Bearing in wrist pin

[Allen3]

I'm designing a machine with a moving joint that looks and functions exactly like the joint between a piston and its connecting rod in reciprocating engine. My thought is that I would use the same type of bearing.

AFAIKT, a typical engine will attach the piston to the connecting rod using either (a) plain sleeve bearing, or (b) a needle roller bearing.

Is my understanding correct, that these are the two options usually used? Which is more common in large or heavy duty diesel engines? If the engine uses a plain bearing, how is it typically lubricated? Which wears out faster in this type of application with a heavy load, a plain sleeve bearing or a needle bearing?

Thank you for any pointers you could provide.

 

[Lou Scannon]

I'm not a bearing engineer, but I understand that wrist pin is a very suboptimal application for a plain bearing, due to the relatively low speeds & reversal of direction; it's not so easy to maintain the desired oil film thickness.
The Napier Nomad aircraft diesel engine that was developed in the 1950s but not put into production had a novel piston pin bearing design that had two slightly eccentric journals at the rod small end; this allowed the oil film to be replenished with each stroke as the pin alternated from one journal to the other.

 

[MikeHalloran]

Due to the limited rotation angle and the high load, wristpins are not a real good application for any kind of roller bearing, with a possible exception being needle bearings with relatively small needles. Not that they are guaranteed to fail; they're fairly common in two-strokes, where lube oil flow may be below 1 drop per minute.

Classical wristpin bearing is bronze or similar, pressed into the rod small end, with the wristpin retained by snaprings in the piston pin bosses.

More modern, at least in The Colonies, is wristpin pressed into the rod small end, bearing on native aluminum alloy in the piston pin bores, no snap rings required.

Schemes for wristpin bearing lube include oil cup (really a counterbore) in the small end of the rod, long hole drilled from rod big end to rod small end set up to align with rod journal oil feed holes once per rev, or in V type engines, notches in rod parting face to squirt oil into opposite piston crown. ... i.e., lots of things will work, but the odds of getting it right on the first try are not large.

I suggest disassembly, measurement, modeling and very close study of an engine design believed to have relatively few wristpin or piston failures.

An alternative strategy is equally close study limited to junkyard engines with rods sticking out of the block, which have the advantage of being cheap, the objective of course being to learn what not to do.





Mike Halloran
Pembroke Pines, FL, USA

 

[Tmoose]

does you load reverse directions? How high are they?

 

[Allen3]

Hi Tmoose,

The load in this application does not reverse direction. It is always compressive.

The machine is cyclical, with a design speed of 1200 rpm. The cycle will start with the joint at a zero degree angle with no load. With the joint held stationary, the load will increase to about 50 kN, growing roughly linearly. The bearing could be as long as 10 cm, so that would be roughly 5 to 10 kN per cm. The joint will then move to an angle of 40 degrees and immediately back to an angle of zero, following roughly a half-sinusoid acceleration/velocity profile. While the joint is moving from zero to 40 degrees, the load will drop to zero, at again a roughly linear rate, and the load will remain zero as the joint moves from 40 degrees back to zero degrees. The cycle would then repeat.

The machine will need to operate intermittently (about 20,000 stop/start cycles) with bearing life of about 2000 hours (150 million cycles) and a premature failure rate of less than one in a thousand.

My general sense is that a plain bearing with forced oil lubrication will work. This would result in boundary layer lubrication. We may in the future build a version of the machine with a longer running life, and for that application, a hydrostatic bearing might be needed.

 

[Lou Scannon]

The loading on the wristpin of the Napier Nomad engine was normally compression only.

 

[dicer]

Most all large diesel engines use a bushing, some are intermitently lubricated through the rod from the crankpin and some are just splash, most diesel engines have piston squirt nozzels to spray oil to the underside of the piston for cooling and upper rod lubrication and thus oil falls out onto the rod small end that sometimes has a hole drilled in the end.
The only place that needle bearings are found that I can think of is small 2 stroke engines.

 

[tbuelna]

Allen3,

With oscillatory rotary motion, a plain bearing would probably be best. Since at the point where the joint rotation stops and reverses direction, both the rolling element bearing and plain bearing would be briefly subject to boundary contact conditions.

The oil flow rates to a plain bearing should be sufficient to provide enough heat transfer to keep the bushing material within acceptable operating temps. Ideally, oil flow to the back side of the bushing's loaded sector would be best. Flow into the journal gap at the unloaded sector of the bushing would also work well.

With your non-reversing radial loads and relatively large oscillatory angles, a needle roller bearing could be made to work. But this bearing would likely have a much lower fatigue life than a similarly sized plain bearing. The needle bearing fatigue life would likely be determined by the number and magnitude of load cycles at the most highly stressed sector of the inner race (ie. wrist pin) surface.

Finally, you have specified a statistical failure rate of 0.10% in 2000 hours, which is quite stringent for a typical rolling element bearing.

Good luck.
Terry