Engine guts 12: Setting deck height

(In case anyone wondered why there was a lapse in updates, it’s because of this little guy who arrived on January 28. It’s gotten a bit harder to find time to work on the engine since then, although it’s so close to the end I’m going to try hard to get it done.)

Axel, the source of engine work delays…

Now that the case is closed, the next step is to set the deck heights. On a VW, the cylinder barrels are separate from the crankcase, and the combination of crankshaft stroke, connecting rod length, piston height, and compression ratio affects how close to the top of the cylinders the pistons come at top dead center. This is called the “deck height”, and must be set correctly since it affects the compression ratio of the cylinders.

The deck height can be adjusted in two ways. First, by adding shims under the base of the cylinder barrel. More shims here puts the cylinder further out from the case and hence increases the space between the top of the piston and the top of the barrel. Second, if the desired change is small, the pistons themselves can have their top surfaces fly-cut down a bit. You obviously don’t want to take away a lot of material this way, but since shims with thicknesses less than a millimeter aren’t common this is a way to get exactly the height you need.

There is another reason you might want to fly-cut the pistons rather than add shims: You can set each cylinder’s deck height individually this way. Since the heads go over both barrels on each side, the two cylinders need to have the same shims or their top surfaces, which seal against the head, won’t be co-planar. Now, ideally, the combustion chambers will have exactly the same volume and hence you want exactly the same deck heights, but in the real world that’s not the case. On our engine, the combustion chamber volumes are the same to within a cc. To get the same compression ratio on the different cylinders, this translates to a different deck height of about 0.1mm. Not much, but if you’re going to fly-cut them anyway, you might as well do it right.

There is another consideration with deck height: you don’t want it to too large. The combustion chamber only takes up a part of the cylinder cross-sectional area (see the pictures of the heads in the linked post above.) The rest, known as the “squish” or “quench” area, becomes very thin when the piston is at the top. If the piston came up exactly to the top of the cylinder, there would be no volume here at all and all the fuel/air mixture would be squished out (hence the name.) This is desirable for several reasons. First, it concentrates all the fuel/air mixture closer to the spark plugs and hence helps it burn more quickly. Having combustible mixture hang out far away from the spark plugs increases the risk of it spontaneously igniting before the flame front has time to reach it, so that makes the cylinder more detonation-prone, which is a bad thing. Second, as the mixture is squished out from the quench area into the combustion chamber, it creates turbulence which helps the combustion process. For all these reasons, you want to keep the deck height small, ideally no more than 2mm between the piston and the deck surface in the head.

All this means you really want a zero deck height. However, for practical reasons that’s not possible. You really don’t want the piston to hit the head and some clearance is needed to allow for thermal expansion, bearing wear, and the fact that the piston can cock itself slightly in the bore. For these reasons, people don’t recommend setting a deck height less than 1mm in the VW. So for an efficient, reliable engine, you’re constrained from both ends here.

With our engine, there is another concern. As shown in the combustion chamber post earlier, the spark plugs for the secondary ignition protrude above the deck surface in the head, with significant variation between the cylinders. When I test-fit the heads and turned the engine over, I could feel a slight interference when piston #4 reached top dead center. After removing the heads, a slight mark on the piston surface was apparent, corresponding to the location of the spark plug. Apparently the piston came too close to the head in this cylinder (which is the one where the secondary spark plug protruded the highest.)

The mark in the center of the picture coincides with the location of the secondary spark plug in cylinder #4. Of all secondary spark plugs, this one protrudes the highest above the head. Apparently, the deck height used previously was too small, letting the piston hit the spark plug.

This wasn’t the correct deck height, I was just using a 1mm shim on all cylinders as a baseline, so it wasn’t an immediate concern. I was surprised, though, because by my estimation there should be clearance to all the spark plugs. Intrigued, I dug out the old #4 piston and, voila, there’s a clear impact mark in the same location! Apparently there has been insufficient clearance here all along.

By sticking some clay on the piston, mounting the head, and turning the engine over, we get an imprint of how close various parts of the head gets. This picture confirms that the electrode for the spark plug has pushed all the clay out of the way and is contacting the piston. It also shows that the edge of the spark plug, which comes in at a shallow angle, also is very close to the piston.

I double-checked this by doing the classic “clay test”. By affixing a piece of clay to the piston tops and turning the engine over, you get an imprint of the head shape. By doing this on the four cylinders, the interference on the #4 piston was confirmed. It also showed that the spark plugs on the other cylinders were dangerously close, certainly violating the desired 1mm clearance.

This presented a dilemma: The 1mm desired clearance obviously applies to the highest point in the head. This normally is the head deck surface, but in our engine it’s the secondary spark plugs. If the deck height is set to clear the spark plug by 1mm, the result would be a too wide squish clearance, maybe 2.5mm, and too low compression ratio.

Since this engine will run 100 octane avgas, there should be plenty of margin to set a compression ratio a bit higher than the 8.0:1 recommended by Sonex for 93 octane and up. Based on what people have run in other aircraft VW conversions using 100LL, I decided to target 9.0:1, which results in a squish clearance of about 1.6mm. This is also more in line with what people recommend for our class of camshaft.

Given that we will also be using a 1.0mm copper head gasket, which adds to the combustion chamber volume, the desired deck heights come out as follows: (The measured chamber volumes include a 1.6mm head gasket, so the 0.6mm difference needs to be taken into account if anyone wants to do these numbers themselves.)

  • Cylinder 1: Combustion chamber 66.4cc, desired deck height 0.70mm, measured 0.33mm, remove 0.37mm from piston top.
  • Cylinder 2: CC 66.9cc, desired deck height 0.62mm, measured 0.20mm, remove 0.42mm.
  • Cylinder 3: CC 66.9cc, desired deck height 0.62mm, measured 0.32mm, remove 0.30mm.
  • Cylinder 4: CC 67.3cc, desired deck height 0.56mm, measured 0.14mm, remove 0.42mm.

The obvious solution to the clearance problem is to make a little depression in the pistons. It only has to be 0.8mm deep to give us the desired clearance. The big question obviously is: would removing 0.8mm material (in addition to the fly-cutting, which will remove up to 0.4mm depending on the cylinder) weaken the piston too much. There’s no easy way to answer this, but asking a few VW people their guess was that it would be OK. The area is close to the edge of the piston, and the piston crown gets progressively thicker the closer to the edge you get. The stresses also go down closer to the edge. It appears the piston is about 7mm thick in this area, so 0.8mm is not negligible but it’s only over a 15mm wide area. We’re also not pushing these pistons that hard, so there should be some margin.

Having decided on this course of action, I programmed the operation up in Fusion 360, figured out how to mount the pistons on the mill, and, after test-cutting on the old pistons a bunch of times, went ahead and modified them.

Modifying the new pistons on the mill. After fly-cutting the tops down the appropriate amount, a shallow depression was cut to improve clearance to the secondary spark plug.

Much more time was spent deciding on a course of action, doing the CAD/CAM, and testing than the actual cutting time. The job only took about 10 minutes per piston, and half of that was in tool change time. I made a movie of the entire job in case you want to vicariously follow along:

So that’s one more thing to check off the list. We should now have correct deck heights and compression ratios, without any risk of interference. In principle it’s now possible to bolt the cylinders and heads on, but I’ll hold off just a bit on doing that because first I want to verify rocker arm geometry and it’s nice to be able to test-fit the heads for that purpose. But it’s getting close!


Engine guts 11: Closing the case

It’s finally time to close up the case. But first a little tidying up.

There’s been a small amount of binding when turning the crankshaft with the Force One bearing mounted. Over perhaps 120 degrees, it was harder to turn than seemed normal. I had measured earlier that the prop hub had a runout of about 0.002″/0.05mm. If the bearing was mounted slightly off-center, it would rub against the journal when the two pointed against each other.

I spoke to Great Plains, who said that most of the time when there is binding in the bearing, it’s because the hole dowel pin that hold the bearing in place isn’t drilled deep enough and it’s squeezing the bearing. I verified that was not the case since removing the dowel pin completely didn’t change things. I also determined that it was not the bearing that was uneven, since rotating the bearing in the case did not rotate the point at which the crankshaft was binding. The only conclusion was that the shop that bored the case to mount the bearing did not get it perfectly centered.

Since the effect must be very small, I decided to try an experiment. I added some lapping compound (30 micron) to the bearing (which is aluminum), bolted everything together, and turned the crankshaft for a few minutes. This got rid of the binding. After taking everything apart and cleaning up the bearing, it was clear that some material had been taken off at a section. I continued with 12 and 5 micron compound to smooth the surface out. After this exercise, the bearing now spins without any noticeable binding.

This is the side of the Force One bearing that was abraded the most by the lapping compound, as evidenced by the matte area.

The other side of the bearing. This one still got cut a little bit, but is mostly untouched. By the pattern it looks like the bearing is actually mounted slightly cocked in the case.

Unfortunately I don’t have the equipment to measure the clearance between the bearing and the journal. It would be good to know whether this operation cut too much material so the bearing now has too much clearance. The guys at Great Plains couldn’t give me a number on what the expected clearance is, though.

Another little thing that had to be done was to notch the cam bearing by the flywheel for the oil return passage. Another example of the half-assed fit of these parts.

The cam bearing on the flywheel end of the case was a tad bit too wide and partially blocked the oil return passage, so it had to be ground back a bit.

Finally, it was time to close everything up. For sealing the case halves, people highly recommended Hylomar, a non-setting gasketing compound that supposedly works very well in situations where it’s important that there be absolutely no buildup, like the case halves that are already a very close fit. It’s a translucent blue, very thin, material that smears out to practically nothing. I applied it all over one of the halves.

Hylomar sealant applied, in a very thin layer, to the case halves without studs, where it’s easier to get it around all the bolt holes.

Checking once again that everything was ready: all the bearings seated correctly, all 8 lifters greased and in place, cam plug in place, cam gear meshed correctly, rubber seals in place on all 8 main bearing studs, cam gear bolts torqued and loctited. I attempted to get the case halves to mate. I’ve done this for fitting purposes probably 25 times by now and never had a problem, but the one time I’d applied sealant would of course be the time where they did not go together properly.

Everything’s ready to go together.

Taking them apart again, I noted a cam bearing had slipped out of place slightly. Luckily the case flange surfaces never actually mated so the sealant was undisturbed, and the second time everything went together OK.

To avoid oil leaks, I also applied Hylomar to the washers for the main bearing studs and cam studs, and then on the underside of the nuts. This was a bit messy even though I attempted to put on as little as possible. The M12 main bearing stud nuts were secured with Loctite red. (They originally had Nylon locknuts on them but I was unable to find M12 fine thread lock nuts in stainless, and given how corroded the original nuts were I didn’t want to put back steel nuts.)

Gradually going around and tightening the big studs and then all the ones around the circumference of the case, you could tell that the Hylomar was gradually getting squeezed out from between the case halves as you came back a few minutes after tightening them and they would then tighten a bit more. I really hope we don’t get any oil leaks here.

The case halves are finally bolted together again.

Once the halves were together, the next step was to mount the main seal and flywheel. The first thing I did was to double-check the crankshaft axial play which, if you remember, was what started this whole ordeal back in October.

Checking crankshaft end play. With the test indicator mounted on the flywheel, pushing the crank back and forth shows the endplay directly. It’s a bit out of focus, but the test indicator was zeroed at 40 and read 30, so the play is 0.10mm/0.004″. The spec is 0.07 – 0.15mm, so we’re right in the middle.

To test this you have to add the shims to the crankshaft and then mount the flywheel. The nut on the flywheel is supposed to be tightened to 250 lb-ft, a crazy amount of torque, but for testing purposes I figured it was good enough to just tighten it real good. It came out to 0.10mm, just like when I had gotten the shims weeks ago. All good there.

Before you mount the flywheel, though, the main seal needs to be pressed in place. It’s about 90mm is diameter, so I was looking for something that would fit over the seal with a suitable hole in the center, so I could use the flywheel nut to pull the seal into position. (I have not had good experience using any kind of tapping on these seals, they always end up leaking.) Turns out the crush plate that came with the Force One hub worked quite well.

The improvised seal pulling tool, the Force One prop plate being pulled in with the gigantic flywheel nut and washer.

I had applied Hylomar to the perimeter of the seal, thinking this would help avoid leaks there. Alas, this was not a good idea. After pressing the seal in place I watched it slowly creep back out until it popped off. Clearly something with more friction was called for here. I had read somewhere that Loctite retaining compound was a good choice, so I decided to try that instead.

Amazingly, the thin layer of Hylomar reduced the friction between the seal and the case enough that it slowly creeped back out.

Cleaning off the Hylomar was not that easy, it’s pretty solvent-resistant. After trying acetone and ethanol without much luck, I finally pulled out the Xylene. That did the trick.

Repeating the operation, now with the retaining compound rather than the Hylomar, everything went smoothly and the seal stayed in place.

After using Loctite retaining compound instead, the seal stayed in place.

Finally, the flywheel could go on. To avoid oil leaks through the dowel pins, I applied Hylomar to the gasket that goes between the crankshaft and the flywheel. The Aerovee instructions call for silicone RTV here, but others discourage that because it builds up too thick and leads to excessive axial play.

The flywheel gasket got a thin coat of Hylomar on both sides, making sure to go on both sides of all the dowel pins.

It was finally time to tighten the gland nut. To be able to hold the crankshaft in position, I had drilled holes in a piece of wood that I bolted to the prop hub. Getting help from Kathy, who held that end of the engine down, I pulled on my largest torque wrench until the piece of wood cracked.

Digging out a long 2×4 from under the house, I quickly (the Loctite on the nut was setting up) drilled a 2″ hole in the center so I could squeeze it between the prop hub and crush plate. With 3 feet on either side, there was plenty of lever arm for Kathy to hold the engine down while resting the other end on the bench so it wouldn’t turn. I basically put all my weight on the 250 ft-lb torque wrench before we got it to click. What an ordeal.

My human counterweight holding the prop end of the crankshaft while I tighten the nut with the big-ass torque wrench to 250 ft-lb.

So there we go. The case is together. Now we can proceed to pistons, cylinders, and the heads. The next operation will be to set the deck heights.


Engine guts 10: Blueprinting the oil pump

After dialing in the cam, I moved on to the final (I hope) grinding work that needed to be done before the case could be closed up: blueprinting the oil pump.

I had to get a new oil pump because the old one had the external connections for the oil cooler, which we will no longer use with the new filter setup we designed. There are a few small modifications that can be made to the oil pumps to increase their efficiency. (These are not new, they’re outlined in the book “Hot to hotrod VW engines” from 1970. Apparently the design of VW oil pumps have stayed the same for the last 50 years…)

The first step is to verify that the inlets and outlets of the pump actually line up with the corresponding passages in the case. VW Type 1 oil pumps are cylindrical and pressed into the front of the case. (The suction passage from the sump and the outlet going to the oil galleries are holes in the cylindrical surface.)

The first step of blueprinting the pump was to literally mark the passages in the case with dye and verify that they lined up with the passage in the pump. The inlet was pretty good, the outlet was not. Note how the blue-free region extends above and to the left of the hole in the pump.

As you can see in the picture above, the outlet in the pump was significantly offset from the case. This was corrected by upsizing the hole in the pump with the Dremel. There was probably also a corresponding place where the case should be opened up to match, but I was not willing to go that far — I’ve already cleaned the case out from swarf and plugged the passages up.

The next thing to do is to smooth the inlets and outlets on the inside of the pump. This is especially important on the inlet side, since pumps are sensitive to flow restrictions there. If there is too much pressure drop, the oil pump can even cavitate, which really ruins your oil pressure. The oil pump is a traditional gear pump, where the oil is transported in between the teeth of the gears. The inlet in the VW pump is mostly aimed at the driven gear, which makes it harder for the flow to go to the idler gear. The solution is to radius the inlet and open it up as much as possible towards both gears.

The pump inlet opening has been radiused to open more gradually, especially towards the front of the pump (although that’s hard to see in the picture) and towards the bottom gear. I could probably have been even more aggressive here, but this is a lot better than the sharp edges of the stock inlet.

The pump outlet shape isn’t as critical since this is a high-pressure area, but it has also been smoothed to avoid sharp edges. The pump also has this vertical outlet hole facing the front of the pump. This is used for attaching full-flow filter fittings, but has no function here. I did my best to round out those edges, too.

The final thing to check is the clearance between the pump gears and the lid at the front of the pump. Excessive clearance here gives the oil a path to leak back from the outlet to the inlet side and lowers the efficiency of the pump. This is especially noticeable when the oil is hot and less viscous. You want as little clearance as possible when cold here, since the aluminum pump body will expand more than the steel gears as the pump heats up. This clearance can be shrunk (but not increased) by carefully sanding the pump body down with sand paper against a flat plate.

The stock pump had about 0.20mm clearance, and the spec is <0.10mm. After some judicious sanding, I was down to between 0.04-0.05mm, as measured with a feeler gauge between the gears and the lid. That should be good.

This is the oil pump being mounted in the case, a small jump forward in time to after the case halves had been mated. (That will be in the next post.) The pump gears are coated with white lithium grease to help the pump “prime” when first started. Since there’s no oil there, the pump may not be able to suck the oil up from the sump without the sealing aid of the grease. The thin, blue coating on the lid surface is the “Hylomar” sealant.

The pump is tapped into the case after the case halves have been mated (which has happened and will be in the next post, when I get a chance to write it up.) The gears are coated with grease and mounted (the driven gear extends through the pump to the cam shaft, where it hooks into a slot) and finally a thin layer of sealant is smeared on the outer perimeter of the pump and the lid mounted. That’s all now done, so one more thing to check off the list.

Next will be actually closing up the case halves.


Engine guts 9: Dialing the cam

One of the few remaining things to check was to “dial the cam”, i.e., to verify that the camshaft and crankshaft are timed correctly. In its most basic form, this is done by ensuring that the marked teeth on the gears on the crank and cam shafts are aligned, and if you were assembling a stock VW made by the factory, that might be the end of it. However, VW’s these days are collections of aftermarket parts and stackup of tolerances can lead to significant discrepancies between the actual valve timing and the design. The only way to be sure is to actually measure the lifter motion and see if it agrees with what it’s supposed to be.

This is a somewhat involved operation because you have to assemble the engine with lifters, camshaft, connecting rod, and pistons. (Technically you only need the #1 piston and lifters since if it’s right on one cylinder it’s going to be right on the others, too.) A way of reading crankshaft position is needed, so you mount a degree wheel on the crankshaft. Then you need to find the top dead center (TDC) of the piston with a dial indicator and align the degree wheel so it reads crank degrees.

Once you’re at this point, you can hook a dial indicator to the lifters and measure where the valves open and close and compare to when they are supposed to according to the cam data sheet. Bob Hoover’s blog has a pretty detailed description of this whole process, with much more detail than I’ll go into here, in case you are interested.

The engine set up for verifying the cam timing. The degree wheel attached to the prop hub indicates the crankshaft angle, in this case 28 degrees ATDC (After Tod Dead Center). The dial indicator is showing the position of the lifter, see the next picture.

The cam in our engine is an Eagle Performance 2234, and I found the card specifying valve open/close times in the Sonex files. However, it was kind of hard to read the writing on the card, and it did not actually specify at what lift open/close was defined at. Since the valve isn’t really either open or closed, it’s just lifted some specific amount, you need to define the amount of lift you are referencing the opening/closing to. A quick email to CB Performance gave me the answer: They use 0.015″ lift, and the spec is: “Intake Opens 33 – Intake Closes 61 – Exhaust Opens 67 – Exhaust Closes 27”. (Those all have implicit BTDC/ATDC, etc, depending on the event. See the Hoover article for the details.)

Now I knew what lift I was looking for. However, hooking up the dial indicator to the lifter turned out to be a bit complicated. The lifters are way down inside the case and the dial indicator can’t reach them. I first tried to just stack a socket on top of the lifter, which kind of worked but was very unstable. I needed something that could sit in the lifter cup and have another cup on the opposite end that the dial indicator could sit in. A pushrod is the obvious thing that can sit in the lifter, since that’s what it’s made for, but the other end is convex. After rummaging through the engine boxes I found an adjustable pushrod, and I realized I just had to make a little head for the top of it where the dial indicator could sit. Enter the JB Weld.

The adjustable pushrod and the chink of JB Weld that goes over it, with a small divot on the end that the dial indicator can sit in.

That worked like a charm, now I had very stable readings.

The dial indicator attached to the top of the pushrod, which in turn is sitting in the exhaust valve lifter. It’s indicating 0.015″ lift. This is the exhaust closing point. The crankshaft angle is the same as in the first picture, 28 degrees ATDC. Since the cam spec is for exhaust close to be at 27 degrees ATDC, we are within a degree.

The first measurements were intake: 35/58, exhaust: 76/24. Compared to the spec 33/61, 67/27, this indicates that the events happen 2-3 degrees early relative to their specified position. (Because the openings are “angle before” and closings are “angle after“, a larger value on the openings and a smaller value on the closings corresponds to “earlier” in the camshaft rotation.) The only exception was the exhaust opening which is off by 11 degrees. I don’t know what’s going on there, but the other three points are pretty much in agreement. 3 degrees is a pretty large error, if you manage to get a tooth off when meshing the cam gear, I think you get an offset of 6 degrees, so this was half of a tooth.

To correct this, you have to rotate the camshaft relative to its gear. There are various fancy ways of doing this but the one requiring no special parts is to simply file the bolt holes in the cam gear a bit larger. This allows you to rotate the cam a bit when bolting it to the gear, and as long as you don’t make the holes large enough to compromise the seat the bolt is bearing on, it’s fine. In this case I had to elongate the holes just under 1 mm, which I did by wrapping sand paper around a drill bit of suitable diameter and filing them out. It took a while but worked pretty well.

When I bolted the whole thing back together and re-measured, I was pleased to note that the timing was now intake: 61.5/33, exhaust 72/28, within a degree of the spec (except for the exhaust again…) Since it’s hard to reliably read anything less than a half-degree increment from the printed degree wheel, and a degree is not going to have a major impact on performance, I called that good. As for the exhaust, I don’t know. Maybe I’ll measure the other cylinders to see if the lobe on the cam is just funny.

There are now only a few more things that need to get done before the case can be closed, and then we have a bit more work to do with deck heights and rocker arm geometry. I’m seeing the light at the end of the tunnel!


Engine guts 8: Balancing act complete

The balancing of the new connecting rods and the pistons is now done. They were within about 1.5 grams to begin with, so there wasn’t much to do.

This picture shows how material was removed from the top of the small end to balance the small end weights, and from the area around the bolts for the big end.

Neither the rods, nor the pistons, had balancing pads. On the pistons I could see where material had been removed around the skirt for balancing previously, so there I continued grinding in that area. On the connecting rods I took material off the small end and around the bolts.

For posterity, here are the final weights:

  • Rod #444: Small end 204.2g, big end 308.4g.
  • Rod #476: Small end 204.2g, big end 308.2g.
  • Rod #827: Small end 204.3g, big end 308.2g
  • Rod #211: Small end 204.4g, big end 308.1g.

Those measurements don’t add up to the total weight of the rod, by the way, because the small end measurement is with the bolts included, but the big end was obtained by taking the total weight, without bolts, and subtracting the small end. Since all we’re looking for is for the numbers to be equal, it shouldn’t matter.

In the end both big and small ends are within 0.3g, not bad. For the pistons, we have:

  1. 489.0g. Pair with rod #444.
  2. 488.9g. Pair with rod #211.
  3. 489.0g. Pair with rod #827.
  4. 489.0g. Pair with rod #476.

The pistons are within 0.1g, and by pairing the heaviest piston with the rod with the lightest small end, we can further decrease the overall imbalance. This should not contribute to noticeable imbalance at the RPM this engine will see. (Of course, the real contributor to vibration is likely the prop. It would be good to get the entire rotating assembly dynamically balanced on the plane.)

One of the remaining things that will need to be done soon is setting the deck heights to get correct compression ratio. I started looking at how to mount the cylinders so they can be torqued down and the deck height measured and discovered that when you put shims under the cylinders, they don’t sit flat against the case.

The case head studs are threaded into steel inserts that are threaded into the magnesium case. (These are called “case savers”, since they save the case from having the threads strip out.) It turns out that most of these inserts were not threaded in deeply enough to go below the level of the cylinder flange surface in the case. This meant that when you added the shims, they actually rested on the edges of these four inserts rather than on the sealing surface on the case.

This is quite bad because, first, the cylinders were unstable and could rock back and forth since they were not sitting against a flat surface. Second, it also means there was a gap between the shims and the case, which is a recipe for a serious oil leak. It would also mean that if the shims gradually compressed where they were resting on the inserts, the deck height would change. The old shims had clear indentation marks where they had been pressed into the inserts.

Two of the threaded “case saver” inserts. These have been carefully ground down so they now don’t protrude above the mating surface around the perimeter of the cylinder openings.

The solution was pretty simple. I went in with a small sanding wheel and very carefully ground down the steel inserts until they were below the flange plane, taking care to not nick the very narrow mating surface at these points.

Hopefully fixing this will mean less risk of oil leaks at the cylinder bases and a more reliable deck height setting.


Engine guts 7: More oil mods

The second step in the “oil mods“, after boring the case to increase oil supply to the left-side (in the airplane) lifters, is to modify the lifters themselves to increase the oil supply to the head.

As you can see below, the lifters have two circumferential grooves in them. One of the grooves has a hole in it, leading to the pushrod bore inside the lifter. When the lifter is not being actuated, the groove without the hole is aligned with the oil supply in the case. Oil will flow around the groove and on to the next lifter, and also oil the lifter bore.

The second groove, with the hole, is aligned with the oil supply when the valve is open. Oil will then not only flow around the groove but also into the hole and into the pushrod, supplying oil to the rocker arms in the cylinder head.

The problem is that this only happens when the valve is at full lift, which it is for maybe 90 degrees out of the 720 degrees of crankshaft rotation during an engine cycle. This means the head is only getting pressurized oil 1/8 of the time. This is the most important time for it to get oil since this is when the rocker arms are under maximum pressure, but experience has shown that it apparently is marginal and more oil is needed to ensure the rocker arms don’t gall, especially if you use high-pressure valve springs.

Furthermore, if an aggressive cam with lots of lift is used, it may even push the lifter groove past the oil supply hole in the case, in which case the head just gets a little blip of oil as the lifter groove passes the hole. Not good at all.

A lifter modified to increase the oil flow to the head. The hole in the upper groove supplies oil to the pushrod. The mod is to cut three connecting grooves between the two grooves, so that the one supplying oil to the pushrod gets a continuous oil supply even when the valve is not being actuated.

The fix, as shown in the picture, is super simple: You just grind passages connecting the two grooves. Three cuts, arranged roughly symmetrically around the lifter, in recommended. The biggest risk is letting the Dremel jump around on the lifter surface, but I managed to do all 8 lifters without any accidents. It only took maybe 15 minutes in all.

This way oil will flow to the head basically 100% of the time. This is supposed to not only decrease the risk of lubrication problems with the rocker arms, but the increased oil flow through the head also increases the heat transfer into the oil, since the cylinder heads are the hottest part of the engine. This should help cylinder head cooling (while increasing oil temperature, of course.)

The only possible drawback of these oil mods (which incidentally VW themselves incorporated into their next model, the Type IV engine) is that it may pump so much oil into the heads that there isn’t enough in the oil sump. I’ll monitor the oil pressure closely for signs of the pickup tube sucking air when we start running it again.

I also examined the wear on the lifters quite closely. While there the surface is noticeably worn, all 8 surfaces are still clearly convex. There is very minor pitting, as shown in the picture below. I was tempted to replace them since I don’t want to have to pull the engine apart again to change the lifters, but decided against it. I think they’ll be fine.

The lifter lobes have clear signs of pitting, but it is very shallow. The picture makes it look worse than it is. They can barely be felt when dragging a scribe over the surface.

The list of things that need to be done before the case can be put back together for good is getting shorter. Top on the list are balancing the new rods and pistons and verifying the cam timing. That’ll be the topic of the next post.


Engine guts 6: More head work

While working on the “oil mods”, I’ve also been trying to get the combustion chamber volumes more even, as I talked about two posts ago.

As so many other things, this became quite an ordeal. It turns out 2cc is a lot of material to remove. I’ve been grinding and measuring for what seems like days. I quickly realized I needed to start hogging out big chunks of stuff to get anywhere, but it’s hard to judge and you don’t want to make some gouges that are so large they can’t be smoothed out later.

The natural spot to remove material is around the extra spark plug, which now is quite unshrouded on the heads that had material removed from them. It’s a big difference compared to the original state, as you can see in the pictures below. This has lead to more shape differences than I liked but there’s not much to do about that. I’ll let the pictures tell the story.

This is the, basically untouched, largest combustion chamber, cylinder #2, with a volume of 67.5cc. As you can see, the spark plug is set quite deep into the head. The only thing done here was to grind away the threads visible in the aluminum, as well as the 1-2 threads of the Time-sert insert barely poking out of the head on top. This also provides needed piston clearance.

This is cylinder #3, which started out with a volume of 66.4cc. Comparing the location of the spark plug with cylinder #2 above, it’s obvious that it’s located much closer to the piston, which likely is the cause of most of the discrepancy. The cavity in which the spark plug sits has been opened up and rounded out to compensate. The area below the intake valve, on the left, has also been deepened to match that of #2 above. After this work, the volume has been increased to 66.9cc.

Cylinder #1, which initially started out as the smallest chamber at 66.0cc. In addition to having a “high” spark plug position like #3 above, also note that the secondary spark plugs, on top, are not mounted symmetrically. In this cylinder, it is toward the exhaust valve, while in the others above it is toward the intake valve. Because the chamber is asymmetric, this cuts the hole more “normal” to the chamber surface in this cylinder, which likely added to the difference. This chamber was treated in much the same way, except the spark plug cavity was made even wider. Even with this amount of work, the chamber has only been increased to 66.4cc.

At this point the chamber volumes are 66.4, 66.9, 66.9, and 67.3 cc. This is a spread of 0.9cc, which I’m going to call good since I’m really not comfortable removing as much more material from #1 as I already have, which would be the required amount to get #1 into better balance with the others.

I’m not so excited about this amount of discrepancy, and the asymmetry of the spark plug placement bothers me. I know it’s not by much, but it’s gotta contribute to uneven combustion between the two sets of cylinders.

While I had the heads and the Dremel out, I decided to do some work on the cooling passages. It’s extremely important that air can flow through the cooling fins around the exhaust ports. These holes are often blocked by casting flash and need to be cleaned out. I’d done some of this back when I had the heads at home for the spark plug thread repairs, but not as well as it should be.

After finding some pictures online about how the holes should look (and probably also having gotten over my hesitation of cutting into the heads) I went ahead and drilled most of the holes. The long “aircraft drill” used in the previous post came in handy here, since these holes are quite deep.

These holes were opened up drastically using drill bits. The exhaust port is right below them, so cooling is critical here. The small hole on the right goes quite near to the casting for the exhaust valve guide. There was no opening here before, and I probably had to drill through 5mm of material to break through. It’s not a big hole, but it should improve the cooling here a lot over the previous situation with no airflow at all.

Having “opened up” those holes, I got a bit innovative and figured I could make some holes near the exhaust flange, where there shouldn’t be any.

This is the area closest to the exhaust port flange, which did not have any air passages before. It’s not possible to go all the way through the head here, because the head stud on the other side is in the way. It is, however, possible to drill diagonally. (Sorry for the bad picture, it was hard to get sufficient lighting here.)

This view shows the other end of the hole near the exhaust flange, and also makes abundantly clear that drilling straight through the head was not an option. The red RTV line to the right marks where the baffles go, so this hole will exit on the low-pressure side of the baffles, as it should.

Hopefully enlarging these holes will provide some sorely needed improved cooling to the exhaust port area of the heads.

While reading about where there should be air passages in the heads, I also came across an interesting note by Bob Hoover. The area between the cylinders has a fairly large air passage, without fins, seen below.

This rather large hole allows a lot of air to pass through and exit the bottom of the head without doing much cooling.

Apparently on the stock VW heads, there is a blocking plate in this location, forcing the air to spread left and right through the cooling fins on the bottom of the head rather than just go straight through. I have not seen this described anywhere else, it’s certainly not part of the Aerovee instruction. It makes sense, though. It should look something like shown below.

I rough-cut a piece of aluminum plate to the required size to see how it would fit. The fins have a square-shaped depression that indicates that there really should be something here.

By forcing that air to take a more useful route, it will accomplish two things: First, it will do some cooling work. Second, because it will encounter more flow resistance, it’ll increase the overall air pressure above the engine which will help push more air through the smaller passages. (Unless we have a big-ass air leak somewhere else, that is.)

I’ll certainly go ahead and add plates on our heads in this location when replacing the heads. It should be simple to cut squares of aluminum sheet and RTV them in place on the fins.


Engine guts 5: Oil mods

I finally got the “aircraft drill bit” (which is a very long drill bit. Why they are called that, I don’t know) needed for the modifications to the oil delivery in the engine case.

Apparently, type 1 VW engines have marginal oil supply to the rocker arms in the heads. Oil gets there through a circuitous route through the center camshaft bearing, to the lifters, through the pushrods, and into the rocker arms. The first problem is that rather small oil passage through the cam bearing. People have long since figured out (like 50 years ago) that you can double the supply by drilling through the case to the cam bearing on the pulley end, too.

At least most of the time, you can. There’s apparently manufacturing differences in the castings that make it so that if you try this on some engines, you’ll drill out through to the outside of the case instead, basically ruining the case. After measuring a bunch of times, I convinced myself that it would be OK on our engine.

This picture shows the basic idea. The 12″ drill bit is inserted from the flywheel end of the case, which has been drilled through all the lifter holes. The lifters get their oil through the hole in the cam bearing saddle in the center,, which is supplied with oil from the other side of the case. The idea is to extend the holes through the lifter bores to the right, and then extend the other hole in the cam bearing saddle on the right to intersect with it. While the hole in the bearing saddle is quite large, note that it is only supplied with oil through the small, square passage cut in the bearing surface.

There are two intersecting holes you need to drill to complete this operation. Both have the potential to ruin the case. Starting with the long drill bit, you extend the holes through the lifter bores as shown in the picture above. To ensure you don’t drill enough, the required depth of the hole is carefully measured and marked on the drill bit.

As I was drilling this hole, the bit made a sound that sounded like when you break through the surface. Which it wasn’t supposed to do! Envisioning dollar bills flying off, I turned the case over but found no hole. It must just have been porosity in the casting. I bravely continued drilling the few more mm needed, without incident.

Next step was to extend the hole in the cam bearing web. Here you’re really drilling straight towards a pocket in the outside so it’s really important to not go too far. Ideally you’d break through to the first hole before that happens, but just to make sure I marked this drill bit with the required depth, too.

Success! After drilling into the cam bearing web, the tip of the long drill is visible through the hole. Next step is to widen the small, square hole in the bearing surface.

Aiming towards the extension of the holes in the lifter bores, it didn’t take long to indeed break into the other hole. Phew. Now the lifters on this side of the engine has twice as much oil supply as they did. To further improve the oil supply, the square passages in the bearing surfaces were also widened a bit.

This drill operation was the final step I’ve been waiting on before all-out cleaning the oil passages from the swarf left over from fitting the Force One bearing and tapping the case for the oil plugs. Using a spray bottle of mineral spirits, I flushed out all the passages, as well as the cavities on the inside of the case. It was pretty cool, the tub I was using was shimmering with flushed out suspended metal flakes! After blowing out all the passages, it was inspection time.

This is the bore for one of the oil pressure control pistons. Looks clean.

This is the main oil gallery supplying the crankshaft and camshaft bearings. Looks nice and clean.

This is route to the oil cooler. This one had some stuff left in it that needed to be cleaned out.

Overall things looked pretty clean. I found a few passages that had little bits of crud in them that had to be brushed out with a pipe brush, but we should now be good to put in all those plugs again.

With that, the case got some paint touch-up since all the handling since it was painted had chipped paint off exposed parts.

That completes the work on the engine case. Part two of the oil mods is to modify the way in which the lifters supply oil to the pushrods. More on that later.


Engine guts 4: Blueprinting

Things are slowly proceeding on the engine front as I’m meticulously going through some things while waiting for parts for others.

I got the cam gear thrust bearing endplay, mentioned in the last post, set right only to realize that this added to the free play in the gears (they are helical, so end play translates to angular play; or maybe it was just easier to notice once the cam turned more freely) to the point that it was outside of the spec. I’m banking on this being because the aluminum cam gear is worn, so I ordered a new one from CB Performance. The cam gears are supposed to be available in different sizes to account for tolerances in the spacing between the crank and cam shafts, but it appears most aftermarket parts only make the “0” standard size. If a new gear doesn’t get the play within spec, I guess I have to go searching harder.

Next topic on my rather long list was to balance the connecting rods. This is a bit tricky, because you can’t just ensure they have the same mass. Because the big end of the rods is rotating with the crankshaft while the small end reciprocates with the piston, they also have to have the same center of mass (and, I guess, also moment of inertia, if you want to go all out). This is accomplished by weighing the small and big ends separately, which requires a jig to hold the two ends such that you can put one on the scale and it can rotate around the other.

After looking at some of these setups I realized I had some small ball bearings left over when I replaced the bearing in the anemometer on the weather station that could be put to use. I designed a simple setup in Fusion 360 and 3D-printed it in polycarbonate.

This is the 3d-printed connecting rod scale. There are two round bushings, made to fit the small and big ends, with two ball bearings each. The non-weighed end hangs on an extra link to minimize sideloads imparted on the scale.

I initially had tried to weigh the small end by mounting the rod on the crank bearing and letting it rotate around there. This did not work, the oil in the bearings apparently has enough viscosity to affect the weight enough to make the measurement useless. A reasonable standard to set is to make sure the rods are balanced to within 0.1g, and the measurement error was several grams.

The new setup worked much better. There is still enough variance that you have to make a bunch of repeated measurements. I think this is because the scale is sensitive to side load and there’s enough flexure in the plastic setup to make it a bit non-repeatable. The scale, with a resolution of 0.01g, also turns out to not be repeatable to better than about 0.1g when you start repeating measurements, even without the side load issue.

The maximum difference between the rods was about 2g (out of ~540g). That’s not bad, but we can do better. With some careful grinding on (hopefully) non-critical parts of the rods, and many measurements later, both the small end and overall weight spreads were reduced to +-0.15g at which point I called it a day.

However, it turns out that was for nothing. I did measure the big end bearing clearance on one of the rods using Plastigage (I haven’t used my set in years and it turns out it was all broken and missing, but I got one last measurement out of it) but I had not looked at the small end. The “How to rebuild your air-cooled VW” book says that the wrist pin should be a light push-fit in the small end bushing, without any perceptible play. Turns out my rods failed that test. Given how much play the old crank bearings had, it seems reasonable that the rods saw more side loads and coned out the small end bushings.

In the “good old days”, this would be fixed by pressing in new bushings and reaming them to size for the wrist pins. This is one of these things that are no longer economical to do when a new set of rods cost $150, you don’t get much shop time for that. So I ordered a new set of SCAT rods that are supposedly already balanced to within 1g. We’ll see, but hopefully I won’t have to do a lot of grinding there.

The other things that was supposed to get balanced was the pistons. The spread there was about the same, a couple of grams. However, in this case I started out measuring the ring groove clearances and discovered that at least a couple had bent or damaged ring lands, so we’ll be getting a new set of those too.

The old pistons were “AA” brand, so I ordered a new set of those. Turns out they make both a forged and a cast model, but the good people at aircooled.net claim the cast pistons are fine for engines that stay under 6500 RPM, which we do by a large margin. Since the forged pistons are about 3 times as expensive as the cast ones, I think cast will be fine.

Moving on to the next item on the list: checking the combustion chamber volumes. If the combustion chambers aren’t the same volume, the different cylinders will run at different compression ratios and generate different power, contributing to uneven and inefficient running. Measuring the chamber volume is easy in principle. Cap the chamber with a piece of transparent plastic and fill it with water until it’s full. In practice it’s a bit more tricky.

Measuring the volume of the combustion chamber requires a bore-sized piece of plastic and a way of measuring volume. This chamber is almost filled with soapy water

Normally when doing this on a VW, you plop the plastic directly on top of the head. This allows you to just measure the volume of the chamber without the squish areas where the full bore is exposed. This volume is easily added later when you measure the deck heights, ie the minimum distance between the piston and the top of the cylinder. On the Aerovee heads, however, the second spark plug protrudes into the squish area, so it’s not possible to put the plastic directly against the top of the head. I had to use the 1.5mm copper gasket ring as a spacer. This is fine, I just have to take that into account when setting the compression ratio later.

There were two things making this procedure trickier than I hoped. First, I discovered that using soapy water worked much better than clean water, because the surface tension of the water otherwise prevented it from wetting the spark plugs and other small cavities. Second, It’s quite tricky to get all the bubbles out. In particular, air tends to get trapped in the narrow squish area and it’s hard to get it to come out to the hole in the middle. You also can’t tilt the head indiscriminately since then water will seep out around the perimeter. To get as good of a seal as possible, I coated the edge of the plastic with grease and set one of the cylinders on top of it to ensure it was pressed against the gasket.

I also don’t have a good way of measuring volume with high accuracy. Instead, I used the same precision scale and weighed how much water I had to take out of a cup and squirt into the head. This is a bit more error-prone since you have to ensure you don’t lose any of the water anywhere else. In practice, I repeated the measurements three times and got a spread of about 0.2cc, well within the accuracy needed. The chambers are about 68cc and with a swept volume of 545cc per cylinder, being off by 1cc results in a difference in compression ratio by about 0.12. It should not be a big problem to reduce the existing spread of about 2cc to less than 0.5cc.

Once the volumes have been measured, you of course have to do something about it. The obvious solution is to grind away material from the chambers with smaller volumes until they are equal to the largest one. But where do you grind? It appears with these heads, the differences are mostly from how deep the seat for the extra spark plug was cut, so it seems reasonable to take out material around that area. This should have a minimal effect on the airflow through the chamber, since you don’t want to end up making that uneven between the different cylinders either.

I’m currently in the process of modifying the chambers, so I’ll report back when this is complete.




Engine guts 3

I finally got the engine case back from the machine shop, so I could proceed with remaining work.

First I finished up drilling out all the oil passage plugs from the case that my Dad and I started before giving the case to the shop. To be 100% sure that you get all gunk and swarf out of the oil passages, you need to remove all the plugs from the case and instead tap them with assorted NPT pipe threads so they can be replaced with screw plugs at assembly.

This task was straightforward except that the largest passages use 14mm plugs that need to be drilled out and tapped for 3/8″ NPT thread. The drill for such a tap is just over 14mm, and the only place I can mount such a drill is in the drill chuck for the CNC mill. The CNC mill, however, doesn’t have enough space to fit the engine case, so that was not a solution. I ended up using the Dremel to get rid of most of the material from the inside of the existing hole (always fun) and then turning the drill bit by hand to let it cut the hole to size. It took about 20 times longer than if I had been able to just drill it, and the holes did not come out perfectly perpendicular. That is luckily not a concern since all you need is for the tap to cut a usable thread, which it did.

The Dremel added to the significant amount of swarf still left in the case from the machine shop cutting it for the prop hub bearing, but I only cleaned it out superficially since I still have one more drilling operation to complete. I need to drill out an oil passage to improve oil flow to the heads, but that task awaits the arrival of a 12″ long drill bit.

Instead, it was time to test fit the crankshaft and make sure the machine shop hadn’t screwed up the job. The first order of business was to make sure everything was in-line and the crankshaft wouldn’t bind in the bearings. Before fitting the crankshaft, I started by bolting together the case halves and measured the bearing bores with my just-acquired cheap set of bore gauges. These require a bit of skill to use accurately, but in combination with the Mitutoyo digital caliper the bores generally came out to 50.03mm, plus minus a few hundreds of a mm. That’s in the right ballpark, since the service limit turns out to be 50.03mm, but the as-new dimension is 50.00-50.02mm so I don’t really have the required amount of precision here. But at least there was no sign that the bores are grossly oval or oversized, so that’s good.

When trial fitting the crank, you replace the locating dowel pins, mount the #2 bearing half (the other bearings are full-rounds so are slid onto the crankshaft) and then lower the crank in place, taking care that all the bearings find their dowel pins correctly. At that point, the crank should spin freely. Which it did, kinda. Not as freely as I would have hoped, but only finger force was needed.

The real test is to do this with the case halves bolted together, though, because the stud tension actually deforms the case. Boring is always done with the case bolted together to the correct torque, so in principle the bearing bores aren’t circular until things are tightened down. This actually turned out to be the case now; as the studs were tightened the crank spun progressively more freely. Kind of cool. This included the new Force One bearing, so apparently the machine shop did their job.

The engine case with the new Force One bearing up front, bolted up and with crank and cam shafts test fit.

The next step was to add the camshaft. Here things looked a bit more iffy. First, the gear driving the camshaft has a lot of scoring on the teeth. I assume this also is from all the particles embedded in the oil.

The gear driving the camshaft has some knarly looking wear and scoring. However, it does not have excessive backlash so it seems OK.

The second problem was that when I added the camshaft in its bearings and tried to spin it, it would bind at a few places. Turns out that a few teeth had gotten dented on the edge. I suspect the shop did this when they pressed the gears off the old crankshaft. To get things to rotate smoothly, I had to dress the affected teeth with a file. No big deal.

A few of the teeth on the cam driving gear were banged up such that it would bind against the cam gear. Minor dressing with a file got the raised parts back in line so the gears meshed cleanly again.

Another little thing I was alerted to by a youtube video was the fact that the thrust bearing’s oil supply hole is not very well aligned with the oil passage in the back of the bearing, such that about half the oil supply hole is blocked. The solution to this is to simply grind away the back of the bearing to open up the hole.

As is visible on the thrust bearing journal, the supply hole is offset from the circular groove on the back of the thrust bearing. By grinding away the back of the bearing where the hole enters, the obstruction is minimized.

There are a few more things to sort out with these bearings. There is still a little binding when rotating the crankshaft/camshaft in the assembled case. It’s possible there’s another buggered tooth on the gear that I didn’t notice before. It’s also possible that this is because the thrust bearing on the camshaft has no end play. This is apparently a common problem with the bearings you get these days; they’re all too wide such that they actually stick on the camshaft. The solution is to carefully sand down the thrust surfaces of the bearing halves until the requisite 0.04-0.10mm appears. If you don’t do this, the bearing will not get sufficient lubrication.

Once these issues are worked out, it’s time for the rod bearings. Hopefully before the weekend.