. Show us how to fabricate an advanced airfoil without basic fabrication ability, and how to achieve greater than 50% friction regen.

This is only half-rhetorical -- I don't think you're being realistic, but I'm also here to learn too, and if you were able to pull something like this off -- again, without fabrication ability -- it would be... well, like I said, a miracle. It would be extremely impressive, and I expect we would all learn a great deal.

I'm genuinely not trying to be negative, but I'm shaped by my experience on an engineering race team where everything was modeled before claims were made. And many users here post wild claims that simply aren't grounded in reality.

 

[Steph]

Frankly, I see regen argument is kind of 'sour grapes' one due to complexity of implementing a manual clutch - once you have it, it is an absolute no-brainer.

Ahh, you made my day
I tried to have a similar discussion and I ended up with the same feeling and being kinda discouraged to pursue, despite having the opportunity of engaging with otherwise smart and experienced people.

Once you have a correct clutch mechanism in place, the only price you pay is the weight of that clutch mechanism (plus in your case, the motor that will be used as an alternator since you specialise it) to carry uphill. The overall should fit within 500g.
On the other hand, it will help converting power that is proportional to the total mass of the ride when dragging downhill. That would be 170kg

There can be a lot of debate as to the efficiency of the conversion, but from mass perspective, the ratio is 1/340 so overall efficiency just needs to be better than that (which is 0.03%).

To reduce losses related to tire deformation, have you considered rolling the 'drive' on the rim?
Much like velogical does, for both their drive and dynamo. Maybe the grip will be less but I assume you just want to create a drag effect (sustained but not that strong a braking) so this might not be an issue.

By the look of it, you seem more determined than me to carry this out short term, so I'll definitely be interested in your progress.

 

[BalorNG]

4080Wh - 3120Wh = 960Wh. It takes you 960Wh extra to lug around 40kg of electrification over 600km (aero).

Because, like I said, a conventional velo is out of the question for many reasons, and if am to build diy one, it would drag even more than that! *sigh*
Compare DIY, unpowered oranges to [strike]clockwork[/strike] electrified oranges, not to fractal cucumbers please.
Unpowered bent would be build to absolutely same spec, just without a motor. Maybe I would build something like you propose (much lighter and more aerodynamic)... eventually, but I'll need more build experience AND lose 40 kg myself. And learn to like climbing... something that eluded me for years, and get worse if anything.

Your argument about needing to burn 1000W while riding uphill is specious. Cyclists have been downshifting to climb hills for 100 years now to keep cranking away at FT. If your single-track streamliner is unstable at that speed, build it as a trike like most other solutions.

Obviously, these are back-of-napkin, zeroth-order approximations, but they illustrate the point.

Sir, are you an engineer or a politician? You pretend the former, but quality of your (motivated) reasoning suggest the latter.

I'm studying to be an engineer -- I have no horse in this race. Show me objective math that demonstrates carrying 40kg up 4000m over 600km is more advantageous than 100 year-old derailleur.

Oh, unfortunately you do - you have brought a lot of *value judgements* into this topic that were neither asked for NOR wanted, and you let them influence your reasons (heavily) - so it all went the good old 'someone is wrong on the internet' debate way... hence the 'politician' remark. Values before facts - that's how the worst things in history began...
Hardly surprizing turn of affairs, but sad nonetheless.

 

[BalorNG]

Post up your model in Flow Simulation showing 0.046 at 11m/s. Show us how to fabricate an advanced airfoil without basic fabrication ability, and how to achieve greater than 50% friction regen.

This is only half-rhetorical -- I don't think you're being realistic, but I'm also here to learn too, and if you were able to pull something like this off -- again, without fabrication ability -- it would be... well, like I said, a miracle. It would be extremely impressive, and I expect we would all learn a great deal.

I'm genuinely not trying to be negative, but I'm shaped by my experience on an engineering race team where everything was modeled before claims were made. And many users here post wild claims that simply aren't grounded in reality.

For someone not trying to be negative you are not doing a particularly good job, to be frank.
But here:

https://www.simscale.com/projects/OTsyganov/kervelo_fairing/

That's symmetry of half-model, at highest resolution I could reasonably run the test (spend lots of CPU hours to make sure how better resolution affects results, the drag is a bit higher not nothing drastic).

Of course, you cannot expect a model w/o any holes and a head faring (I'll add it on later) to accurately represent real world drag values, hence I use 'double that' as an estimate, that meshes well with 'middle of the pack' velomobiles, which my 'contraption' is about the size of, but should have better aerodynamics due to less wheels and more refined (I hope) airfoil shape.

The shell will be manufactured with an EVA foam sandwich (very thin cloth), that is just stiff and and light enough to provide good aerodynamics, some vibration damping and heat isolation.

Here is a similar project, using 'bare' foam:
http://www.recumbents.com/mars/pages/proj/tetz/manual/1mold.html

I'll have a the plug wire cut though, that should allow good accuracy - I already have a contact to do this for a resonable sum, but due to 'reasonable sum' (that I can afford) other project of their take precendents and this takes a while - COVID-related stuff, at first they had to shut down for half a year, and now they are permanently overwhelmed .

Here is a test run of the nose fairing from a 'scrap piece' of foam - look quite reasonable, minimal post-processing required.

 

[BalorNG]

Frankly, I see regen argument is kind of 'sour grapes' one due to complexity of implementing a manual clutch - once you have it, it is an absolute no-brainer.

Ahh, you made my day
I tried to have a similar discussion and I ended up with the same feeling and being kinda discouraged to pursue, despite having the opportunity of engaging with otherwise smart and experienced people.

Once you have a correct clutch mechanism in place, the only price you pay is the weight of that clutch mechanism (plus in your case, the motor that will be used as an alternator since you specialise it) to carry uphill. The overall should fit within 500g.
On the other hand, it will help converting power that is proportional to the total mass of the ride when dragging downhill. That would be 170kg

There can be a lot of debate as to the efficiency of the conversion, but from mass perspective, the ratio is 1/340 so overall efficiency just needs to be better than that (which is 0.03%).

To reduce losses related to tire deformation, have you considered rolling the 'drive' on the rim?
Much like velogical does, for both their drive and dynamo. Maybe the grip will be less but I assume you just want to create a drag effect (sustained but not that strong a braking) so this might not be an issue.

By the look of it, you seem more determined than me to carry this out short term, so I'll definitely be interested in your progress.

Point is, I can kinda see why it makes no sense for a 'conventional kind of e-bike' - using a motor to climb hills is fun, but than dragging the brakes downhill while *pedalling* as though on a flat is hardly fun unless you want to eke out the last bit of efficiency from the system, which is not usually a high priority for a typical e-biker. If you think about it, regen braking downhills should actually be *more* important for something not aerodynamic: on a velo you can 'porpoise' rolling hills pretty effectively even when purely human powered I've been told, but that still require a specific set of circumstances, while hitting even 'mere' 40 mph on a downhill on something upright is like dragging a chute (I've hit 72 kmh on one of my bents once, and the wind nearly ripped the helmet off my head!)

I like solving 'optimisation puzzles', hence I don't mind at all, plus I should have plenty of fun 'bombing downhill' even while extracting a lot of power back out of the system, and while our area is not 'montaineus', there are still *plenty* of hills when it comes to 400+ km distances... And after 300+ kms in, you really get to *dread* the climbing part, when your legs are already nearly shot, so you have to crawl those uphills, and than waste that energy on air drag downhill. Like I said, not fun at all.

Btw, just recently it was brought to my attention that there are *hugely* effective motors our there that also happen to be relatively cheap and actually pretty light:
RC *inrunners* from large scale models.

https://www.aliexpress.com/item/4000127447150.html

I mean, look at this beauty! That's like Astro motor at quarter of the price!
Unfortunately, I'm not exactly sure that those specs can be believed, but it looks *reasonable* at least... would still require double reduction though, and I'm sure that it would sound like a dental drill in operation.

I DID think about rim drive, but that would require a new wheel with v-brake rim and a system to actuate the motors against it - possible, but much harder in to implement given my limited abilities, plus I'll need to buy even more motors... and again, there would still be friction and hysteresis at roller/rim interface.

An other of my tentative ideas is to use a front drive with Sturmey Archer fixed gear hub - one of 'unintended' consequences of it's construction that it has a 'neutral gear'!
Unfortunately, it is rare and expencive, and it is said that finding this 'neutral' gear is fiddly (and partial engament might brake the hub), otherwise it would be a perfect fit I guess...