D
duivendyk
Guest
Accasionally problems seem to occur with this drive system,especially after it has been in operation for some time,consequently I deemed it worthwhile to subject it to some scrutiny.On of the problems that seems to crop up is that the belt starts skipping teeth on the drive sprocket under certain conditions,Which is obviously not good for the belt or the drive sprocket.Since I don't own such a GeBee system and am not likely to so in the future,some of my dimensions may not be quite accurate but represent reasonable assumptions.The effect on the general conclusions will be slight however in my opinion.
In this setup the drive belt runs at surprising speed,at 5500 rpm with a 1.5" drive sprocket,the belt velocity works out to be 11m/sec or about 35 ft/sec (24 mph).When the belt goes around the sprocket in the process of reversing direction, it is subject to considerable accelleration which tends to lift it off the sprocket,the accelleration computes to over 60 g's (g =9.81 m/sec^ (^means squared).The belt tension,produced by the chain tensioner on the slack side of the belt is intended to keep the belt on the sprocket,on the driven side of the belt there will be tension on it from the reaction of the driven ring on the rear wheel.What the belt 'likes' to do so to speak is to keep on going straight up unless pulled rapidly towards the front by the preceeding portion of the belt to keep it wrapped around the sprocket,so that the sprocket teeth engage the belt for transmitting the engine torque
Obviously if there is no slack at all in the belt ,or if the belt tensioner excerts enough force to maimtain adequate tension on the belt to keep it on the sprocket every thing will be fine.The amount of tension reqd. can be readily calculated if the weight/length ratio is known (measure weight &length of belt),but I don't have this info available.
Since the driven ring is never perfectly concentric,the motor mount can shift, and the belt is liable to stretch ,there inevitably will be residual slack.The Gebee vendor leaves it for the belt tensioner to cope with this slop and also with any adjustments related to changes in drive sprockets, to the best of my knowledge.
It pays to examine the relationship between the reqd. sideways movement of the adjustment roller and a change in belt length,excenricity etc.For the geometry in question the following relationship can be demonstrated to hold aproximately L= S^/10. (^ means 'squared').L is 'apparent' change in belt length,S is lateral tensioner movement.For instance assume the excentricity of the ring is 1/16" (Total length change 1/8" or L= 0.125 " then the sideways movement to make up for it is 1.1" approximately,quite substantial.Assume now that the belt stretches 1/16", then the extra travel needed is another 0.25 " and for the next 1/16 " stretch it goes up only 0.20 ".So in order to minimise the back&forth movement of the belttensioner it would seem to make sense to allways have some belt slack since for the same change in belt 'slop' there is much less tensioner movement reqd.It is worth remembering that the rear wheel is running at 400rpm (30mph), about 7 rev/sec.
This seems like a good idea but there happens to be fly in the ointment.This has to do with decelleration and braking.When decellerating the bike drives the motor and the slack side where the tensioner is, becomes the driving part of the belt consequently the slack will immediately appear on the normally taut part.This is likely to result in tooth skipping and will not be good for belt&sprocket.The reason is that the motordrive does not behave like a freewheel because the cent. clutch will stay engaged untill the speed has come down enough for it to disconnect the engine.The MORE slack there is the worse this will be. The remedy for this is to have belt tensioners on BOTH legs of the belt.Preferably these would be linked with downward force applied to the combination to always mantain enough tension on both sides of the belt,while permitting sideways movement.In my opinion this would come long way towards minimising problems with the belt drive.I am not saying that it is a bad system but it could be made more trouble free.
In this setup the drive belt runs at surprising speed,at 5500 rpm with a 1.5" drive sprocket,the belt velocity works out to be 11m/sec or about 35 ft/sec (24 mph).When the belt goes around the sprocket in the process of reversing direction, it is subject to considerable accelleration which tends to lift it off the sprocket,the accelleration computes to over 60 g's (g =9.81 m/sec^ (^means squared).The belt tension,produced by the chain tensioner on the slack side of the belt is intended to keep the belt on the sprocket,on the driven side of the belt there will be tension on it from the reaction of the driven ring on the rear wheel.What the belt 'likes' to do so to speak is to keep on going straight up unless pulled rapidly towards the front by the preceeding portion of the belt to keep it wrapped around the sprocket,so that the sprocket teeth engage the belt for transmitting the engine torque
Obviously if there is no slack at all in the belt ,or if the belt tensioner excerts enough force to maimtain adequate tension on the belt to keep it on the sprocket every thing will be fine.The amount of tension reqd. can be readily calculated if the weight/length ratio is known (measure weight &length of belt),but I don't have this info available.
Since the driven ring is never perfectly concentric,the motor mount can shift, and the belt is liable to stretch ,there inevitably will be residual slack.The Gebee vendor leaves it for the belt tensioner to cope with this slop and also with any adjustments related to changes in drive sprockets, to the best of my knowledge.
It pays to examine the relationship between the reqd. sideways movement of the adjustment roller and a change in belt length,excenricity etc.For the geometry in question the following relationship can be demonstrated to hold aproximately L= S^/10. (^ means 'squared').L is 'apparent' change in belt length,S is lateral tensioner movement.For instance assume the excentricity of the ring is 1/16" (Total length change 1/8" or L= 0.125 " then the sideways movement to make up for it is 1.1" approximately,quite substantial.Assume now that the belt stretches 1/16", then the extra travel needed is another 0.25 " and for the next 1/16 " stretch it goes up only 0.20 ".So in order to minimise the back&forth movement of the belttensioner it would seem to make sense to allways have some belt slack since for the same change in belt 'slop' there is much less tensioner movement reqd.It is worth remembering that the rear wheel is running at 400rpm (30mph), about 7 rev/sec.
This seems like a good idea but there happens to be fly in the ointment.This has to do with decelleration and braking.When decellerating the bike drives the motor and the slack side where the tensioner is, becomes the driving part of the belt consequently the slack will immediately appear on the normally taut part.This is likely to result in tooth skipping and will not be good for belt&sprocket.The reason is that the motordrive does not behave like a freewheel because the cent. clutch will stay engaged untill the speed has come down enough for it to disconnect the engine.The MORE slack there is the worse this will be. The remedy for this is to have belt tensioners on BOTH legs of the belt.Preferably these would be linked with downward force applied to the combination to always mantain enough tension on both sides of the belt,while permitting sideways movement.In my opinion this would come long way towards minimising problems with the belt drive.I am not saying that it is a bad system but it could be made more trouble free.
