Automotive cylinder head finish

[Rickabilly]

(07-28-2012, 03:38 PM)Sunset Machine Wrote: "copy the existing finish".

Yes, but Hmm, No.

When was the last time that the replacement head gasket was anything like the OEM one that was present from manufacturer? I know with many of the DAF engines we were using we seemed to be getting a different gasket each time we rebuilt one which was at least one every month usually more than that, in the very last BIG CAM Cummins we did the gasket set came with two different head gasket styles in the one set, three heads two gaskets of one style and one of the other, when we enquired we were told that they were interchangable and not to worry as they both ran on the one part number, we used them without issue. Then I did some Subaru stuff, and the replacement gaskets were consistant but the OEM ones seemed to vary greatly. This is just my most recent work, I've been inside engines for a few decades now and even designed and manufactured a few cylinder heads in my time, I have also worked with designers of engine line machine tools and the requirement has always been flat and smooth, "finish" was not really discussed beyond "Acceptable or not"

My opinion is, With the greatest respect to Tony Wells for giving the correct answer, Which there is no doubt he did. But still my opinion is that; surface flatness and a lack of "leakage path grooves" is more important than "Finish".

While I respect that there is a difference in thermal expansion between Ally and Cast iron, when I do the sums the result is that thermal expansion is no big deal really. Given a 20" long Ally head on a Cast Iron Block the difference in growth over 60 degrees C of change in heat which is about normal in most places where people use cars the total difference in change is about 0.014" then considering that this is halved because the centre will stay aligned and the two ends will move, and also given that there will be some deflection in the head and block as they are often dowelled together on many engines, the "Slippage" required is less than about 0.006" at the very ends with much less than this (about 0.003")at the most common failure points (between the cylinders).

Given also that the common gasket thickness once compressed is about 0.040" the deformation gradient across the gasket at the most common failure point is less than 1:10, I believe that the movement can be completely absorbed by deflection of the gasket with little or no requirement for slippage, Indeed any slippage would result in "Fretting" of the surface against which the slipping takes place, this kind of Fretting does take place on the Cummins BIG CAM Engines but as both head and block are iron it's not due to differential thermal expansion.

Ahead of any questions on this, in doing these calculations I used published figures for the thermal expansion coefficients of Ally and Cast iron, knowing that the actual rate of expansion for the types of cast Ally used in cylinder heads is smaller than general purpose Ally as a result of the effect of differing grain structures, of course all of these factors are documented and understood by those in the business of designing engines, it should also be noted that Cast iron is quite flexible at the stress and strain figures that we are discussing here, it would also be likely to move sufficiently to absorb a great deal of the expansion, only a full FEA of the design in question would give "watertight" results.

My point is; Keep the head and block surfaces flat and free from deep grooves and you'll probably be fine, if you match your gasket material accurately to it's ideal finish you will likely have even more success, and given the advances in gasket technology things are only getting better, considering that most of us work on older engines as a hobby rather than cutting edge stuff for Supercar manufacturers,

I most often use a fly cutter in a Horizontal Mill, with the head sat up on either the inlet or exhaust port face (Whichever is square to the head face) as the Horizontal mill is more rigid than a turret and it's really easy to clamp and set up the face accurately, I use two stepped setup widgets to set the head up just hanging over the edge of the table and parrallel within a few thou, I bolt the widgets on using bolts through the head stud holes, then slide the head back until both widgets touch the table and clamp the head down, after tightening I run a dial across and along the old face and then skim a new surface.

Regards
Rick