Involute Spline Fretting (constant mesh pto)
Involute Spline Fretting (constant mesh pto)
(OP)
I have a hydraulic pump with a 7/8"-16/32DP straight involute external HT alloy steel spline shaft. The splines are experiencing contact fretting. (See spec's below and pic attached). The pump connects to internal splines of a constant mesh PTO, which in turn is bolted to a truck transmission and is powered by a heavy duty diesel engine. The spline is greased via a shaft centerline thru-hole which enables cleansing of the fretting debris.
Diametral Pitch: P = 16
Number of Teeth: N = 13
Pitch Diameter: D = 0.8125 in
Face Width: F = 1.31 in
Torque: T = 1550 in-lbs
Depth of Engagement: h = 0.05625 in
Compressive Stress: Sc = 3983 psi
In addition to the compressive stresses under load, there are torsional vibrations imposed by the drivetrain. Would you expect that the fretting is primarily caused by insufficient lubrication, alignment error, torsional windup, edge loading, shock loads, or something else? What are some options to limit contact fretting? My immediate options are a larger shaft, 1"-15T, and/or an oil bath lubrication system.
Thanks,
David
Diametral Pitch: P = 16
Number of Teeth: N = 13
Pitch Diameter: D = 0.8125 in
Face Width: F = 1.31 in
Torque: T = 1550 in-lbs
Depth of Engagement: h = 0.05625 in
Compressive Stress: Sc = 3983 psi
In addition to the compressive stresses under load, there are torsional vibrations imposed by the drivetrain. Would you expect that the fretting is primarily caused by insufficient lubrication, alignment error, torsional windup, edge loading, shock loads, or something else? What are some options to limit contact fretting? My immediate options are a larger shaft, 1"-15T, and/or an oil bath lubrication system.
Thanks,
David





RE: Involute Spline Fretting (constant mesh pto)
Ron Volmershausen
Brunkerville Engineering
Newcastle Australia
http://www.aussieweb.com.au/email.aspx?id=1194181
RE: Involute Spline Fretting (constant mesh pto)
Thanks,
David
RE: Involute Spline Fretting (constant mesh pto)
Ted
RE: Involute Spline Fretting (constant mesh pto)
The application is a hydraulically driven fan motor. The fan produces a variable load on the pump. The fan control is proportionally driven and the fan motor has built-in trim speed control.
There are 5 components, (1) the pto, (2) the pump, (3) the fan motor with integrated flow control, (4) a fan drive control assembly (FDCA), and (5) an electronic control module (ECM).
The ECM produces an electrical PWM signal which feeds the fan drive control assembly which then sends a proportional hydraulic signal to the fan motor flow control valve. This allows the ECM to gradually ramp up and ramp down fan speed and to arbitrarily set the fan speed.
At engine idle most of the pump pressure is routed around the fan motor - the pump is mostly freewheeling. The engine produces significant torsional vibration at idle which is fed thru the transmission and results in clattering of the PTO drive gears. I don't know how to quantify the cyclical and possibly reversing nature of the loads on the splines during idle.
The manufacturer is in favor of an oil bath lubrication system which draws and returns fluid from/to the transmission. I'm concerned over passing spline debris into the transmission.
RE: Involute Spline Fretting (constant mesh pto)
Ted
RE: Involute Spline Fretting (constant mesh pto)
I expect the shaft that mates with that female spline in the hub is junk now, too, so will destroy a new hub in short order.
As other's said, rev it up.
More flywheel on the engine side might help smooth things out.
I's be thinking about a way to apply enough torsional preload to keep the spline from rattling too.
RE: Involute Spline Fretting (constant mesh pto)
Thanks,
David
RE: Involute Spline Fretting (constant mesh pto)
https
Thanks,
David
RE: Involute Spline Fretting (constant mesh pto)
The width of the wear lip is uneven suggesting that there is an uneven load distribution.
I can't see any evidence of "torsional vibration"......this would create wear on both flanks not just the loaded flank.
A once off application of grease is the worst thing you could use in this environment.......you'd be better off assembling the parts dry.
Ron Volmershausen
Brunkerville Engineering
Newcastle Australia
http://www.aussieweb.com.au/email.aspx?id=1194181
RE: Involute Spline Fretting (constant mesh pto)
Ron Volmershausen
Brunkerville Engineering
Newcastle Australia
http://www.aussieweb.com.au/email.aspx?id=1194181
RE: Involute Spline Fretting (constant mesh pto)
The pump external shaft uses DU bushing journals, one on either side of the pump gear, width: 1.125" x ID: 1.30".
The PTO helical shaft is supported on cone bearings with a lock ring. It's an SAE 'B' shaft. I don't know the bearing dimensions. The shaft is 1.25" OD, and the bearings are approximately 3.5" apart.
In my OP I had the wrong 'Face Width'. Here are the corrected numbers:
Diametral Pitch: P = 16
Number of Teeth: N = 13
Pitch Diameter: D = 0.8125 in
Face Width: F = 0.8335 in
Torque: T = 1550 in-lbs
Depth of Engagement: h = 0.05625 in
L/D = 1.026
Compressive Stress: Sc = 6260 psi
I'm not sure if I can risk a wet spline lubrication system. There's no filtering of the spline cavity output fluid before returns to the transmission.
David
RE: Involute Spline Fretting (constant mesh pto)
Here is a link where you can check the numbers:
http://www.zakgear.com/External_Spline_IN.html
Ron Volmershausen
Brunkerville Engineering
Newcastle Australia
http://www.aussieweb.com.au/email.aspx?id=1194181
RE: Involute Spline Fretting (constant mesh pto)
A smaller pump would reduce the pump's torque reaction capacity and reduce the resistance to the engine impulses and should reduce the spline loading.
Ted
RE: Involute Spline Fretting (constant mesh pto)
Ron, I used your link to check my numbers and they do checkout, but these are steady state conditions. There may be transient conditions which likely push the compressive stress well beyond my 6K psi steady state.
The pto is constant mesh, so the pump is always spinning. A common condition is having the fan turn on when the pump is spinning close to it's max RPM and delivering it's max GPH at low pressure.
At this point the ECM should turn the fan motor on gradually to avoid pressure spikes. In older equipment gradual fan turn on may not be effective, so there may be a pressure shock wave sent thru the system as the system tries to adjust to it's new equilibrium.
So I may be seeing metal fatigue do to hydraulic transients. I need to brush up on my fluid dynamics.
Ted, The fan motor requires 40 hp to move the CFM need to cool the heavy duty diesel engine. Unfortunately this is the smallest pump available that does the job at the PTO rpm's required. Thanks for the suggestion.
I think you guys are correct, the splines and/or material isn't adequate for the job. The stresses are likely higher than I thought.
David
RE: Involute Spline Fretting (constant mesh pto)
Torsional vibrations can be measured with battery-powered strain gage telemetry system or optical encoder and F-V (torsional) converter. Test the system with new components to identify torsional natural frequencies. Install a rotating mass, incease shaft stiffness, or add flexible (elastomer type) coupling to detune the system, as necessary. If you really want to get rid of this problem forever!
Walt
RE: Involute Spline Fretting (constant mesh pto)
What are the flow and pressure requirements of the fan? Just curious.
Ted
RE: Involute Spline Fretting (constant mesh pto)
Can you post a screen shot showing the entered values?
Ron Volmershausen
Brunkerville Engineering
Newcastle Australia
http://www.aussieweb.com.au/email.aspx?id=1194181
RE: Involute Spline Fretting (constant mesh pto)
You're correct. After paying special attention to the correction factors in your spreadsheet (at zakgear.com) it appears that the compressive stress of the SAE 'B' shaft is on the edge.
I have the following factors:
Application Factor, Ka, 2.2: (driven by diesel engine, light shock)
Load Distribution Factor, Km, 1.5: (0.004" per inch of Face Width)
Life Factor Limited by Fatigue, Lf, 0.5: (number of torque cycles)
I tried to be reasonable conservative with these factors.
Also, I'm not yet sure of the material so I chose a reasonably conservative option: Surface Hardened Steel.
Thanks,
David
RE: Involute Spline Fretting (constant mesh pto)
Attached is a chart showing pressure, rpm, and flow requirements.
Thanks,
David
RE: Involute Spline Fretting (constant mesh pto)
I'd certainly like to see, at the very least, a 1.5x – 2.0x safety built in to the design if alignment cannot be guaranteed.
Most of the splines we work with in the mining industry generally operate at 10x – 15x safety factors.
Ron Volmershausen
Brunkerville Engineering
Newcastle Australia
http://www.aussieweb.com.au/email.aspx?id=1194181
RE: Involute Spline Fretting (constant mesh pto)
1) a 'compressive stress in teeth' margin-of-safety of 0.349 versus -0.129, and
2) a 'shear stress in teeth' margin-of-safety of 0.487 versus -0.032.
Better, but nothing to write home to mom about - certainly not good enough to pipe the effluent into my transmission, unfiltered.
David
RE: Involute Spline Fretting (constant mesh pto)