Gallery, Projects and General => How do I?? => Topic started by: smthrll on April 27, 2017, 10:30:28 PM

Title: Roller Project
Post by: smthrll on April 27, 2017, 10:30:28 PM
I was hoping to get some input from some folks on a project I'd like to tackle.  I'm still new to machining and I roughed out a sketch that I hope will help explain.  I'd like to create:
- although shown as 2 separate rollers, the inertia roller will be placed inside the outer roller
- the inertia roller will have bearings pressed into the ends (shown in blue), and it will spin on a fixed shaft.
- the stub on the inertia roller will protrude beyond the outer roller, so it can be driven by a belt

My question:
- can I press a bearing in the outer roller and have it ride on the inertia roller's stub
- I'm guessing there will be a problem with the bearing rubbing on the face of the inertia roller - I'm not sure of the proper way to create a gap between the bearing and the face of the internal roller.  The inertia roller will be running about 3x faster than the outer roller.   I thought maybe leaving a bit of a raised surface on the stub shaft to ride against the inner ring of the bearing. 

My apologies to all the engineers and machinists that are doing a face palm right now....

Title: Re: Roller Project
Post by: awemawson on April 28, 2017, 02:58:08 AM
It probably me being slow on the uptake, but I'm having a problem visulising what you are trying to achieve.

It seems to me you are trying to construct three elements, all mutually co-axial and separated by bearing races, is this the case?

Perhaps were you to reveal your application it would be clearer as I've no idea what an inertia roller is!
Title: Re: Roller Project
Post by: mattinker on April 28, 2017, 04:45:37 AM
Like Andrew I don't know what an inertia roller is!
To prevent the face of the outer bearing touching the "inertia roller, you can either space it away with a washer or leave a small step on the end of the said "inertia" roller.
Maybe I'm completely off the track!!

Regards, Matthew
Title: Re: Roller Project
Post by: smthrll on April 28, 2017, 10:04:03 AM
Thanks for the replies.  You are on the right track completely and have described the question better than I.

-project is for the back roller for a set of bicycle rollers.  It is a trainer that allows the rider to bike as if he was outdoors.  I've attached a picture of Trutrainer's product.  It has an internal flywheel in the last roller - this allows the rider to coast, as the inertia keeps driving the outer drum on the middle roller even when stopped pedaling.

So yes, all 3 elements are mutually coaxial and separated by bearing races.  Maybe a washer or a small step is all I need for separation?  I was worried about wearing out the bearing race of the outer roller rubbing on the end face of the inertia roller. 

I can understand the roller rotating on a fixed shaft, but I wasn't sure if there's problems with the outer roller rotating on another spinning shaft.  Tried reading the SKF bearing catalogue - bearings are actually pretty complex.  Once everything is assembled, there is really no way to provide lubrication. 


Title: Re: Roller Project
Post by: awemawson on April 28, 2017, 10:52:49 AM
Probably just a collar or thick washer is all you need.  :scratch:
Title: Re: Roller Project
Post by: JHovel on April 29, 2017, 08:00:11 AM
That looks clever. So the front and middle roller drive the internal (heavy) flywheel roller at something like 2 or 3 times the speed of the outer roller. So when pedalling stops, the middle and front rollers are driven by the internal flywheel roller for a little while by the drive belt, while the outer rear roller is only ever driven by the bike tyre.
Do I understand this correctly?
Title: Re: Roller Project
Post by: smthrll on April 29, 2017, 10:43:47 AM
That's it exactly.  Some trainers have an external flywheel that try to accomplish the same thing (kreitler rollers for example), but this is a nice compact way of doing it.   I've obtained some 2.5" steel tube,  I.d. being 1".  It weighs 17 lbs so I think it'll make a nice flywheel.   Writeups I've seen on the Trutrainer product say a 13lb flywheel spinning at 12000rpm if the rider can do about 50km/h.  They use a 4:1 reduction ratio  (3.5" diameter roller / 0.75" stub shaft)

I'll probably use the suggestions given here,  space the outer roller's bearing with a thick washer.   Maybe use a c-clip on the outside of the bearing to hold everything in the proper place on the stub shaft.

I'm certainly open to ideas though.

Title: Re: Roller Project
Post by: philf on April 29, 2017, 01:44:24 PM
Hi Rollie,

The TrueTrainer literature quotes:

"Additionally, viscous drag forces between the flywheel and roller simulate wind and rolling resistance."

The gap between the flywheel and roller may be critical.

I'd rather be riding on the road! 30 minutes on a turbo trainer and I'm drowning in a pool of sweat. :(


Title: Re: Roller Project
Post by: smthrll on April 29, 2017, 02:32:37 PM
I agree 100%, but it's almost May and I used my snow blower last week!   I just kind've figured that if I start this project now,  maybe I'll have it done  for when the snow comes in September again.  Maybe that drag will allow 200 watts resistance.

I also was thinking that if I can't get it quite right,  maybe I'll just pull the flywheel out and position it as a fourth roller,  kind've like the kreitlers except it would span the full width from rail to rail.   
Title: Re: Roller Project
Post by: smthrll on June 28, 2017, 06:49:55 PM
I've been making some progress on my project - usually by doing things at least 4 times til I almost get it right.  I have the inner flywheel with the stub shaft complete.  I ended up welding a steel plate onto the ends of my .75" walled cylinder.  I then  bored some bearing pockets and pressed them in.  I then turned the cylinder between centers (no stub shaft at this point), and took a finishing cut.  It seemed to work well, I can spin it up to 2000 rpm with no vibration and almost no run-out.   Not sure if it's proper form to use deadcenters on bearings, but it worked good.   

Trouble started when I attached the stub shaft (originally used 82 Degree countersunk machine screws).   Everything was running true, I cut the shoulder for the bearing to sit against, cut the micro-grooves for the belt,  the snap ring, and everything was good.  However, I didn't realize how important those countersunk pockets were.  These were cut on the drill press, and I don't think my accuracy was as good as it should've been.  When I removed the shaft from the cylinder, and then re-attached it again - good luck!  Out by .006"!   I closed up the countersunk holes, ruining my pretty bright finish, and used hexbolts.  Now I can adjust the bit of wiggle room and everything runs true again. 

I attached a picture of my outer aluminum drum as well.  I'm really struggling here - it's about 20 inches long, and my lathe only has about 16" of travel.  I had a really nice finish on the aluminum, but then had to swap the drum end for end in the steady rest.    The result is a fine line trying to match up the finishing cut on one half, with the finishing cut on the other.  I promptly ruined the finish with some fine sandpaper, trying to get rid of that line.  I'd ask for solutions, but I suspect the answer is going to be "buy a bigger lathe" if you want to do bigger work.

This forum is great,  thanks for all the ideas and problem solving tips I've found in other posts.

Title: Re: Roller Project
Post by: smthrll on September 09, 2017, 05:27:05 PM
Here's a few pics to show that I've followed thru with my roller machining experience.  It's been great to learn the basics of turning, facing, boring...and trying to hit exact dimensions and get good finishes. 

I solved the problem of the finish on the aluminum drums - I pressed in the end caps on the aluminum pipe that contain the bearings.  These were easier to turn between centers and didn't require use of a rest at all.  The finish turned out pretty well.

Since these pics were taken, I've welded feet onto the bottom and taken it for a test spin.  For those that are cyclists - Phil was right.  It turns you into a bucket of sweat after about 30 minutes.  Each drum spins easily enough, but there is something about that flywheel that adds pretty good resistance.   Makes me wonder if there is something to this "viscous drag" that he mentioned (while I could usually spin at 20mph, I now drop to a much lower  42x19 gear to maintain 13mph).