The guts of the SRAM dual-drive exposed in pics :)

[liveforphysics]

These are for AussieJester KiM, and also for the rest of the folks interested in the dual-drive hubs of course.

I wanted to see what they looked like inside to see if I could get a rough idea of the torque handling ability of the hub.

Pics first.

 

[deecanio]

go Luke,
been waiting for this and your advice before i purchase!!!
looks better than the sa mate no?????
i only want to put 4kw through it but wonder if it will stand up to some jumping more than torque??
let us know what you think mate - if you give me the nod i will buy one

D

 

[Miles]

This is the exploded view from the manual:

 

[liveforphysics]

The gear on the axle is the smallest, as well as the highest torque loaded gear in the hub. This makes for the highest tooth loading, and the likely failure place. With trannys, you only need to analize the possible weak links rather than the whole system to find the practical limits.

We have an addendium diameter of 5/8", and a diametral diameter of ~1/2" (This is the one that matters to us).

The pitch is 36.
The width is 1/4", loaded diameter of 1/2"
Triple planet driven arangement.

This tells us, if they went with 70,000psi tensile strength steel, we've got a safe shared tooth loading of ~149lbs
On the bright side, the shaft did feel hard as hell, and showed blueing from induction hardening, so I think it's well above a mild steel 70ksi, likely around 125-150ksi would be more reasonable, which would put us in the ~270-320lbs tooth loading range (keeping the safety factor of 3).

 

[Miles]

SRAM service manual 2010: http://www.sram.com/_media/pdf/sram/dealers/SRAM-TechManual-MY10-GearHubSystems-English-RevA.pdf (22.5 MB)

Dual-Drive operating instructions leaflet: http://www.sram.com/_media/techdocs/DualDrive_Ins_E_12_02.PDF

 

[liveforphysics]

go Luke,
been waiting for this and your advice before i purchase!!!
looks better than the sa mate no?????
i only want to put 4kw through it but wonder if it will stand up to some jumping more than torque??
let us know what you think mate - if you give me the nod i will buy one

D

The strength of the hub for riding abuse seemed very strong. Large diameter cone bearing on the drive end, and a normal diameter cone bearing on the non-drive side.

As far as torque loading goes, 4kw at 1rpm is like at least 10,000ft-lbs (just guessing). In other words, it's going to depend on how things are geared, and the weight of the vehicle to know if it's going to break. They seem to hold up well for all the torque human power can deliver.

 

[deecanio]

well from Gary's threads i know that his setup is nearly identical to mine, and he is a big guy, but runs the sram with no issues thus far, the main difference is he's running 20's and im on 25's ish.
Kim rightly warned me to hold off for a bit given Matt P's sa failure.
i'd counted on being able to use the sram to give me three motor gears and three pedal gears without the need for a delta/wye switch immediately.
I can work around not using the sram by going back to 3 rings upfront and a delta/wye on the motor but it is a lot more hassle.
Given that i will hopefully use etards monster 219 sprocket to avoid using any reduction at all will this increase the stress on the sram??


D

 

[liveforphysics]

Ultimately, the hub only needs to be strong enough to wheelie the vehicle inside the range of it's safety torque loading figures, and use the 3X overhead to handle the shock loading and werid dynamic loading situations that gears encounter.

I think these will be strong enough to wheelie most any non-recumbent bike, so it should be good to go.

There is another failure mode possibility though... This is gear mesh seperation. When gears get loaded to the point the carriers flex out away from each other, the gears get loaded at the tip rather than the pitch circle area, which makes them much weaker. This happens to some models of Honda tranny, and it doesn't matter how strong of a gear set you put in them, until you build a brace to reinforce the case bearing locations, the tranny will fail (were're talking 1,000hp loads here). Fortunately, the planets appear to be in a robust carrier, but they are only bushed rather than on bearings, so if the bush wears, the ring gear should help to control planet mesh to a degree, but I do see it as a potential failure for high torque high RPM loading for long periods of time. Hitting them with a little grease once in a while might be all that is needed to avoid this failure though. The hub I took apart had a few years of frequent use before I bought it, and the only place I found any wear at was on the corner of the dog hub teeth with some mild rounding on the floating hub only, which indicates only one side gets hardened. It wasn't a level of wear to cause it to slip out of gear or mis-shift or anything though. Unlike crappy synchro-crappers, dogs can endure a ton of punishment and still perform perfectly.

 

[deecanio]

excellent reporting!!
ok Luke straight out - do you think i'll be ok with an sram in my setup? im not going for crazy power but i do ride a little rough?
one other thing to mention is the ability to ride electric well.
i consider myself e-trained now meaning i know when to ask for power and when to crank to give my setup the best chance of long life - using the sram will be like riding a puma imo, if you ride boggy mud uphill sitting on your ass and max throttle your likely to peanut butter your gears, help yourself and you'll be fine.
in your opinion should i buy one?

D

 

[liveforphysics]

I would go for it D I'm personally going to be hitting it with 2-3 times the power you're going to be throwing at it, and I'm thinking it will be OK if I try not to shock-load it.

A bicycle can only apply as much torque as it takes to wheelie (well, a little more due to the delay as your body accelerates upward and you start to flip over).

I think for a mountain bike style geometry bike, this will be a solid safety net for protecting this hub.

We've also got some real-world high power testing with Gary's 3220 bike. Though on a smaller rim, he is a fat guy, and he was powershifting it, which means it's taking fat-guy +3220 shock loading

As far as longevity goes, the biggest thing for reducing gear-set fatigue failure is the undercut, which I completely forgot to look at.

I will take it apart again tomorrow night at work and bring my DSLR with the macro lens and get some shots so we can see how the undercut (if any) has been done.

It requires an additional step to put an undercut fillet on a gear, and the skip it sometimes with little gears. This means they inherently have focused stress-riser points at the base of each tooth, which fatigue and eventually fail overtime (when loaded near capacity at least). If they do have a properly undercut fillet, then they will handle abuse for a long time because they spread the stress and fatigue over a wide area.

BTW- This is what I'm talking about for anyone who might be confused:

 

[deecanio]

hot damn, wooohooo
thats good enough for me to give it a whirl, many many thanks Luke, i'll be ordering tomorrow so i'll take a look back here beforehand.
All i need now is etard to come good and im ready to build.
Parts for sale!!!! Parts for sale!!!

p.s. dont forget to let me know what i owe for return shipping i'll pm my address now


D

 

[liveforphysics]

hot damn, wooohooo
thats good enough for me to give it a whirl, many many thanks Luke, i'll be ordering tomorrow so i'll take a look back here beforehand.
All i need now is etard to come good and im ready to build.
Parts for sale!!!! Parts for sale!!!

p.s. dont forget to let me know what i owe for return shipping i'll pm my address now


D

Ok. Once I finish broaching them, you should PM me like twice a day everyday with a reminder to mail them, because I think I might be actually retarded when it comes to remembering normal-life stuff, including mailing. lol

Here is one of the pictures above highlighting the only wear areas, and some tips on how to keep the dogs in good shape.

 

[spinningmagnets]

lfp, you posted to shift hard and fast for long dog-life (to ensure full engagement), should power be let up for a half second when shifting? (thanks for this thread and pics!)

 

[liveforphysics]

lfp, you posted to shift hard and fast for long dog-life (to ensure full engagement), should power be let up for a half second when shifting? (thanks for this thread and pics!)

Yep, for this style of tranny, minimize torque being carried by the tranny, then shift as fast as possible. This puts the absolute minimum wear on the critical pieces inside the tranny. The most damage you could do would be to half-push it and hold it there so it's starting to engage, and then popping back out. This is how newbs wreck superbike trannys, by being scared of downshifting, so they slowly/softly push the lever down...mashing and mushrooming the edges of the dogs until they mis-shift, and then can't go into gear, or stay in gear. A superbike tranny that is constantly ridden in anger tends to always look brand-new inside, and stays working like it did when new. This is pretty much the opposite of a car tranny with syncrapos wants to be treated for maximum life. Synchros = ghey.

 

[Thud]

Just a couple off points:
Garys video is using a Sturmey-Archer SRF3
http://endless-sphere.com/forums/viewtopic.php?f=28&t=12267

[youtube]IWVJUtFP-bs[/youtube]

I doubt your going to find undercuting on these gears. That normaly occurs when fitting a small gear into a much larger gear. The solution is to increase the pitch (make the teeth smaller) or adjust the presure angle/nose radius on the smaller gear.(to not under cut the larger) It appears these are very close to optimal for the intended applicaton. Again, I think the biggest issue will be the "shock"loads that these will be subjected to.

Spin, don't confuse "Hard" with "quick deliberate exicution" of a shift. A properly adjusted linkage will be critical to survival in these applications. & yes you can do it under full power (the old Macio's were renownd for their full power upshift-ability)
but, unless you racing ...why abuse the eqiupment?
There were/are(?) aftermarket Air-shifter kits that incorperated an "ignition interupt" to insure dog engagement without slipping of the clutch.

 

[deardancer3]

Thanks Luke.

I have 3 sram DD on the shelf, waiting to go on 360w cyclone chain drives.

One A recumbent with 20" wheels, one a velomobile 24" wheel, one to be built for wife for a manual bike.

I feel good proceeding in the spring.

Thanks again

D

 

[liveforphysics]

Its apart again. Its got a fully radius'd fillet, and no undercut.

This is better than it could be, but it won't be quite as immune to cumulative fatigue as an undercut gear. But tt should have a higher tooth sheer strength than undercut, which is good for this application I supose.

KiM- throw up the pics with the stripped SA hub so we can see what its shaft gear looks like.

 

[GGoodrum]

Though on a smaller rim, he is a fat guy, and he was powershifting it, which means it's taking fat-guy +3220 shock loading

Hey, I prefer big-boned...

 

[deecanio]

Hey Gary,

nice to see you, any input on the sram please?

D

 

[GGoodrum]

Hey Gary,

nice to see you, any input on the sram please?

D

I think I'd go with Luke's advice, and give the DD unit a try. It looks pretty robust to me, but I don't have anything to compare it to. Maybe if someone has looked at how it compares to the regular SA, SRAM and/or Nexus 3-speed hubs, in terms of how the internals compare, we might have a better idea. My SA3 still works, but I've only ridden it a few times, since that video was posted.

Matt P. has managed to trash his hub, so i guess we now know these won't last for a long time at high power and/or higher levels of abuse, but the DD might be a different animal. Luke will eventually break his, no doubt, but he should be able to document/predict how these might do with more normal levels of extreme abuse. :lol:

-- Gary

 

[swbluto]

As far as torque loading goes, 4kw at 1rpm is like at least 10,000ft-lbs (just guessing).

or :lol: ? Can't tell.

He probably means an input power of 4 kW rather than an output power of 4 kw. I would probably expect a 20 N.M. motor torque with a gearing ratio of 20:1 to have a geared hub torque of 400 N.M. = 295.02486 foot pounds. Do you think this would be close to causing this dual-drive to fail?

 

[liveforphysics]

As far as torque loading goes, 4kw at 1rpm is like at least 10,000ft-lbs (just guessing).

or :lol: ? Can't tell.

He probably means an input power of 4 kW rather than an output power of 4 kw. I would probably expect a 20 N.M. motor torque with a gearing ratio of 20:1 to have a geared hub torque of 400 N.M. = 295.02486 foot pounds. Do you think this would be close to causing this dual-drive to fail?

I worked the numbers just for fun 21,120ft-lbs of torque @ 1rpm = 4kw.

300ft-lbs would be beyond the 1/3rd safety factor, but should be inside the the calculated tooth sheer point. Often times if you go beyond the safety factor with a tranny, strange harmonics and shock loading can cause instantaneous 2x-10x instantaneous loading, which rapidly fatigues a gear until it develops stress cracks and sheers a tooth off.

For example, when my civic's on the jug, I generate about ~350ft-lbs of torque at the crankshaft (at ~8,400rpm), and yet I manage to shatter gears on the tranny input shaft that can easily hold 2,000ft-lbs of static loading. My engine can only generate a constant 350ft-lbs, but the shock loading when I launch with tire-shake is able to put instantaneous shock loads into the tranny at 5x-10x what my average torque value would be.

An electric motor is different of course, and a bicycle frame is different of course as well. I don't know if it makes shock loading peaks higher or lower, but I'm thinking perhaps on the lower side.

 

[AussieJester]

The gear on the axle is the smallest, as well as the highest torque loaded gear in the hub. This makes for the highest tooth loading, and the likely failure place.

Just so happens to be the gear that failed in the SA of Matts too buddy :-(

KiM- throw up the pics with the stripped SA hub so we can see what its shaft gear looks like.

Cheers for taking the time to pull it down, my moneys on you blowing one first with your mammoth power
young fella, if you don't your not trying hard enough

KiM

 

[enoob]

me no likey the looks of that.

even if your able to show some restraint on launches if your under hard acceleration and the tire slips and then hooks up again its gonna be touchy.

 

[AussieJester]

Yeah looks pretty much the same size and tooth count to the SA to me what is it Luke around 20 teeth on that gear axle?

I think they will hold up long enough for users to get a false sense of security then at the worst possible
time they will find out they aren't up to the job LoL...

Best of luck fellas.... My coins on Luke destroying the first one...


KiM