Thesis on the fretting fatigue phenomenon in splined couplings.

[gearcutter]

Here's a link to another excellent thesis on the fretting fatigue phenomenon in splined couplings.
The work was conducted from an aerospace perspective.

REPRESENTATIVE FRETTING FATIGUE TESTING AND PREDICTION FOR SPLINED COUPLINGS.
By DEAN HOUGHTON, BEng, University of Nottingham.

http://etheses.nottingham.ac.uk/1013/1/Thesis_submission-final_5.pdf

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[tbuelna]

gearcutter,

Thanks for the link to this 200+ page PhD thesis on the subject of fretting in spline couplings. The author coordinated this study with some folks at Rolls-Royce. I have worked for Rolls-Royce myself, and I have spent more than 10 years designing components for aircraft propulsion systems, including many highly loaded spline couplings. During this time, the one thing I have learned about the design of spline couplings for critical aircraft applications is to use the established conservative approach, otherwise your analysis will not be accepted by those reviewing your design. The reality is that the aircraft world is still very conservative when it comes to things like spline analysis. The crowned splines of turboshaft engine output shafts (from GE, P&W, R-R, etc.) are all still based on very modest P/A contact stresses and a specific flow of lube oil through the spline to flush out debris generated from fretting.

In reality, there is not much to be gained from more precise analytical techniques for splines, in terms of weight or cost. So this PhD thesis is mostly of academic interest.

Best regards,
Terry

 

[gearcutter]

tbuelna

What you say is probably right.

You guys in aerospace are so lucky that the systems you work with include continuous lubrication for splined couplings.
One of the few applications where this is still possible.

Without the luxury of a continuous flow-through lubrication system; I feel that designers need to keep themselves familiarised with the latest developments and study in the field, particularly with the fretting phenomenon.
Large non-lubricated gear couplings are common in the steel and mining industries with some being well over 1 meter in diameter and costing 10s of thousands of dollars to produce.
As materials and manufacturing technology rapidly changes; the push to make industrial transmission components lighter and therefore cheaper are steering many away from the typically conservative approach.

As you know; the same is occurring in the gear industry.
Typically; a modern 500Kw gearbox, for example, can be up to 3x lighter and cheaper than a similar box would have been 30-40 years ago.








Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[spigor]

I have seen some helical external splines in the thesis. It says, that they are helical to handle axial loads. So the axial force generated by the helical spline must be greater then the axial force to handle (and of course opposite direction). If there's a peak in the axial force- would it disconnect the spline coupling? Does the mating spline have the same helical angle?

 

[tbuelna]

spigor,

I don't recall ever seeing a specific example of what would be considered a helical spline being used to intentionally create an axial force at the joint. However, I have seen many instances where lead correction was applied to the spline teeth to compensate for torsional deflections under operating loads. This is common with splines that have large L/D ratios and/or thin walls backing them up.

I'll have to read thru that paper again, to see what I missed.

 

[spigor]

tbuelna,
exactly! Check out the Figure 1.3 p.10, Figure 2.25 p.69, Figure 2.29 p.71, and on the p.2: "The spline teeth are usually involute in form, and axially they may follow a helical profile in order to accommodate axial loads."

 

[CoryPad]

In the automotive industry, there are some applications that use a small helix angle to prevent axial separation.

 

[tbuelna]

spigor,

I took a look at the two figures you noted, and they both are definitely examples of splines with extreme L/D ratio and low relative torsional stiffness. Normally, lead correction or crowning is required with involute spline joints when the L/D ratio is around 1.0 or more.

 

[tbuelna]

You guys in aerospace are so lucky that the systems you work with include continuous lubrication for splined couplings.....One of the few applications where this is still possible.

Designing spline joints for aircraft propulsion systems is not the cakewalk you describe. You should consider that the reliability required from these splines must be close to perfection. For example, the spline of a modern 2500hp turboshaft engine output shaft must operate flawlessly for 6x10^9 cycles or more, at rotational speeds of 15-20K rpm, and with the absolute lightest weight possible. Plus the fatigue life expected from aircraft splines, gears, bearings, etc over the past few years has doubled or tripled. A decade ago, helicopter gearboxes were typically expected to last for around 2000 hrs, but now they are expected to have an MTBR of around 5000 hrs. And the accessory gearboxes used on commercial turbofan engines are currently designed for an MTBR of 30,000 hrs or more.