Rocker Shaft Seals

[chris_seven]

This is an old problem and so far I am not sure that the available solutions are particularly effective.

The RSR Seal can be difficult to fit as they simply have a tendency to shear and they are not always effective depending on the scoring of the bore.

We tried to find a design which used a shaft made of more than one part but this would mean that the rocker load would have transferred to the carrier through a relatively small cross-section which we didn't like and abandoned.

One available solution uses O rings that work on the ends of the shafts and this makes fitting quite straightforward.

We looked at this solution in detail but we weren't happy with some aspects of the design.

The Bush and Nut use a long taper that expands an O Ring outwards and effectively presses it into the bore of the cam carrier forming a seal.

This sounds quite good, makes seals easy to fit and relatively inexpensive but is does have issues.

The volume available for the O ring is very small and even at a 1mm cross-section the O Ring is too large.



Rubber is, of course, incompressible although it will readily change shape.

This means if the available void is too small for the O Ring any axial load will transfer through the O ring into the carrier and will not expand the end of the rocker shaft.

In effect the O Ring will not only seal it will also retain the shaft.

Rubber will also, after a certain time at temperature will take a set and the ability of the O Ring to locate the shaft will reduce.

We changed the basic design of the Bush and Nut to have a parallel section which uses the O ring in a conventional manner and meets the design criteria outlined in the Parker Hannefin Design Manual.



This works as a seal, has space for the rubber and will transfer load to allow the end of the shaft to expand and clamp into place.

It does, however, leave the nut sticking out of the carrier and the small groove it produces will almost certainly trap dirt which we think is undesirable.

If the nut is positioned so it is flush with the end of the carrier then the distance from the groove in the shaft to the edge of the carrier reduces to 0.4mm.

Whilst this is probably OK it doesn't look great.



One solution is to remove the groove.

A 'grubby' FEA analysis shows that the groove doesn't contribute to the expansion of the shaft end so it could be eliminated.

We feel, however that the sharp edge does provided some grip and also we think that the removal of the groove would be unacceptable to many potential users - so the groove stays.

The only other option is to reduce the length of the machined section of the shaft that radially expands.

This will, of course, mean that the load transfer capability will reduce and that clamping forces will need to increase.

As 'overtightening' the bolt is commonly recommended this is an issue.

The solution is to re-manufacture the Rocker Shaft in a Grade 4 6AL4V Titanium Alloy.

The lower modulus of this material allows the shaft end to expand about twice the amount for a given axial force, it is therefore much better at transferring load.

By reducing the length of the shaft in this area we would only need a 10% increase in clamping force if the shaft were steel so we still have a shaft that will clamp more effectively than the standard component.

The overlap becomes 1.4mm which is identical to the standard part and 6AL4V has more than sufficient strength for the required duty.



The only issue concerns the ability of Titanium to tolerate the oscillatory motion and this means that a coating needs to be applied.

We will use a Titanium Nitride finish which will be applied in a PVD chamber.

We have tried some basic samples from a fit point of view and it all works.

We have just ordered a batch of shafts and custom made Viton O rings with an optimum diameter and cross-section.

[afterburn 549]

I have not ever had a problem with a dab of silicone grease on them.

[porterdog]

Neat.