Downhill project motorcycle transmission

[gwhy!1]

Gwhy, are you saying one could run both controllers for one motor? To have the advantages from both controllers?

So at low RPM I use them hall sensors and a regular controller - that take care of all stand still launch issues and let you launch by throttle from dead stop.

Then as speed and RPM rises the "launch" controller falls back and the RC esc takes of and now run the motor without any care for the hall sensors.


This would actually solve many problems. It would be easier to use gyro sensors to control acceleration and avoid wheeling. The sensored controller would work far from it's limits because it would not be used to push max current etc as it only needs to work trough part of the motors band width. It might even let you use a lower cost or smaller foot print controller as the rc esc takes over anyway. That could meanie you might get away with the adaptto mini rather then the double sized and double prized maxi. So now you can use the smaller and cheaper of the two adaptto's.

I think the Achilles heel would be a seamless fallback/takeover. But that should be solvable. It might be so easy as to use a relè and when a certain RPM is reached power up rc esc and power off the sensored controller.

No .. I cant think of a situation where you can run 2 off the shelf controllers driving the same motor without also having the need for addition electronic hardware...
There are some e-bike controllers that already do this .. start in sensored/non-sensored mode then from x amount of rpm the sensor signals are ignored so then becomes sensorless.

Edit: one major plus that could be done is with rc controllers and that is to just add sensor inputs to the controller and in the firmware just have the option to use them or not... this would one of the problems solved for ev use...
Also the use of a device such as a CA that can make open loop control become a closed loop control and this will take care of the other issue with using a rc esc on a ev

 

[panurge]

Ecobogan has yet noted during the thread, that the only reason to keep the clutch would be to have a smoother sensorless control at low rpm....
Actually most of the Astro users with a Recumpence drive, run them with a slipper clutch (me included), If you keep it a bit lax, it could act like an automatic clutch exactly like you figured out. That should be perfect, were manually acted and not automatic....but A standard disc clutch, even the smaller, would be overweighted I think, for this application. Having a manual clutch, would even open these systems to the regen world, ideally....
However, when you're used with these systems, you would keep the torque limiter tight just enough to almost never act, if not to avoid transmission's disasters as consequence of any fault (motor chain derail/pedal chain derail/freewheel fault etc.)...

 

[Ecobogan]

Well chaps t'would appear as though the torque through the drive system has cleft the hub in twain creating quite some mischief about the rear end. Seemed like such a marvellous afternoon to venture forth and practice ones wheelstands, being my birthday and all.
I'd take that as a good sign, if one were to be glass half full about it. Far from the most awesome downhill hub but a downhill hub nevertheless, a Bontrager King Earl. Twisted clean in half.
I was up the back of the creek tracks and had to power mono over a boggy rut across the track, the track was super marshy due to some overflowing drain, and just ended up 12 oclocking and lost it.............again. Not a full flip but more a side lose and kind of like what I was trying not to do, nothing major but ended up in the cessey crap and shite just the same, bike and all. Nearly held it but was just too vertical to work which prompted me to get stuck into some old fashioned wheelstand practice on the way back home.
It was good fun and starting to get a go on, doing repeated wheelies then the back end just crumbled. Quite funny really.
Up until then the bike has been running strong and is seriously quick through the single track and any short straight work,with more dialling to do, you should come and have a go.

 

[cheekybloke]

Hi mate
how much power did you throw at the hub when it exploded?( i would wall mount it for posterity )
I am just building a strong rear wheel for a high power mid drive and attempting to use the freehub for a 3 speed set up.
The hub i am using is a Chris King with the stainless freehub and will also be going with stainless chain and sprockets.
The only other hub i would use is a DT hugi as it has manly engagement too
Darren

 

[Ecobogan]

Not too sure on the wattage as the meter tops out at about 6200w so somewhere over that, perhaps 9 even spikes of 10? kw.
I'll look in to that hub of yours, do need one,and some spokes and a better rim.

 

[gwhy!1]

I have never seen a hub go like that before !!! well you learn something new every day.. I have seen plenty of exploded freewheels and freewheel hubs but thats a first for me.

 

[Ecobogan]

Me either and it doesn't make a whole lot of sense. It couldn't twist that far with no broken spokes I'd imagine. The cast alu may have work hardened somewhat when I lightly bent the hub in an early chain derailment.
The hub should be designed to handle more braking torque, even though it's a rear, than what the motor could produce, surely? Maybe it was a dud

 

[toolman2]

Me either and it doesn't make a whole lot of sense. It couldn't twist that far with no broken spokes I'd imagine. The cast alu may have work hardened somewhat when I lightly bent the hub in an early chain derailment.
The hub should be designed to handle more braking torque, even though it's a rear, than what the motor could produce, surely? Maybe it was a dud

Again Ecobogan, you have excelled.
having a look at your pic there are milled sections where it broke, then search for pics of your hub and you get this: http://www.ebay.com/itm/Bontrager-32-Hole-10-X-135mm-8-9-10-Speed-Hub-Body-6-Bolt-Disc-Rear-Hub-/111505945617

It broke because it was designed by a marketing expert literally the carbon fiber tube on that ebay one was shoved through and exposed by milling right through a structural part of the hub, just for appearance then logo stuck on -highly amusing! :lol:

I know ive suffered from hub tube flex like this, because i was breaking far more spokes on the sprocket drive side -it can be from no other reason, just enough twist so that half your spokes dont share the drive load -the bigger the (minimum) hub 'tube' the better for this stuff..

And im not sure if Miles is into this type of thread, but im sure he would agree that its not the watts at the meter at play here, its the torque.

This kinda stuff is pure gold, its so much fun and so much more to be learnt by finding the true limits of things in this way, im definitely enjoying this thread more than a theoretical and eternal debate with circuit about the perils of these low inductance motors.

 

[ksithumper]

Excellent piece of field analysis to determine the rear hub safety factor Ecobogan! (Seems it's less than 1 )

Glad you weren't hurt, it is a timely reminder to us all that once you have a few kW and/or km/h, it can all get quite serious.

Further good news is that your chain and freewheel aren't (yet) the weakest link. Find a stronger hub AND a good wheelbuilder, and then let us have some more video please!

 

[Ecobogan]

Yes I noticed the hub design when I first bought the bike and just figured 'they' would have catered for it's exposed centre some other how, it being a downhill hub and all that.
It certainly didn't let go slowly and those centre 'spokes' are missing altogether. Well done Astroflight.
CK does make a nice looking hub as do a few others, any more suggestions kids? They can run into some pretty sums of cash. Will most likely do the front also due to the weight increase of the bike and am considering an oversize front rotor. A machined to 1.8mm thick 250mm MX jobbie perhaps.
The freewheel wouldn't be under any more load than standard I imagine due to the drive coming off the disc side, the chain being of go kart ilk should take a beating for a wee while longer. There is no noticeable wear on the chain or sprockets as yet.
Will put together another vid soon for sure. Had a great ride yesterday and I really don't mind crashing and wrecking things at speed, it's all part of the field research as you guys mentioned!!

 

[panurge]

Incredible!
I've experimented some time ago with bonded hubs for testing and research purposes during the develop of our alloy-wood, alloy-carbon and Titnium hubs. I may say that hot rotors, not well cooled and not floting have been sometime cause of fault on the epoxy.
One test has been made with the same exact hub (3 pieces bonded alloy without any mechnical coupling) disc and non-disc with 3M DP and the disc faults while the test rim hub is still working ok after 2 years.
That means huge temps can reach the hub body.....sure it is not your case looking at your rotor (more likely a torque and brake torque question), but looking close at the tipical alloy specs I wont really trust a disc hub with that stupid shape and thin wall even on a regular DH bike:

Tensile Strenght (U)
6061 T6:
310 MPa @ 24.0 °C 45000 psi @ 75.2 °F
234 MPa @149 °C 33900 psi @ 300 °F
131 MPa @ 204 °C 19000 psi @ 399 °F

And just for reference:
6061 0:
124 MPa @ 24.0 °C 18000 psi @ 75.2 °F

6061 T4:
241 MPa @ 24.0 °C 35000 psi @ 75.2 °F

7075 T6:
572 MPa @ 24.0 °C 83000 psi @ 75.2 °F
214 MPa @ 149 °C 31000 psi @ 300 °F
110 MPa @ 204 °C 16000 psi @ 399 °F

 

[Drum]

Hi Eco, these high power beasts are definitely overloading standard bicycle components, even full-on downhill ones.

I think the torque that can be transmitted to the rear hub under acceleration is MUCH higher than the braking torque because of weight transfer: When you are braking, weight is effectively transferred towards the front wheel because the centre of gravity is higher than ground level.. with the high centre of gravity of our bikes this is very significant. The back wheel has less force acting downwards on it, so the maximum traction it can develop is reduced and the wheel locks easily, reducing the torque developed.
Under acceleration, the weight transfer is towards the rear, so the back wheel has more downwards force and can therefore develop more traction, so more torque is needed to break traction (or lift the front wheel!!). Therefore the rear hub & spokes can be loaded up with more torque under acceleration than braking so with this issue plus more weight (and higher speed on the flat and uphill) the rear hub is seriously overloaded.

As the top-line downhill hubs are quite expensive, it might be worth going to a machine shop and pricing a special hub to be made.. look at Adam Mercier (BZHWINDTALKER)'s LMX P2 build blog: he has had special hubs made using standard sealed ball races and a strong axle, with a disc brake flange either end so a sprocket can be fitted to one end and the disc to the other end.
It is much cheaper per hub to get several made at once.. Adam has posted on your blog, maybe he would sell you one of his hubs if you asked??
Alternatively a group of ES builders could get together and order a batch of 10 hubs.. between us we certainly have the skills to design it!

I bought a double-disc flange hub from another ES member in USA and it looks good and strong.. we could take dimensions off that but beef up the disc flanges a bit to allow for larger bolts on the sprocket side. Happy to help if anyone's interested.

Jules, Thanks for posting those figures for strength of aluminium alloys at various temperatures, that's really interesting. I knew the strength reduced at elevated temperatures, but hadn't realised how quickly the strength fell away.
Dave

 

[recumpence]

I can attest to the fact that production hubs are fine. I run 3 times this much power through standard hubs all the time without an issue.

He just has a bad hub.

Matt

 

[gwhy!1]

its clear that it was the torque linked with a very bad hub that caused this and not so much the power.. it will also be the same with freewheels.. but i do agree that most hubs should not have a major problem.

 

[bzhwindtalker]

Cassete FW from 12mm DH hub can not take the torque of a 5kw system for long (I had a shimano ZEE pretty much explode), that's why we use custom hubs with either flanged FWs or Iso 6 hole M6 mounts for sprocket on our high power bikes.

We designed a new hub for dual chainlinne and FW on the rigth side. But this design may be tigth to fit into 135mm dropouts...

Matt, I think loads and strains are vastly superior on an offroad bike that will be bashed and abused than on a on-road bike where you can only put so much power to the ground before the front wheel decides to get some air The worst load case Imo would be when re-accelerating in a jump, where you have the wheel spinning full blast and you land at way slower speed, or while doing trials (or trying to )

Adam

 

[recumpence]

I stand behind my statement. He has a bad hub, period.

You will notice the hub broke where it was machined very thin. This was caused more from side loading of the wheel than rotational torque.

I pulled 35,000 watts through a basic no name hub on my yellow trike without any problems. Angular loading broke his hub.

Matt

 

[Ecobogan]

Yeah I understand the weight shift under braking principle . That's why my street bike runs massive twin rotors and 4 pot radials up front compared to the small rotor single pot rear set up ,and all of that.
I was more pointing out that even though it would be very unlikely it is still POSSIBLE for a dh bike to find itself in a situation where the rear wheel could be dealt 100% traction and a massive load. Picture this, A heavy rider coming down from a very big jump and landing near vertical with the back wheel locked on a very grippy surface. A recipe for a good sized crash and an unlikely situation but possible just the same and one that would torque load the rear hub massively, more than I would have thought the motor would.
Anyway for this to be the hard ware for a major company's front line dh bike seems a little odd albeit 6 years ago, A bad design and as Matt suggested other better brands out there would be fine.
If it was more an angular load issue then it must be the weight increase of the bike, near doubled, that'd be a defining factor. As I'm not pushing the bike any harder than a fast downhiller I reckon.
That hub of yours does look rather tasty bzhwindtalker might look into that as well.

I'll just have to keep and eye on the front brakes coping ability. But it's being maxed out for certain, especially now that I'll be adding more battery weight, I think a 250mm rotor and perhaps a 4 pot caliper would do it. Very interesting alloy heat specs too thanks for that mate I'll be referring to that in future. and relevant mpa info on my 'as welded' 6061 frame and swingarm additions, the parent metal being 7005.

 

[Drum]

When I posted this morning I was a bit pushed for time, and I hadn't looked at the photos of the original shape of the hub. Now I have done that, I agree with Matt.. the hub was almost certainly not strong enough to start with, and (unless the spokes were loose) side force was probably more of an issue than torque.
I also missed that for Eco's setup he needs a freewheel and cassette setup, so the twin flange hub idea doesn't match the situation.. I should have waited till this evening when I had more time...
I like the look of Adam's hub with twin chainline / freewheel, but as he says, it might be a bit wide for the standard dropout width.

Glad no injury or damage to other parts, Eco.
Dave

 

[toolman2]

I can attest to the fact that production hubs are fine. I run 3 times this much power through standard hubs all the time without an issue.

He just has a bad hub.

Matt

Your a smart guy and i agree with the above, side loads help break things and that you could also pull more than 35,000 watts without these issues.
Thats because its the torque that hurts drivelines, not ridiculous power numbers measured at the battery wires.

Heres an output power graph from a bike a had like this one, short DH bike, and moto gears, its measured at the tyre and if you pick the kind of spot where these breakages often happen, say 10kph, we only have about 3kw on the Cycle Analist, yet its well over 300NM of torque on the hub and its on route to get a 90kg rider to 50kph in under 2 seconds.

The next photo is the result of these off road capers, spokes broken mostly on the drive side can only really be caused by slight hub flex that you otherwise would never notice, if someone cares for doing the maths you'll find that we are already a fair way towards the tensile strength of the spokes from motor torque alone and when they are driven by the hub at only a 30mm radius (60mm hub diameter) theres a real need for larger hub and flanges to allow us to make use of awesome absorption and load sharing ability of butted bicycle sized spokes.

i agree that a normal/decent DH hub without this design flaw would still be cool for now, there are going to be way higher cycle numbers and peak loads on the hub for this type of machine than anything you would see for say a 20" wheel used at X million battery watts geared for high speed and used on the road, as adam has rightly pointed out, its when you wheelspin over large rocks or land a jump with the throttle pinned -you can prolly ask how he knows this

A relevant question Ecobogan, what is your final ratio when your in 1st gear? and what do we think the peak (say saturated) torque of this astro would be? -if anyone is keen, send me one (im in Australia) i have a torque measurement rig like the one justin used for hubs, and ill post them up here with other motor results and freight it back.

BTW most hub motors cannot make any more than 120NM and the big ones mostly less than 200NM, and thats giving away over 5000w of heat trying. -i hope its ok say that as we are in the non hub section :lol: however with a small geared rc motor you can apparently get far more than that at the wheel with the total loss including ALL drive train losses of around 700w, thats what these machines are so good at, and eventually when you know your only limited by battery capacity its all about trying to keep the CA watts as low as possible to get the job done.

I reckon just get a normal DH hub with decent flange and 'tube' diameter and use butted spokes, or stick another wheel on it and your away.

 

[panurge]

We used to weld 6000 series Alloys to 7000 too for the early welded swingarms, mostly 7000 tubings to 6000 Billets/plates the problem here is that 7005/7020 are AutoT alloy that need more ability to be welded but little or no re-threat, while 6061/83 are not and thus terrible if not retreated, and, as you see from datas, there's an huge difference among the various T states for the same Alloy code.
In fact to Tig these different alloys together would mean easily or to loose the working advantages of one of the 2 or to loose reliability and performances, and is typically due to limits on stocking them in the various forms, or to other required characteristics that makes unsuitable a part of the work to be made with a particular alloy like corrosion and heat resistance, or machineability
interesting that 7075 could be considered AutoT as well, but it is referred obviously not to welding operations (not weldable) but to parts that although subjected to significant heat, may return in a quite similar T state and strength. When a 6061 alloy is subjected to 200+ C° it loose his original T state strength not temporarily but definitively being declassified to a lower T state depending on the amount and duration of heat.

Not sure any kind of side load from a 26" diameter wheel could do that to a small diameter alloy hub before to chew spokes and rim.......sure the poor design + a defect and/or a previous damage ( that's your report of the first chain fail:

Snapped the chain clean in half. Jammed in the caliper mount, bent the disc and chewed up some teeth on the sprocket and also bent the hub.

Totally Agreed Toolman2: may you clarify the total reduction ratios of your 1st gear? from your speed reports and assuming a 26x2.5 wheel and the 4t in star config, I would say about 30:1 that means 210Nm peaks from the targa specs of the Astro 3220 (7Nm max sust torque for 60 sec.)

 

[gwhy!1]

im sure if human power on a trials cycle can make high quality freewheels explode and snap chains/crankarms/axles then its not the power that doing that ( its human power ) so its got to be the torque.

 

[recumpence]

I can attest to the fact that production hubs are fine. I run 3 times this much power through standard hubs all the time without an issue.

He just has a bad hub.

Matt

Your a smart guy and i agree with the above, side loads help break things and that you could also pull more than 35,000 watts without these issues.
Thats because its the torque that hurts drivelines, not ridiculous power numbers measured at the battery wires.

Heres an output power graph from a bike a had like this one, short DH bike, and moto gears, its measured at the tyre and if you pick the kind of spot where these breakages often happen, say 10kph, we only have about 3kw on the Cycle Analist, yet its well over 300NM of torque on the hub and its on route to get a 90kg rider to 50kph in under 2 seconds.

The next photo is the result of these off road capers, spokes broken mostly on the drive side can only really be caused by slight hub flex that you otherwise would never notice, if someone cares for doing the maths you'll find that we are already a fair way towards the tensile strength of the spokes from motor torque alone and when they are driven by the hub at only a 30mm radius (60mm hub diameter) theres a real need for larger hub and flanges to allow us to make use of awesome absorption and load sharing ability of butted bicycle sized spokes.

i agree that a normal/decent DH hub without this design flaw would still be cool for now, there are going to be way higher cycle numbers and peak loads on the hub for this type of machine than anything you would see for say a 20" wheel used at X million battery watts geared for high speed and used on the road, as adam has rightly pointed out, its when you wheelspin over large rocks or land a jump with the throttle pinned -you can prolly ask how he knows this

A relevant question Ecobogan, what is your final ratio when your in 1st gear? and what do we think the peak (say saturated) torque of this astro would be? -if anyone is keen, send me one (im in Australia) i have a torque measurement rig like the one justin used for hubs, and ill post them up here with other motor results and freight it back.

BTW most hub motors cannot make any more than 120NM and the big ones mostly less than 200NM, and thats giving away over 5000w of heat trying. -i hope its ok say that as we are in the non hub section :lol: however with a small geared rc motor you can apparently get far more than that at the wheel with the total loss including ALL drive train losses of around 700w, thats what these machines are so good at, and eventually when you know your only limited by battery capacity its all about trying to keep the CA watts as low as possible to get the job done.

I reckon just get a normal DH hub with decent flange and 'tube' diameter and use butted spokes, or stick another wheel on it and your away.

View attachment 1

the system I was talking about, was my trike that weighs 100 pounds and I weighed 200 pounds at the time. That trike ran 0 to 70 miles per hour in 4.3 seconds. & I never had a problem with the hub. I completely understand that torque is the issue not watts.

 

[Ecobogan]

The final ratio for 1st is 38.6:1. That being 3.041:1 on the primary drive, 2.846:1 as 1st and 4.461:1 on the final drive (13/58).There is still a vague ? as to whether this is a 6t or a 4t, whatever it is it's wired in delta. A few wheels options have popped up on the dark web and should have something more macho soon.

 

[ksithumper]

An Astro 3220 provides 10 in-oz/amp of torque according to the spec sheet.

At 100 amps peak and with that gearing, that is over 270Nm (200lbft) at the rear wheel.

Add a shock loading from jumps and I'm surprised the hub, not to mention the spokes, lasted as long as they did.

A pit bike is similar in power - look at the size of the spokes and hubs on those.

 

[panurge]

Isn't that label clearly showing a 3220 4t on the MODEL line?

A 4t in wye has 169Kv => *1/9.5478 = 0,05649Nm/A => @100A=5,649Nm => *38,6= 218Nm (continuous)
A 4t in delta has 169√3=292.7Kv => *1/9.5478=0,03262 Nm/A => @100A=3,262Nm => *38,6= 126Nm (continuous)

About speed:

A 4t in wye @50v => 50v*169rpm*60 *2,113m/38.6= 27.7Kph
A 4t in Δ: 50v*292.7rpm*60*2,113m/38.6= 48Kph

Seems to me that you've reported max speed in first gear as about 30....that's why I was thinking you still run it in Wye.

You could Put a simple good speedo on the rear wheel and work back from no load speed@given voltage to get the turn count of your motor, and anyway the no load Kv at the wheel in the real world, that's what you need, in fact.....

My feeling with my bike (3220 4t, 19:1, 55kph, 107Nm cont.) is that only few times I have wished higher torque than that.....say 22:1, but more likely I would like an overdrive to get a 70Kph top speed....for a 3 speed 40/70/100 should be an amazing top speed value for each gear...

Anyway That's an epic thread......