VW cylinder head machining - increasing compression ratio

Well this guy came to the shop with two VW cylinder heads that he wanted some machining done to. The idea was to remove 2 mm from the combustion chambers in order to raise the compression ratio. Not being a specialized VW shop, or a car engine machine shop for that matter, here's one way to do it:

First, find a a big faceplate for the lathe. Somewhere there's bound to be an old one with a roughly 4" by 4 tpi mount, so make a threaded stub shaft that fits, generously loctite the threads and assemble with a cheater bar and BIG hammer. Machine a slug with a centre hole. Make it so it just fits the cylinder seats, toss the faceplate-on-a-shaft in the three-jaw, pinch the cylinder head against the faceplate with a live centre pressing against the slug inserted in the cylinder seat, bolt it onto the faceplate with suitable fasteners and it should look like this:

Big faceplate with VW cylinder head - iPhone 5s

Remove the tailstock and slug, indicate the cylinder seat and tap it around a bit if you need to, tighten the bolts, mount some scraps to balance the faceplate, and bore 2mm deep. Paint the first bore with a magic marker so you can touch off and set your dials for diameter and depth, then bore to size and depth on all four.

Of course you need to use the slug and centre on all four holes while setting up. It's so much easier when you don't have to fight gravity. Depending on how flat the faceplate is you may also have to use shims. I did not want to add that fuss so I took a skim cut after assembling the stud shaft and faceplate. Next time around though, I might have to shim since I don't want to turn the faceplate into chips. It's only so thick to begin with.

Well the job was a success and I made myself a faceplate that can go in both of the big lathes without removing their chucks. Fun stuff!

Cylinder head resurfacing

My take on cylinder head resurfacing in the mill is to use a single point tool - a flycutter. Multitooth cutters can speed things up, but if one cutter edge gets chipped you´re in trouble. Flycutters take longer time to use since only one cutting edge is doing all the work, but if you have a mishap you only waste one insert. And swapping from cast iron inserts to aluminium inserts is quick, you change the one insert and go to town.

Now, I seldom do heads more than a few times a month. If all I did was cylinder heads you can be sure I had a big multitooth cutter or even a cylinder head resurfacer, but since I don´t I haven´t. Thats why I use what I already have - a vertical mill and a flycutter.

If the head is warped, which it probably is if you intend to resurface it, you have (at least) three options; pull it flat against the table, shim the blocks to fit the twisted head or straighten it before resurfacing. The easiest of course is to just tighten it down and start cutting. If there is only slight twist this may be the best approach. Shimming might be necessary if the head is too warped to pull flat. But if the head is so bowed or twisted that the cam bearings are out of line, maybe even causing the cam(s) to bind, it really should be straightened first. Or junked.

Straightening typically involves twisting or bowing the head the opposite way it's already warped, and then baking it in heat to let it "take a set" and end up straight. The bearing bores, that is.
While this sounds easy in theory it may be tricky in reality. After the head is straight 'nuff, the gasket surface can be resurfaced.

If the cam bearings are out of line, perhaps you could also resurface the head as is and then grind some off the bearing caps making the bore smaller and then hone them out again. Obviously this only works if there are removeable bearing caps. The end goal is to have the cam(s) sit straight and parallel to the gasket surface.

For a typical 4-6 cylinder head, I usually place the head on ground blocks, tighten it down just a little and feel if any blocks are loose - you want the head bearing on all corners. Watch out for burrs, dents, grit and gasket remains. After any needed shimming I continue to gradually tighten the hold downs until its firm and secure. I then touch off, snug the table locks (don´t lock the feed axis) and take a skim cut of 0,001" or so. This lets me see if any corner or side sits high or low. Add shims as necessary to average out the differences. Newsprint works well for small adjustments. When you are satisfied, dial in and take a cut. I typically run 269 rpm, 0,004" depth or so and 6-7 ipm while roughing. Adjust the cutting depth according to how unflat the head is. When it's almost flat I take another 0,0015" or so at 2 ipm to finish it off, hopefully cleaning up all areas.

Rover head starting to clean up - iPhone 4s

Note in the above picture how the o-ring grooves are starting to disappear at the ends, but are still there in the middle - this was a banana for sure, and the grooves seem to be cut crooked aswell.


Now, the above procedure assumes the previous gasket surface is factory original or at least resurfaced parallel with the valve cover surface. This is important as you are averaging the cut over all four corners. If the head is banana shaped you want to hit the ends equally. If it is twisted you want to hit diagonal corners equally. If it is both banana and corkscrew, well.... you get the idea. But if the head has been cut wedge shaped by the guy before you, and you also cut more on one end than the other, you´ll have a head that sits tilted when its done. Perhaps not by much, but still.

Of course you could touch off the skim cut, find the lowest point, dial in one roughing cut to remove all warp and then take one finishing cut. But fiddling around finding the low point may take just as long as just cutting away until things start to look flat. I tend to engage a cut and then do something else while the mill hums along.


How do you measure the removed amount? I like to touch off, set zero, cut until its clean and then see how much was removed from touch-off. You could also mark a spot that you measure with a micrometer before and after cutting and see how much you removed. If you measure the middle of a concave head you could end up with 0,001" although you removed say 0,01" at the ends. I feel this is misleading and thats why I prefer to measure from touch-off. That gets you the max that was removed.
If you don´t have readouts you could set up a dial indicator on a magnetic base as a poor man´s readout.