Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment   Page 2 of 2      Prev   1   2
electricpete

Sr. Member
Registered:
Posts: 647
Reply with quote  #16 
I don't know if the comments about building this design in a desperate situation are joking or not, but it got me thinking about the design a little. 

The torque-carrying component of the force is perpendicular to the axial direction. The tension in the band is approximately Torque / [Radius * sin(theta)  * B] where theta is angle from axial and B is the number of bands.  And of course the resulting tensile stress in the band would be tension divided by the cross sectional area of the band. 

The designers of that monstrosity at least had enough foresight to put enough slack into the band so that it formed an angle approximately 45 degrees from axial under load.  In that case approximately sin(45)~3/4 of the tension is in the torque producing direction (so the tension in the band is 4/3 of that needed to transmit the torque at that radius).  If they had less slack so that it was let's say at 30 degree angle, then only sin(30)=1/2 of the tension is in the torque producing direction (so tension in the band is twice what is needed to transmit torque).  If they made it really tight so there was almost angle zero when stationary and let's say 15% under load, then only sin(15) ~ 1/4 of the tensions acts in the torque direction and tension is 4 times as high as that needed to transmit the torque at that radius.

All of the above is under some very optimistic and simplistic assumptions.  

The most troubling assumption to me is that the band is able to rotate at the bolt to the proper direction.  Of course we could loosen the bolts to make it easy to rotate, but then we would lose the clamping action onto the rubber around the bolt, making it more likely for a tear to occur at the bolt.   So loosening bolts is not an option.  I imagine you'd want to set the rubber in the proper orientation for carrying load at the time the clamping bolts are torqued.  Looking at the stationary photo I see some evidence of that, but not as much as I expected (I wonder if their bolts are loose enough to allow rotation... which would put a tearing stress on the rubber at the bolt hole). 

Another assumption in my paragraph above is that the load angle is constant (so we have an ability to match it if we are careful enough). But actually the load angle would change as the load changes and there is different stretch.  So if the load angle is changing but we can only clamp the rubber in tightly at one angle, then we may end up in a situation where the rubber is clamped at an angle different than where it "wants" to be.  That means a non-uniform tension across the width of the band, increasing the local stress. Probably the best we could do is set it up for load angle associated with the maximum expected  load, and accept the non-uniform tension across the width at lower loads.  

Another assumption implied above is that if I allow the band to rotate to the proper direction, then it will be under simple tension as if it is pulled straight. But of course it still has to wrap around some corners which creates non-simple stretching with a twist component to it.  And if there is friction as the band wraps around a corner so that the band does not stretch along it's length uniformly under tension, there is higher stress in certain portions of the band.    


Reaction forces:
  1. Of course the bands have to be very well matched to share the load so that the radial components of the tension force cancel out among all four bands when you add them together (to minimize the radial forces on the hub).   Likewise matched to prevent mechanical unbalance. 
  2. While you could in theory eliminate the net radial force on the hub by perfect matching, you'll never eliminate the axial direction forces on the hub under load.  Those would be unavoidable and of course would affect the axial loading of the machine bearings.   We can calculate the axial force from the coupling as Tension * B * cos(theta) = Torque * cotan(theta) / Radius.  Smaller angles (band closer to axial direction) result in higher axial force. 


That's just a few minutes thinking about it. Obviously I don't recommend anyone to build this, even in desperation, and certainly not using my feeble attempt at analysis.  



Todd Smith

Avatar / Picture

Member
Registered:
Posts: 9
Reply with quote  #17 
Wow what that's a nice looking want to be Lovejoy coupling, what hp is the motor 450? Rpm can't be that fast 600 to 800 rpm? Must be trying to save on cost? Falk coupling with grid or Pix Powerware rubber coupling, maybe put in an soft start so its not so hard on the start up of the unit.
Very interesting, thanks for posting.
fburgos

Sr. Member
Registered:
Posts: 670
Reply with quote  #18 
Its a 1188rpm and 430kW about 600hp, he is trying to save costs.

last week new rubber belts and bolts were installed, everything same size, and to my surprise vibrations are "good".

pump vibration.png 

new rubber.png 




MarkL

Avatar / Picture

Sr. Member
Registered:
Posts: 1,121
Reply with quote  #19 
I suppose hardest thing is getting it started. Once running can't see why it wouldn't work.
Very agricultural :-) looks like how we would fix an issue on the farm if a machine broke .


fburgos

Sr. Member
Registered:
Posts: 670
Reply with quote  #20 
Can't see why bushings made of the same belt wouldn't work, it's faster and cheaper with less uncertain of unbalance condition or the axial pull from the tension of the belt... Now he's happy, and I learned something in the way
electricpete

Sr. Member
Registered:
Posts: 647
Reply with quote  #21 

Quote:
I suppose hardest thing is getting it started.

I agree, the bands have to reposition from their current angle theta = 0 to approx. 45 degrees under load.  That means they rotate under the clamping plate.  How quickly are they going to do that, and what happens in the meantime. I wonder. 

Quote:
axial pull from the tension of the belt

Torque (in ft-lbf) = 5250* Power (in hp) / Speed (in rpm)

Torque = 5250 * 600 / 1188 = 2650 ft-lbf

Assuming the radius is 0.5 ft, that is a circumferential-directed force of 2650/0.5 = 5300 lbf

If the angle under load is 45 as shown in the video, then the axial force will be the same as the circumferential directed force:  5,300 lbf. 

The motor bearings normally see no thrust.  The pump looks like a single stage double suction pump which also wouldn’t normally have much thrust. I assume both machines have rolling bearings. Certainly the axial loading will reduce the expected life by some amount (for deep groove ball bearings, axial loading counts approx. twice as much as radial load in terms of impact on bearing life according to the X and Y factors).   If you know the part numbers, maybe the expected effect on bearing life could be estimated quantitatively.  Otherwise, you can wait for the next bearing removed due to degradation and ideally try to examine it to see if axial loading contributed. Then you have more ammunition.   

Other hazards besides the bearings are the coupling itself degrading to where one strap breaks, then increase the load on the others so they break. Assuming no-one gets hurt, there’s still an impact on production.

The force on each band is 5300 / (4*sin(45)) ~ 1875 lbf.  With an apparently loose clamping force (to allow those things to swivel), that force is transmitted through the strap via the bolt hole in the strap (although that force is probably reduced by the the friction at the clamping surfaces and the friction where the belt wraps over the edge of the hub).   I have a hard time imagining why the bolt doesn’t tear through the strap at the bolt hole.

I have a hard time believing this would save money in the long run.

In spite of all my reservations, it looks like it operated fine for 6 months the  pump seal leak, then had a problem due to something that was done during maintenance.  It makes me wonder: was the pump bearing changed out in order to gain access the seal, or was there a bearing defect? 

 

fburgos

Sr. Member
Registered:
Posts: 670
Reply with quote  #22 
Quote:
Originally Posted by electricpete
In spite of all my reservations, it looks like it operated fine for 6 months the  pump seal leak, then had a problem due to something that was done during maintenance.  It makes me wonder: was the pump bearing changed out in order to gain access the seal, or was there a bearing defect? 


The modification was done with this repair, it used to be a standar pin and bush at least for last 5 years, repair was mainly because the seal leakage (wore off) and bearings were in the way.

Pump has 6321C3 both bearings

Motor has NDE: NU222 and DE: NU224-6224
Todd Smith

Avatar / Picture

Member
Registered:
Posts: 9
Reply with quote  #23 
Well this coupling could be the culprit of the seal failure? Is the seal carbine, which I think the rotating element is and the stationary as well. This vibration from the coupling could have an effect on the seal. Did the drive side seal have the failure? Or was it the non-drive side.
fburgos

Sr. Member
Registered:
Posts: 670
Reply with quote  #24 
Quote:
Originally Posted by Todd Smith
Well this coupling could be the culprit of the seal failure? Is the seal carbine, which I think the rotating element is and the stationary as well. This vibration from the coupling could have an effect on the seal. Did the drive side seal have the failure? Or was it the non-drive side.


perhaps in the future, this new seal/coupling (artisan, homemade from scrapped belts) has not failed its from june repair.

the original seal and original/coupling (pin and bush) lasted for longer than 4 years until june.
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.