Belt Drives and design

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

plncraze

Well-Known Member
HBA Supporter
Joined
May 11, 2006
Messages
2,104
Billski said in his "Springs,etc" post that a reduction drive frequency must be 25% away from the engine's critical frequency. Den Hartog fixed a Pratt engine by using a softer element in the gearbox. The post above shows why that works. I really want to have a pusher prop installation so until I change my mind I am stuck with long shafts. The budget on this project kind of forces me not to buy new parts.
 

plncraze

Well-Known Member
HBA Supporter
Joined
May 11, 2006
Messages
2,104
The attached publication from Gates has a formula in it that is supposed to give Young's Modulus for fiberglass reinforced belts. I have been playing with this using numbers from EPI-ENG's website about the Gates belts and Prodrive and some of DanH's numbers for the Panther belt. The formula is listed below:
T*L/E1*W
T=tension, L= span length, E1 is the elongation of span and W is belt width in inches. As in the bifilar example that DanH has posted the units are in inches and pounds so the answer should be in inch-pounds. Now I have to convert my answers to foot-pounds per radian. What am I missing? Thanks in advance
 

Attachments

plncraze

Well-Known Member
HBA Supporter
Joined
May 11, 2006
Messages
2,104
One last question for the day. What is the correct way to use the driven inertia adjustment in the spreadsheet that was posted a while back? I just used the prop's inertia since it was the largest and used that number as the final inertia in the Holzer program rather than the individual pieces downstream of the belt. For those who have not used the Holzer program on Tom Irvine's website you enter the MMOI numbers for your inertias and the stiffness numbers for your shafts in Ft. pounds per radian.
 

plncraze

Well-Known Member
HBA Supporter
Joined
May 11, 2006
Messages
2,104
After playing around with formulas this morning trying to calculate belt stiffnesses I tried adjusting belt stiffnesses in the Holzer program. Going from 60000 ft lb per rad to 80000 ft lb per rad in 5000 increments with everything else being left the same the lowest critical rpm I had was 7144 and the highest was 9024 rpm. As Rotax did I might have to limit rpm if all this remains accurate which it probably won't. I have not yet done any calculations which include the springs in the clutch plate.
 

DanH

Well-Known Member
Joined
May 20, 2019
Messages
52
I have not yet done any calculations which include the springs in the clutch plate.
Uhh, John, although they may provide some personal educational value, we probably don't need to read through GIGO examples.

First establish a practical drive configuration, with weight and packaging foremost. So far, what you have is rather clunky. Long shafts are not typically necessary for pusher airframe.

If money is a concern, perhaps it would be a practical educational exercise to model (i.e. reverse engineer) a really simple system which flew successfully. For example, consider the VW redrive developed by Gene Smith. Doesn't get more basic, and it appeared to work very well. "Appeared" is the operative word, so see what a model says.

 

DanH

Well-Known Member
Joined
May 20, 2019
Messages
52
John, to your questions...

What is the correct way to use the driven inertia adjustment in the spreadsheet that was posted a while back? I just used the prop's inertia since it was the largest and used that number as the final inertia in the Holzer program rather than the individual pieces downstream of the belt.
All the inertias on the driven side are adjusted, so here the upper sprocket, propshaft, prop hub, crush plate, bolts, spinner if applicable, and the blades.

The attached publication from Gates has a formula in it that is supposed to give Young's Modulus for fiberglass reinforced belts.......Now I have to convert my answers to foot-pounds per radian.
Note the sentence on page 7: "The deviation from this curve by any single belt can be as great as ±30% due to normal material variations." That being the case, and already having the sprockets and belt in hand, you're probably best off to set up a fixture and just measure directly rather than fool around trying to get data from Gates.

Firmly fixate the big sprocket. Put the small sprocket on an axle with a socket on the end so you can pull it with a large torque wrench. Fasten a digital protractor to the small sprocket. Tension the belt to a static value like you will use in service. Now pull the torque wrench to five values, 40 through 120 ft lbs, since your operating torque will be in that range somewhere. Record the number of degrees of rotation at each torque. Plot them. If the plot (or even just the upper three values) forms something close to a straight line, use the one closest to your nominal operating torque to extrapolate torque per radian. For example, if an 80 lb pull on the torque wrench results in 1.7 degrees on the digital protractor, equivalent torsional spring rate of the belt and sprockets is (57.29578 x 80)/1.7 = 2696 ft lbs/radian.
 
Last edited:

plncraze

Well-Known Member
HBA Supporter
Joined
May 11, 2006
Messages
2,104
Very practical! I had not thought about using a torque wrench.
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
8,226
Location
Saline Michigan
Here is a question for the group: I am using the friction part of a factory Chevy Metro clutch. This has springs in it. Do I calculate these in like a stiffness? Also does anyone have a figure for Shor A stiffness? My Lovejoy spider talks about angular limits but does not provide any torsional stiffness numbers.
I am still working on figuring driven inertia as well.
Bolt it down, put on a degree wheel on the output side, and twist this stuff with a big torque wrench. Now you know how stiff they are. For the coil spring units arranged as torsion springs, you can also do standard spring rate and stress calcs (right out of Shigley or the spring industry literature) and then figure in the working radius to calculate the rates...
 

DanH

Well-Known Member
Joined
May 20, 2019
Messages
52
Bolt it down, put on a degree wheel on the output side, and twist this stuff with a big torque wrench. Now you know how stiff they are.
Yep. Illustration from the wayback machine, a measurement made 20 years ago...clutch springs for a 2000 Subaru Legacy 4wd. Still have the spline shaft with a welded socket for the torque wrench. For accuracy, the degree wheel was about 3 ft radius, with a length of gas welding rod for a pointer.

Dual rate springs, hits the stops at about 5.5 degrees.
 

Attachments

Top