Synchronous Belt Drives - any rules of thumb?

[John in CR]

Why do the manufacturers want everyone to have a degree in belts? Chains are pretty simple, it's easy to find a chart for rpm limits and power. I want to try a belt drive, and the problem I have seems to be pretty straight forward, but darned if I can find a good place to even start.. What's the recommended belt width and pitch if I want to be able to run at high as 20kw with a max motor rpm of 2500, about a 2:1 reduction, with an acceptable drive pulley diameter of 70-100mm and acceptable width up to 1". If 8mm Polychain GT is a good solution, then I can get the pulley stock from MDD0127 as long as I don't end up stuck with only some hard to find expensive belt to fit my 460mm centers.

I won't run near that power or rpm at first, but I don't want to have to change anything if I'm able to get there.

If you guys can just point me in the right direction, I'll run with it.

John

[mdd0127]

The closest you can get to a 2:1 ratio using a 70-100 mm driver pulley with the stuff I have will be using a 38-75 combination.

The 38t has a 96mm pitch diameter and the 75 tooth has a 191mm pitch diameter.

This will give you a 1.974:1 ratio, and a center distance of 19.36" (491.7mm) using an 8mgt-1440 180 tooth belt. With a 12mm wide belt, at 2500rpm on the drive pulley, this setup will handle well over 30hp. You can make up the excess center distance with an idler/tensioner, or adjust your motor location to suit.

20kw is 26.82 horsepower so this setup would be just about perfect other than the slightly longer center to center distance.

If you ever think you might go over 30 hp and you have room for 1" wide pulleys, you can just make them 1" wide and run the 12mm belt width for now and get great efficiency. When you're ready for crazy power, you can pick up a 21mm wide belt and it will handle 50hp (37kw).


One cool thing I noticed when working with the 8mm pitch stuff is that the pitch diameters are one tenth of the tooth count. I.E. 22t= 2.20" pitch diameter, 38t=3.8 and so on. Knowing that can really speed things up when your at the initial stages trying to figure stuff out.


Everything is covered in the polychain design manual but there are so many numbers it can definitely make your head swim. If I haven't sent it to you, I gladly will. If you have any other questions, feel free to ask. The system on my dream bike uses five belts and one of them is a dynamically registered drive so going through that design process forced me to get pretty familiar with this stuff!
:smiley w/smoke coming out of ears:

[John in CR]

Wow, a half inch wide belt can handle that kind of power??? I think I'll go the next size up just so it's doesn't look so small, probably the same reason you got the 23mm wide belt. I would like to try finding a belt closer to the right length. I have some adjustability and tensioning potential with the rear wheel, but not that much. If I have to use an idler, so be it.

Dream bike...I don't know. 5 belts sounds like a nightmare instead.

Now I gotta decide how much of that pulley stock I want. I'm sure some of my customers will want pulleys too while I'm making mine, though shipping down here and then back stateside gets costly.

John

[mdd0127]

Wow, a half inch wide belt can handle that kind of power??? I think I'll go the next size up just so it's doesn't look so small, probably the same reason you got the 23mm wide belt. I would like to try finding a belt closer to the right length. I have some adjustability and tensioning potential with the rear wheel, but not that much. If I have to use an idler, so be it.

Dream bike...I don't know. 5 belts sounds like a nightmare instead.

Now I gotta decide how much of that pulley stock I want. I'm sure some of my customers will want pulleys too while I'm making mine, though shipping down here and then back stateside gets costly.

John

I only got the 23mm wide belt because I did a timing belt replacement on a Subaru and the belt looked good so I kept it. I haven't figured out a use for it yet.

I'll be running all 12mm wide belts on the prototype bike and even those will be total overkill. I want to make a belt converted hammerschmidt jive with an in frame nuvinci and the swing arm pivot is a jackshaft too so it's going to need some belts. The cool part is, they're all enclosed and should last at least as long as a timing belt setup on a car engine does so the only things that will ever need replaced with on the bike will be tires, brake pads, hand grips, and batteries.

A 22/45 combo would get you a 465mm center to center distance/2:1 ratio, and a 26/63 setup would be 459.5mm center distance/2.4:1 ratio, but it wouldn't handle as much power and it sounds like the 22 and 26 tooth would be too small.

[Chalo]

Wow, a half inch wide belt can handle that kind of power??? I think I'll go the next size up just so it's doesn't look so small, probably the same reason you got the 23mm wide belt. I would like to try finding a belt closer to the right length. I have some adjustability and tensioning potential with the rear wheel, but not that much. If I have to use an idler, so be it.

Careful using Gates's recommended limits on carbon belts for bike chains. Those things are expensive and do not last very long. I'd rather overspec a belt than get stuck replacing it often at high cost.

If your gizmo is going to pound out torque like a Harley, you could just use the belt and/or sprockets from a Harley. They won't be the most efficient, but they'll be reliable and not outlandishly expensive.

Chalo

[mdd0127]

Wow, a half inch wide belt can handle that kind of power??? I think I'll go the next size up just so it's doesn't look so small, probably the same reason you got the 23mm wide belt. I would like to try finding a belt closer to the right length. I have some adjustability and tensioning potential with the rear wheel, but not that much. If I have to use an idler, so be it.

Careful using Gates's recommended limits on carbon belts for bike chains. Those things are expensive and do not last very long. I'd rather overspec a belt than get stuck replacing it often at high cost.

If your gizmo is going to pound out torque like a Harley, you could just use the belt and/or sprockets from a Harley. They won't be the most efficient, but they'll be reliable and not outlandishly expensive.

Chalo

Harleys use 1.5-3.5 inch wide, 8mm pitch primary belts and 1.5 inch wide, 14mm pitch belts in the final drive that would be terribly inefficient for 30 peak horsepower. The belts themselves are also way more expensive than the belts for industrial applications because harley retailers buy the belts from the source and mark them up, adding an unnecessary middle man. The pulleys are extremely heavy and would have to be machined to fit anything other than harleys anyway, only adding to the cost.
http://www.ebay.com/itm/BELT-DRIVES-LTD ... vi-content

The polychain design manual is set up for industrial applications where drives run at their peak output for 16-20 hours per day. The stuff that is being used on bikes looks similar but has been weight weenied to death and I still haven't heard of any failures that weren't due to manufacturing defects or incorrect installation. In fact, all of my research has shown that people absolutely love their belt drives and my three friends that own bike shops have never seen a belt drive fail or wear out since they started coming out in 2008-09. There are even a few people here on the forum that have used 8m pcgt on high power builds and not only has no one reported a problem, no one has even reported any noticeable wear. A properly spec'd belt drive system that's kept clean will outlast the recommended lifespan of most frames. I spent about six straight months of my life researching the heck out of this before I decided to ditch my whole 219 chain driven design and invest $3k and four more months of design time to convert to belt drive.

[Chalo]

(...)1.5 inch wide, 14mm pitch belts in the final drive that would be terribly inefficient for 30 peak horsepower.

An 883 Sportster doesn't have much more power than that, and likely less wheel torque than John's bike. It uses a 1-1/8" belt. I can't speak for its efficiency one way or the other.

http://www.motorcycle-superstore.com/3/14/140/36608/26292/PITEM/Goodyear-1-1-8-Falcon-SPC-Rear-Drive-Belt-2011-Harley-Davidson-Sportster---XL-883N-Iron-883-Parts.aspx

John is already talking about making custom pulleys from blanks; I don't see how modifying an existing pulley would take more work or expense than that.

I can't even imagine how much you hate chains that you would go to such lengths to use a belt instead. As for me, I don't like that chains are greasy and occasionally noisy, but I like that they are versatile, strong, efficient, inexpensive, and modular.

Chalo

[comradegerry]

Whilst looking for a planetary gearbox for a project I found these guys which I had not seen mentioned before....

They make a heap of different sizes and ratios but the 2 smallest kind of line with ebike application.

What I really liked was on the AL series on which they have integrated a belt drive which could be useful

The one I emailed them for info and pricing was AL070, 20-1 reduction, >94% efficient, 1.2kg, <56db at 3000rpm, max input of 10,000 rpm (4000 continuous), 70mm wide. The next model up would be small motor bike size and a bit over 2kg.

Might be a useful way of getting the motor rpm down to what is useful at the rear wheel by belt, just a thought.

I tried uploading the photos and spec but I am having issues, their website has all the info though.

http://apexdynamicsusa.stage.thomasnet- ... -gearboxes?

[mdd0127]

I got interested in belts for two reasons. Low noise, and reliability. I've worked in industrial plants and on motorcycles for years and replaced tons(literally) of chains. I've never had to replace a toothed belt other than for preventative maintenance reasons on cars with well over 100k miles. I've seen a few final drive pulleys on big bikes wear out but unless something nasty gets sucked up by the belt, the belts are bombproof. On the industrial side of things, the chain drives needed constant maintenance and attention but the machines that ran synchronous belts ran constantly and reliably. If I'm ever able to put my bike into production, I want the people that buy them and ride them to be able to ride instead of fix and to be able to hear what's going on around them, whether it's wildlife, traffic, or just enjoying the silence.

[Punx0r]

This may be a daft question, but are belts really HP/KW rated rather than torque rated (presumably with an upper speed limit)?

[mdd0127]

This may be a daft question, but are belts really HP/KW rated rather than torque rated (presumably with an upper speed limit)?

They are hp rated. The hp rating increases with pulley size and the speed of the fastest shaft.

[Miles]

This may be a daft question, but are belts really HP/KW rated rather than torque rated (presumably with an upper speed limit)?

They always give a driver pulley speed and pulley size/tooth engagement at which the power rating applies. Tables are usually for a given width of belt.

[Miles]

I've always assumed that the specification isn't just directly proportional to torque because of the service factor.

[Punx0r]

Ah, right. That makes sense, else you could transmit infinite torque at an infinitely slow belt speed

So presumably there is a max. belt tensile load spec, and this is expressed as HP for different pulley sizes for the convenience of the user?

Just trying to understand the data tables because I got confused a little while ago when trying to find the minimum belt width (of a given pitch) needed for a drive.

[Miles]

So presumably there is a max. belt tensile load spec, and this is expressed as HP for different pulley sizes for the convenience of the user?

The limit is usually set by the number of teeth engaged in the driver pulley (ratcheting). Only on the larger driver pulley sizes can you get near the span tension limit. The torque capacity rating falls as speed increases because speed affects the service life and hence the service factor (my presumption, there may be other factors).

[bigmoose]

The one I emailed them for info and pricing was AL070, 20-1 reduction, >94% efficient, 1.2kg, <56db at 3000rpm, max input of 10,000 rpm (4000 continuous), 70mm wide. The next model up would be small motor bike size and a bit over 2kg.

Could you share their quote? Ballpark is OK. The right angle version would sure package well.

[Punx0r]

Thanks for the explanation, Miles

[John in CR]

I noticed MDD's pulley stock for sale, and that's what got me started in the direction of belts, especially since he had just the right size I need. No matter what pulleys I get, there will be machining. On the drive side, a chain sprocket would actually be a bit more difficult to implement on the cast steel bell side of the motor that has otherwise wasted width

I'm mostly interested in the silence, because these motors are significantly more quiet than typical hubbies and I'm sure sine wave controllers aren't far away for us regular guys. That makes near silence a reality, so I don't want to listen to chain noise. Plus I'll likely start selling some custom boy toy really high power ebikes locally in the near term, and little touches like belt drive and carbon fiber add a lot of cool factor that those with the bucks will love.

John