Drive Train Shock Loading and Weakest Link Over ride

[DingusMcGee]

The worst Drive Train Shock Loading arises in the drive train when a free spinning full powered rear wheel drops to the ground in the lowest gear and there is a high coefficient of friction[no snow or mud] with great resistance to continue in the direction of wheel travel. This situation arises frequently on difficult dry rock off road biking when the hill is steep, the path is rocky enough and steep enough that the rider has to move fore and aft above the seat to keep the front down after hitting a rock and he has to slide back to get the bouncing rear wheel on to the ground asap as traction is needed to move the mass uphill. Conditions of the like typically put the highest loads on the drivetrain. Higher loads than wheelies unless maybe you have a clutch?

A long travel front suspension fork will reduce the bounce of hitting a rock and a light weight rear wheel with quick response suspension will get the rear wheel back in contact with the ground sooner than a hardtail and thus lessen the wheel speedup while off the ground.

But alas these necessary design changes to make a good hill climbing bike are beginning to make the bike look like a motor powered dirt bike -- why bother with the petals and crank? Surely to each his own as to how he proceeds in designing this ultimate hill climbing bike/machine.

Rather than continually destroying various part from overloading we can add a safety link or weakest link. The easiest safety link to impliment is setting the max amps with a fuse or programmable motor controller. But if we put an additional safety link in the drive train we specifically protect that system while having higher amps. The time immemorial method of weakest link design for over torque is the use of the shear pins and this method could be easily implemented in off road ebikes. Simply remove the allen screw on one of the motor pulleys, drill through the shaft and then insert a shear pin of that diameter. Alum shear pins are perhaps the weakest and chrome moly steel shear pins may be the strongest. You can also drill out a shear pin to make it weaker. Drilling of the shear pins make them adjustable to our particular safe loading. You would also have to remove the key stock from keyway to get the shear pin to load.

I have sheared the pawls on the Shimano cluster carrier with the Cyclone 3000 and torn a gates belts on the LR Sm Bk while operating these at 52v. The cluster carrier and the gates belt may be thought of as the weakest link in the respective drive trains. In either case the parts run about $25 when getting them at a reasonable price off the internet rather than buying them from the inflated price LR parts store or some local no deal bike shop. With a weakest link design we end up not twisting the motor frame or bike frame while not severely loading other parts also.

Another method of saving the drive train would be to put a over torque release device in the drive train. The big Milwaukee right angle drills and RPD have such an over torque device on them to keep the drill from spinning the operator around when the drilling bit hits something solid. Such over torque devices would when set properly save all the drive train part in severe drivetrain shock loading.

A big portion of any ebike modification has both eloquence and kludging together. My criteria: kludging is allowed except the kludging of motorcycle parts -- hence the Milwaukee over torque release is allowed but not motorcycle rims, big chains and thick chainrings etc. I want my ebike to be more bicycle like and will live with that.

 

[FastJohnny]

Great point. I wish it was as easy as adding a "missing link" for the bike chain, but one made with a shear pin in it. There could be a market for this.

 

[DingusMcGee]

fastJohnny,

I hadn't thought of this one but it would be quite easy to replace the master link unless the loose traveling chain did some damage. The situation of using a destroyable safety link is that when the safety over ride happens you are down and will have to spend time replacing the link.

About all motorcycles have clutches except for the group of Föppel see: https://endless-sphere.com/forums/viewtopic.php?f=28&t=78861 where they are in a continual armsrace of beefing up the next likely failing/undersized part.

For an ebike a slip clutch would need no release lever like a true motor cycle clutch to act as a real time safety clutch and there would be no down time to replace a discrete link like the gates 25mm wide belt in the LRSB that will easily fail in Drive Train Shock Loading -- DTSL.

 

[DingusMcGee]

Drive Train Shock Loading -- DTSL -- occurs additionally but maybe less severe when we get air time and land at WOT. It also can occur when coming out of mud hole with WOT.

The Mechanical Solution to DTSL is either safety links or clutches.

There are Electronic Solutions to DTSL: Here is a solution using an Arduino Uno Board [C programmable micro processor] with 2 sensors and one actuator. We assume the overload/destructive condition arises when the mid direct drive motor inertia (angular momentum) is reduced so rapidly from the very rapid rear wheel speed retardation of gaining contact with the ground that the forces and torques on the drive train exceed their elastic recovery design limit. Rather than beef up these drive train parts, we say let's detect the onset conditions that lead to such high force and torques in real time but reinstate driver/rider intentions asap [maybe WOT]. This electronic solution under DTSL conditions would sever throttle input until conditions less than DTSL were emerging.

The conditions that can lead to DTSL are very high rear wheel accelerations[SENSOR 1] with almost no current draw[SENSOR 2]. The necessary condition for DTSL is a very fast spinning rear wheel that got in this condition very quickly from having WOT while the wheel is in free air. So the action of the electronic over ride [ACTUATOR] is at the onset of very high acceleration while at very low current sever throttle input.

The condition for reinstating throttle input is that as soon as the free spinning wheel's velocity[RPM which is continually monitored] lapses faster than its free air lapse rate engage the severed throttle input since we assume the wheel has hit the dry ground. To keep the duty cycle time of WOT high the velocity sensor will need more than one magnetic pick up per wheel revolution. It is possible to symmetrically mount 12 magnets on the rear brake disc. This would allow for as little as 30+ degrees error in wheel rotation for the quickest reengagement of WOT.

We would numerically differentiate wheel velocity data to get wheel acceleration. There are 2 pins / hardware on Arduino for acquiring real time velocity data when the sensors are discrete so as you do not have to continually read the input pin. All readings from the shunt on current demand are in real time. A signal type SSR would certainly be a fast enough relay to sever and reinstate the throttle input to keep the duty cycle of WOT high.

This system could be tuned to reduce DTSL on mud, changing snow conditions, and air time landings also. Think of it as an anti-spin device that reduces DTSL loads to the drive train by the max torque/load that the motor power can put on the drive train. This setup does not null the drive train inertia but will not allow the motor to add the additional force s/torques it could add by WOT motor power to the inertial forces.

 

[spinningmagnets]

Recumpense added a friction clutch to his drives as an option. When users would pop wheelies, do sudden burnouts, or program their drive for max sudden acceleration...it previously caused issues. I'm not saying a friction clutch is the best option for what you are describing, but I'm just saying it might be worthy of consideration...

 

[DingusMcGee]

spinningmagnets,

but I'm just saying it might be worthy of consideration...

Thanks for the lead to Recumpence and their clutch.

Absolutely the clutch is worth looking at & maybe installing. Ideally a [slip]clutch could take all the torque/force the mid drive motor puts on the drive train in simple WOT pulling -- and maybe a little more. But above this threshold slipping would be necessary so as to eliminate inertial loading that arises upon high speed wheel contact at WOT -- the conditions of DTSL. And the nice trait of a clutch is that they can take repeated overloadings -- many more than a non weak link drivetrain can take -- without failure.

What is worthy of consideration? Low tech? Torque is lowest at the motor pulley so the smallest shear pin location would be here. Would a snug fitting birch wooden dowel pushed through the cog pulley then through the motor drive shaft and ending out of the cog withstand motor torque at that location? Certainly water poured on the wooden dowel would very quickly tighten up the pulley and the dowel.

I can get a number on the tangential chain force at the motor pulley from knowing my c-3000 bike with its gearing can climb a little more than a 50% grade and so with getting numbers real design begins.

 

[DingusMcGee]

The idea of reducing drive train shock loading is resolved for me. My solution doesn't involve the use of clutches, Arduino modified CA_V3 response circuit or heavier components used. Here is my line of thought from experience:

I have 3 ebikes with mid drive motors and have used them on severe hill climbing attempts. The LR SB failed on the second attempt of a quite difficult hill when the Gates belt tore [bike: specialized fat bike hardtail]. The 1 st c-3000 build had several chain breaks and some pawl failures[bike: Specialized hardtail Pro racing bike]. My Specialized FSR has had none of the type of failures I would attribute to drive train shock loading while riding the same hills that produced the failures in the 2 hardtail bikes. I suspect the rear suspension lessens shock loadings.

My solution is to choose a bike that has adjustable rear suspension stiffness in the range that can alleviate the inertial effects that I mention in the above posts. It seem the little bit of chain play happening during suspension pull up while increasing torque can reduce harmful peak loadings enough to end them ?. You may need to adjust your rear spring stiffness for more than your body weight + motor & battery weight to get this benefit because if you bottom out too early you can introduce a big portion of the shock loading when the load hits the shock's/springs end of travel.

So the "suck up" from rear suspension is useful in lessening shock loading effects.

 

[2old]

SRAM has a new 8 speed ebike specific group that may be of interest to you. Supposedly, the chain has been strengthened. I've had good luck with BBS02, 8 speed cassette with 11-17-28 (the 17 is aligned with the front), 30 tooth chainring for ascents over some fairly gnarly terrain, but less tortuous than you describe.

 

[recumpence]

This is one major problem with bottom bracket drives. Driving heavy loads through the pedal chain on the right side is very problematic. This is why I choose to drive my bikes through the left side of the rear wheel with a large sprocket at the rear wheel and relatively High chain speed to that sprocket. This drastically reduces the load on the chain and other Driveline components and shifts that load purely to the rear hub. I have no problems pulling as much as forty thousand Watts through my bike's this way. I am not exaggerating by that, I pull 40,000 Watts through a bicycle chain and a standard rear wheel on one of my bicycles.

That bicycle is pictured in my avatar. It has three Astro flight Motors @ 14000 watts each.

 

[DingusMcGee]

Recumpence,

This is one major problem with bottom bracket drives. Driving heavy loads through the pedal chain on the right side is very problematic.

Yes, I suspect the torque you get from 40,000 watts would likely be a problem with the pawls of the cluster carrier vs. a left side drive with no freewheel at the rear wheel. Bike chains are good for some 2200 lb f and the Connex barely stronger than the Shimano but at 3x the ebay cost.

http://www.cantitoeroad.com/assets/images/products/docs/connex_by_wippermann/Connex_Breaking_Load_Test_Results.pdf

But I obviously ride in a tighter circle than you do. I see your enduro-like builds and those of others have a wheelbases >50 inches while my hardtail is 41 inches and the Specialized FS is 43" with them both weighing in at < 50 lbs making carrying it over downed logs easy. These 2 bikes have superb nimbleness but I would have zero need for 40,000 watts. In fact the C-3000 watt capability is more than I could possibly want navigating through a boulder field and for technical rides I do.

As I have mentioned in some of my post above there are several ways to lessen drive train shock loading. You have chosen to implement additional drive trains which add more weight and complexity while the Foppel group of Germany has chosen to beef up components of the right side drive train. My choice: Carry a Shimano free hub body on long rides. This is added to my bike weight as it and chains are cable tied under the seat.

My personal ebike "philosophy" is that high angle fork-tight geometry and big wheels work best for my type of country rides -- which is just the likes that some of the mountain bikes are designed to do with only a person. So it has been a discovery for me that by tuning the rear shock I can accommodate some inertial loadings and enhance the life of the ordinary bike components for the way I use them.

My Goal: If you would pull off the motor and battery (connections and all) would the bike look like a bike that had never had a motor modification?