How gear train losses vary with speed

[j2bprometheus]

Hi-

I am want to know how gear train losses (friction) vary with RPM. I suspect that losses increase in a non-linear manner.

Suppose that I had two spur gears with a 5:1 reduction ratio, and I increase the speed from 1000 RPM to 5000 RPM, how would the plot of friction losses versus RPM look?

I found a little bit on gear train friction at this link:

http://gltrs.grc.nasa.gov/reports/2006/TM-2006-214344.pdf

Can anyone recommend other references?

 

[plasgears]

When you go from low speed to higher speed, there is more heating in the lube, and a lower friction will result. In addition, a film will develop at rotating interactive surfaces. This is common to all machinery. I never used it in my gear train calcs. A blanket 95% effic. per stage should suffice for most applications. Worm gears and similar gears have effic. as low as 50%.

A turbine design engineer would have a governing design point of 'cold start,' when figuring starting torque. The existence of ice can greatly boost this starting torque.

 

[j2bprometheus]

plasgears-

Thanks, but I am confused.

When I look at the experimental results in the attached plot from the NASA link above, it shows the efficiency being lower (and hence friction higher) as the speed goes from 15,000 RPM to 12,500 RPM.

This seems to contradict the argument about " more heating in the lube and a lower friction will result."

Any thoughts?

 

[hydtools]

I suspect that the improved efficiency indicated in the NASA data is the result of better hydrodynamic performance of the lubrication in the tooth contact area resulting in less tooth friction. The higher speed better feeds lubricant into the tooth contact area. Similar to the effect of speed feeding lubricant into hydrodynamic bearing surfaces, improving load support and reducing friction.

Ted

 

[vc66]

The NASA graph agrees with Plasgears. As speed goes down, friction goes up. That's what Plasgears said.

Am I missing something?

V

 

[hydtools]

vc66, no I don't think you are missing anything. The OP is.

Ted

 

[j2bprometheus]

Guys-

The NASA plot shows efficiency versus load, NOT friction versus load.


The NASA data shows that low speed (black line) has better efficiency than the high speed (red line)

This contradicts what plasgears said above.

 

[BobM3]

That's a confusing plot. It seems to show that efficiency is lower at the higher speed but you have to be careful because it's a plot of power vs efficiency (so, different torques at the two speeds). I think you really want a plot of efficiency vs speed with the torque held constant.

 

[j2bprometheus]

BobM3-

You are correct that efficiency is plotted versus power. So torque is varying - good catch.

I am used to how friction torque varies in IC engines: goes up slightly faster than linearly with speed over the normal engine speed range.


Surely there is a plot in a book somewhere that shows how friction in a gear train changes with RPM, assuming constant torque.

 

[BobM3]

I'd try looking at

http://www.AGMA.org

or

http://www.SAE.org

for papers.

 

[tbuelna]

j2bprometheus,

With gears it's not really RPM that's important, it's pitch line velocity. At low pitch line velocities (<5000 fpm), sliding losses at the tooth contact will likely predominate. At higher pitch line velocities, gear tooth windage losses and bearing churning losses can get out of hand quite rapidly. At pitch line velocities above 10,000 fpm, windage and churning losses can be several times the gear tooth sliding losses, unless the oil scavenge system is carefully designed. Just picture the gear teeth as little propellor blades spinning through a dense atmospheric mixture of air and oil droplets.

And ultimately, all of these windage and churning losses are converted to heat in the lube system.