Wear 1

[rm2]

Hello

We are having some heavy wear failures on field returns from camshaft-tappet.
Later I will attach a document where you can appreciate the magnitude of the failure, also some images from parts we catch during testing.

Much work has been done with no successful results.
Some of the theories are:

Cleanliness, metallic debree.
Lack of injection oil to the cam body caused by orifices plugged at the test stand, oil conduct from tappet plugged, body non magnetic chips, etc. Injection timing was increased. A huge cleanliness project is ongoing, levels are much better, failures still occurring.
I have turned on and off the failure by “locking” the tappet (basically putting some loctite).
I will attach later some drawings from the cam and the tappet. We have not found dimensionally parts out from the drawings.

We tried changing to synthetic oil, no success at all, we continue to have failures.

I will really appreciate if you can guide me on how to solve this issue. I would like to understand the load that the camshaft and tappet roller sees (I will expect this is a key point).
I read interesting upload DLC post from a pdf from M. Priest, C.M. Taylor, is it possible to know under what regime of lubrication this design is working?
How can I check the compressive fatigue strength?
How can I check if the proper tolerances and profile placed on the drawing are correct?
You will see that the tappet rotates apparently under hydrodynamic regime over a bronze pin which is assembled by pushing it through a cryogenic shell, any suggestions on how to avoid wear (eventually we find bronze shavings)? How can I establish the optimal clearance between roller an bronze pin? Anyone has experience on ceramic rollers, is it a better option?
Is it a good idea to start using Diamond like carbon?

Thanks all in advance for your help.

 

[ivymike]

the loads in the valvetrain can be simulated using software (Ricardo Valdyn, for example). The resulting stresses in the cam and follower can also be evaluated in software, given the geometry and the loads.

from your post it was not entirely clear to me what type of engine you have, what the valvetrain layout is, and whether you're using hydraulic lash adjuster (hydraulic tappets). It's also not yet clear where your failures have been - the bronze pin, or cam-follower interface?

Are you sure that you have adequate valve lash (clearance) in the system?

 

[rm2]

Sorry for not providing full information.
This is happening on a fuel diesel pump.
Failure is happening between the cam-follower interface (you will understand everything once I upload the photos, hope to upload them tonight).
I do not discard that the follower might got stuck and then friction will cause cam-follower to heat up, that was the reason for providing details on how the follower(tappet) bronze pin is assembled.

 

[rm2]

here the photos

 

[tbuelna]

rm2,

Based on your posts and the drawing you provided, it looks like this is some sort of cam-driven injector pump. You also stated that the follower is a roller type, the pump housing follower bores are bronze sleeved, and you have experienced follower seizures and bronze metallic debris particles in the oil.

If the follower bodies are a ferrous material (ie. steel) operating in a non-ferrous liner (ie. bronze) and housing, then if you have a sliding clearance at room temp, the sliding clearance at operating temps should only increase. So thermally-induced seizures should not be occurring.

The only issue that I can imagine, that is causing failure of your lifter bore liners, is due to the side loads produced by the pressure angles of the cam profile against the roller follower. If you have a large diameter roller follower and a very aggressive cam lift profile, the side forces produce by the instantaneous cam/roller pressure angles may be causing excessive side loads on the follower body.

Good luck.
Terry

 

[WGJ]

Could see a drawing, but no pics

Bill

 

[rm2]

It might be the case, the analysis attached was based statically, which is the incorrect way (it should be evaluated dynamically I will expect). However you can have a better picture of the problem with this file.
Is there a practical way I can acquire the true load that the cam and follower sees?
Any comments or recommendations?

 

[tbuelna]

rm2,

Sorry for the tardy response.

I'm a little bit confused. In your original post you mentioned that the debris being generated was non-magnetic (ie. non-ferrous). But in the later posts you state that the failure is between the cam and roller follower, which I'm assuming are both steel?

The analysis document you linked shows a relatively mild displacement of <10 microns at the contact zone. The roller follower should have some profile crown ground along its surface, so that small deflection should not be a problem.

You did not provide the diameter of the follower body, but estimating from your slides, it looks like there is more than adequate contact area between the steel lifter OD and it's bronze-lined bore for any side loading it might experience. The lifter side loads are a function of the instantaneous pump pressure forces plus return spring forces, acting at the instantaneous cam/follower pressure angle combined with a friction coefficient. However, since it appears that you have a small diameter roller and a large base circle cam, I would assume that the max follower pressure angle is probably very small.

The last thing you might want to check is the materials and heat treatments of the roller follower and cam lobe. The roller follower should be through hardened to Rc58 or better, and the cam lobes should be case hardened to Rc58 or better to at least .030 inch deep. Checking case hardness normally requires a (destructive) micro-section of the parts.

By the way, what type of elements did you use in your FEM to simulate the contact conditions at the roller/cam lobe interface? The gap elements used should have nodal DOF's that allow only compressive forces to be transmitted along the element. These types of contacts are very difficult to simulate in a linear type FEA, as your slides show.

Hope that was helpful.
Terry

The location and configuration of the oil feeds is probably not too critical, unless heat transfer is an issue, which does not appear to be the case.

 

[dicer]

Wrong choice of material. Didnt' see any pictures. DOC's most always can have a virus attached.