Head Rebuilding

by Richard Atwell
(c) Copyright 2005-2020


The weakness of the air-cooled motor is the cylinder head. Unlike its water-cooled brethren which have an almost unlimited cooling capacity, there are limits to the amount of heat (related to power) that can be produced by an air-cooled motor before it overheats. There are also emissions concerns which was another reason VW switched to the water-cooled based Audi engine technology from the Porsche air-cooled designs of the 60s and 70s.

Porsches have been producing high horsepower, low emission engines for decades so how did Porsche solve this problem? First, each cylinder had to produce 1/6th the total HP compared to 1/4 on the VW motor. 2nd, the heads use large valves and a hemispherical combustion chamber. Third, Porsche also employed modern 3-way catalytic converters to reduce emissions. If you've priced one of these exhaust systems, they are not inexpensive!

Apart from being an entirely different design, the biggest difference is that each cylinder had its own cylinder head and the cylinders themselves are made from Aluminum. VW cylinders are cast iron based and the reason for this is that aluminum is not suitable for use as a cylinder wall surface.

Because Aluminum is much better at radiating heat (cooling) than cast iron, Porsche and Mahle worked to develop a wall plating that would allow the piston rings to seal against the cylinder wall. This plating is called Nikasil and it's a popular method (even my motorcycle cylinders are plated with this material). Mahle invented this plating in 1967 and you have to wonder where the 911 engines would be without it. Production cost is what makes these cylinder sets so expensive. The average P&C set on the Porsche costs 5-10x the price of a VW set.

Porsche engines producing these HP levels can go 200k miles before a rebuild if taken care of. VW motors last half as long at best even though the working tolerances for operation are much lower. Even with this technological advantage Porsche heads can still overheat when pushed which goes back to my original statement that the weakness of the aircooled motor is the cylinder head.

Of the various repair work that must be done to rebuild a 4-cyl VW boxer engine, THE most important aspect involves the cylinder head. The original heads lasted 100k miles+ but the rebuilds do not despite all their so called upgrades. Why?

Heads are a complicated topic and I've tried to cover as many of the aspects and myths as possible in this article.


There is a specific bolt pattern for each head and the torque requirements are different between T1 and T4:

Type 1:
T1 torque pt1 T2 torque pt2
Type 4:
T4 torque

Part Numbers:

Heads have numbers cast into them that do not match the part numbers in the fiche. This is because these casting numbers refers to the style and the part number for a particular series of engines that used that casting. Head castings were factory modified for smog ports, breather tubes, CHT bosses, etc. The unique part numbers which were printed in blue letters above the intake manifold studs are useful for ordering parts but the casting numbers are more useful for identification.

Part Number Engine Code Serial Numbers
021 101 353B CB 000 001 - 060 640
021 101 351Q CB
062 001 - 110 000
000 001 - 011 000
022 101 351C ED 000 000 - 000 639
022 101 351H ED 000 640 - 025 000
021 101 352 AW 000 001 - 037 000
022 101 351L GD
000 001 - 055 800
000 001 - 039 331
039 101 351K GE 039 332 - 055 786

Note: the 79 bus cylinder head is one year only. Although it shares the same square exhaust ports with the Vanagon CV engine, the valves are not the same size.

Casting part numbers are nicely covered over at Type 4 Secrets Revealed. You might discover that your head came from a 70-76 914, 1976 912E, 69-74 411/412, 72-79 Bus or 80-83 Vanagon.

Head Care:

Unlike a water-cooled car with an engine temperature that is largely a function of ambient temperature, air-cooled engines run at much hotter temps (350F vs. 170F) that depend on rpm. It's normal for the bus head to increase 100F in temperature when going from idle to fast highway speeds (this means the cooling system can't keep up).

Consider that the bus is a big box driving down the highway and it takes a lot of effort to go 60mph because the power requirements increase non-linearly due to the wind resistance. In other words, the engine has to work hard and gets hot as a result. The difference between 60mph and 70mph can mean a further increase in temps by 50F. At some point the head becomes heat stressed. VW dished the bus pistons (15cc on stock Mahle 94mm 2.0L piston) partially to reduce the compression ratio in order to lower the heat stress to the heads (a flat top or domed piston is not a suitable bus piston).

Every water-cooled car has a temperature gauge for the coolant but for some reason VW never installed a temperature gauge into bug, bus or any other model (you may find some mechanical gauges installed on some industrial (stationary) versions of the engine). The first step towards preventing a meltdown is to install a temperature gauge.

Oil temperature is not a good indicator of the engine temperature. A better indicator is the cylinder head temperature. Because the spark plug is conveniently accessible, you can gauge the temperature of the head by using a thermocouple crimped to a ring terminal that sits under the spark plug in place of the spark plug crush washer. Monitoring the temperature of the head is essential to long head life. When the heads overheat you loose compression in the mild case and suffer catastrophic engine damage in the extreme case which will cost thousands of dollar to repair. Since one leads to the other I highly recommend buying a temperature gauge, wiring it up to spark plug #3 (traditionally the hottest cylinder) and keep your head temps (at that location) below 400F as you drive on the highway as best as you can.

My favorite CHT gauge is made by Dakota Digital. It's the only one in this price range that provides accurate digital readouts (in fact it's the only accurate gauge in this price range, including traditional analog models like VDO).

You might wonder why more probes and gauges aren't required. For the paranoid, installing 4 gauges/probes is the only way to feel reassured. Checking only one head gives you a 25% chance of catching a failure before it happens and that's all. However, all other parts being equal, it's often a detuned engine that causes the head to fail rather than a specific part failure. In this case, when one head gets hot, the others are probably hot as well. Of course, if there is a cooling blockage at once cylinder this won't necessarily be the case and so you can why the paranoid are so paranoid. :-)

I'm sure there was a clever engineering solution but this complication may be the reason why VW never installed a gauge in the first place.

Construction and Cooling:

In order for the internal engine parts to cool properly, the oil temperature must be kept low because oil has a finite ability to absorb heat. The head temperature must be kept low because the hotter the parts get the more they expand and the combustion "seals" break and you lose compression. If you see low oil temps, you may see high head temps because they are not directly connected. When your oil breaks down you can replace it. When your head breaks down you have to shell out a lot of money.

A VW cylinder head shares the same design with any other modern aircooled cylinder head design: fins transfer the heat to the cooler air passing over it. As the engine rpm increases so does the cooling fan output. Up to about 65mph the fan is able to cool the head sufficiently assuming the rest of the engine is working properly, but after that speed head temperatures rise dramatically because they cannot shed the heat from combustion fast enough.

When I talk about seals, I mean all of the mating surfaces exposed to combustion gases must be closed tightly. When overheated, the cylinder to head mating surface can distort and introduce a leak which can be hard to detect with the engine installed in the vehicle. This is because of the engine tin that obstructs the cylinders and the cooling air that passes "through" the cylinders both mask the noise. Although problems can be caused by assembly errors, failure is often caused when the engine is pushed beyond its limits for long periods of time.

Sometimes when the leak is large, it can be heard as a squeak. Sometimes you can hear the heads slapping around on an older engine when you start it up which goes away as the engine heats up and expands in width.

Few designs are perfect but there is an important flaw with the original design: there is a thin metal gasket between the head and cylinder and the combustion gasses can burn through and escape even though the head is tightly clamped down with the correct torque. Once the gasket starts to burn, a gap quickly appears because the rest of the gasket keeps the head under torque and fails to heal itself. The result is a growing loss of cylinder compression once the gasket burns through. This is one of the two types of head failure that Type 4 heads suffer especially when installed in hot running bus engines.

The photo below depicts a typical meltdown of a head. The combustion chamber looks burned, the valves have sunk into their seats and the cylinder lost compression at the bottom of the where you can see the fuel/oil/carbon was pushed out.

head gasket failure

The contact surface between the valve and the seat surface is also important. Any place the head is leaking will eventually lead to a failure. If the intake valve isn't closing properly then combustion gasses will escape through the intake and possibly cause damage to the carb or FI (heard as popping which is a type of backfire). More important is that the exhaust valve seals properly because if the exhaust valve overheats, it may falls off or dislodge its seat and physically damage the piston and head. This is an expensive failure and is the 2nd kind of failure common to Type 4 heads.

The exhaust port deserves special mention because it is difficult to seal. A new head and old heat exchanger require that the tops of the exhaust manifolds be perfectly flat in order to properly seal. If the exhaust port does not seal, then the head will overheat from the escaping exhaust gases and this can also cause the exhaust valves to stretch and fail.


Power is analogous to compression: in order to turn the molecular energy of the combustion mixture into mechanical energy in the form of pressure to push the piston down the combustion chamber must make a tight seal. The seal is not 100% perfect because there is some leakage past the piston rings. If the seal was too tight then the friction and wear between the cylinder and piston rings would be too great and this could cause the piston to seize in the cylinder.

As the engine ages, the rings wear and so do the valves. Both of these conditions can be monitored with a leak-down test while the engine is still in the bus. A special gauge called a cylinder differential gauge is used to determine where the air is leaking from. Air is pumped into the combustion chamber through the spark plug hole with both valves closed. If air is heard/felt through the crankcase breather or oil filler tube, the air is leaking past the rings. If the air is coming from the intake port through the carburetor or intake manifold then it's leaking past an intake valve. Leakage from the exhaust indicates a problem with the exhaust valve. If there is a pressure drop and it's not caused by any of the above, then a loose cylinder seal is to blame.

The test is repeated for each cylinder and with a new motor this leakage might be 4% (for example if you pumped in 100 psi of air, 96 psi of pressure would remain in the cylinder). When the leakage is in the 0-20% range it is also important that the leakage between cylinders be limited to a difference of 20%. This limit narrows as the engine performance requirements increase because it's an indication of the state of balance much as a compression test would indicate.

As power is lost, the engine has to work harder to maintain speed and therefore has more difficulty cooling. When you have new parts and have a compression issue investigate it immediately before you experience an unwanted head meltdown.

Valve Seat Issues:

In addition to power generation properties, the geometry and construction of the parts are designed to keep them cool. Heat from the combustion process is constantly being radiated to the metal parts and they will eventually fail from heat fatigue if overheated.

The intake valve because it is constantly cooled by the intake charge of air and fuel is far less susceptible to overheating. The exhaust valve is not so lucky: instead of a cool flow of air over it, hot exhaust gases are continually flowing past. Not obvious to the average home engine builder is that the hot exhaust valve cools itself by making contact with the "cooler" head which wicks away the heat.

The designers came up with additional cooling solutions such as using sodium filled valves. The sodium liquifies at operating temperatures and the liquid is flung from the head to the tip inside the valve stem by the action of the valve to help to cool it. Some of the 1.8L Type 4 VW engines used sodium valves and you can identify them by the dimple in the center of the head and the 0.008" valve adjustment specification for this model instead of the normal 0.006" clearance.

The contact face between the valve and the seat is only about 1.5-2.0mm wide. Because the head is an aluminum design, the valve seats are inserts that are pressed into the heads. The material is much harder than the head and takes the abuse of the valve contacting the seat over and over, especially during startup on engines with solid lifters.

Type 4 heads can suffer from a problem called a dropped valve seat. All VW heads have this problem but Type 4 heads are particularly susceptible because a) most of the production were installed into heavy vehicles like the bus and b) the 914 folks liked to drive their engines hard. The seats in the photo below were loose in their counterbores and only the valve spring was helping the valve to hold them in place.

sunken valve

What often happens is that the valve seat falls out and jams the head of the valve causing it to break off which can jam the piston. At the very least a chunk of metal in the combustion chamber usually results in a hole beat through the piston crown and a beating of the combustions chamber. The seats were originally installed with an interference fit meaning that the OD of the seat was wider than the ID of the pocket in the head that holds the seat in place. When the head overheats, it expands so much that the seat falls out. This happens because the seat is made from a harder material and expands at a much slower rate.

Another problem is a sunken valve. You can see the intake valve in this photo has worked its way into the seat and sunken below the level of the bowl of the combustion chamber. This is another sign of overheating. Also notice that the larger intake valve is the one that has failed even though its cooled by the intake charge. This happened because either the head of the valve distorted or the seat softened from overheating and changed shape.

loose seats

These are the #1 reasons you should be driving with a CHT gauge in your bus. By keeping the heads under 400F you can be assured that the seats are doing well. If the temperature goes to 425-450F the seats aren't going to fall out right away but eventually they will fail as they age because this is the nature of the design. Some rebuilders attempt to remove the old seats by placing the head into a casting oven that can reach 400F and above. Sure enough, some seats just fall out because they were previously overheated.

Now, grill your head rebuilder and ask him a) what kind of valve seat material he uses and b) how much interference fit he uses. If the answer is stock or he says it's a proprietary secret, find someone who will. If you've guessed that most head rebuilders use sub-standard materials and rebuild your head too quickly you've guessed right.

Because these metals expand at different rates, unless the seat is properly fit and made from the correct material, the seat may fail prematurely because the work isn't up to VW factory standards. Techniques have been developed to help make the air-cooled head more durable and like anything else there is a cost to have this kind of work done. The difference between a correctly rebuilt head and a substandard one is always a few hundred dollars extra vs. an engine drop and many more hundreds or thousands of dollars.

Leaded gasoline:

Back in the days of widespread availability of leaded fuels, TEL (tri-ethyl lead, ethyl or simply referred to as lead) lubricated the contact area between the edge of the valve and the seat as well as the valve guide. This allowed heads to be made of low cost grey cast iron along with the valve guides. Hardened seats and bronze guides were aircraft technologies from the 1920's and although american car manufacturers rolled back these improvements in the name of low cost, VW with its aluminum head design did not.

Lead also helped to transfer heat from the valve to the seat and this acted as a self-healing "metal" gasket between the valve and the set. All of these advantages were side effects discovered when lead was added to the fuel formulas as an octane enhancer. Octane was boosted so higher compression ratios could be used to make more power without the engine knocking.

The downside to using lead was that it polluted the environment. When catalytic converters were introduced to reduce emissions, the lead coated the precious metals inside and ruined the cat instantly. Lead had to go.

When leaded gasoline was in the process of being phased out people freaked out thinking their VW lead loving heads were going to fail (in fact every lead-using vehicle owner had the same reaction). Naturally the aftermarket additive producers jumped on this and introduced various lead substitutes but the fact is that auto manufacturers had been aware of this change long before it occurred. Due to the need to be compatible with the varying gasoline formulas around the world VW had anticipated the change and also relied upon the replacement additives added to gasoline by the oil companies. As a result, VW engines continued to run as well as ever. Cast iron heads however, on other makes of cars did not last as long but they didn't fail right away either because of the substitute additives.

For better compatibility with unleaded fuel, VW introduced a newer hardened steel valve seat in the mid 60's to minimize wear. Every baywindow bus has seat material compatible with unleaded fuel (at least the fuel formula of 1978).

So what does all this have to do with VW heads? If your head rebuilder is making claims that he's modified the head to take the unleaded, lower quality fuels of today think about what he is offering:

In other words, the factory built head was perfectly able to run for 100k miles using unleaded fuel so that's not the issue. VWs that used unleaded from the beginning were far better built than any budget head rebuild I know of.

What about the implied knock reduction ability when they use the term today's fuel? A modern car has a knock sensor which can retard the timing if the fuel causes the engine to knock (pre-ignition) but an air-cooled VW cannot. When it is being suggested that the lead acted as a superior knock control that helped us when the engine got hot and was more susceptible to knocking, I simply don't believe it unless you consider that you could buy 100 octane at the pump at one time.

What's more realistic is that as the fuel chemistry has changed and therefore so has the combustion chemistry which may lead to knock under certain conditions.

So given the facts that the lack of lead is not an issue, that most rebuilders are not improving the original design or using better parts upon inspection, what the rebuilder is saying is that he's only changed the head to deal with knock resistance of today's fuel. That's generally a big lie when you are paying bargain prices because the only way to prevent knock is to lower the compression of the engine. After a valve job, the heads of the valves may have sunk a little (they shouldn't have if the job is done right) but otherwise the chamber will be the same and give the same compression!

What about using parts that are more heat tolerant or seats installed to withstand higher temperatures? While it's true that this type of work will improve the head for long term use, it has nothing to do with the fuel.

Further, these same budget head rebuilders who rebuild stock engines and provide with those engines the same kind of heads often state in their warranties that you must use 92 octane fuel as per VW's recommendation in some unknown bulletin that they can't send or quote you. What good are the modifications they've made to take the "unleaded, lower quality fuel of today" if they are telling you to move up two grades of octane? Of course, if VW gave a damn about their aircooled engines still running today they might make a recommendation that 92 octane fuel from 2004 is factory recommended but this still has nothing to do with the rebuilder's claims.

So the budget head rebuilder improved head design is stretching the truth if you ask me. One type of modification that sometimes occurs is a step-cut to the sealing surface. Generally you get this for free because your head is not a VW casting but instead from AMC in Spain. AMC step cuts their heads to improve the head/cylinder seal by eliminating the head gasket (see section below). This step cut doesn't by itself have anything to do with unleaded or low octane gasoline. I've even seen rebuilds with one AMC head and one VW head that has not been step cut!


So what grade of gasoline should you be using?

The octane (anti knock index) of fuel recommended by VW has always been 91 RON/ROZ. However the gas pumps are rated using the US system (RON+MON/2) which is the source of some confusion for a lot of VW owners (the manuals are dated sometimes partially translated as this RON criteria shows). What do all these terms mean?

RON is computed using a special variable compression engine that determines when a zero octane fuel begins to knock. MON is computed in a more traditional engine that simulates load/heat using 100 octane fuel.

Converting units, 91 RON is the same as 87 (R+M)/2 at the pump (91 RON + 82.5 MON = 173.5/2 = 86.75 octane). Some, actually a lot of, VW owners get this confused thinking that 91 grade gasoline is what they have to use in their engines because VW says so. If their engines are stock, they are incorrect. Once you increase the compression ratio with your rebuild you are forced to use better grades of gasoline.

The reason rebuilders recommend 92 is to protect their warranty. Because they assemble the engine so quickly, they do not cc the combustion chambers or make sure the deck height and compression ratio are properly set. By paying for this higher grade of fuel YOU are buying insurance for the rebuilder against knocking because one or two cylinders prefer to ignite more easily (higher CR) than the others. The real reason you might need a better fuel is based on your driving habits.

Boston Bob wrote a good writeup on VW and fuels. The VW head is subject to far greater range of temperature values than the average watercooled car filling up on the same pump gas you are. When you are hill climbing on a 100F day the engine may benefit from a higher grade than 87 because you are buying some anti-knock protection. Although octane requirement decreases with elevation, head temperature increases so once again the aircooled design suffers vs. the watercooled design.

Not convinced? Install a CHT gauge, fill up 3 times with the 3 different grades of gas, drive the same hilly route on a hot day and see for yourself.

Valve, Guides and Springs:

Valves come in several sizes depending on the stock displacement of the engine. See the references at the end of the article for the specifics.

Valves tend to suffer more than the other parts of the head. They warp, break, stretch, bend, etc. The grooves in the end for the keepers wear away, the end of the steam mushrooms and the list goes on.

The guides are made from a bronze alloy and they are lubricated by oil that is drawn from the valve covers through the clearance between the guide and the valve stem by the suction of the gases on the other side. Stem to guide clearance is important to maintain proper lubrication. When the valve begins to rock it begins to hit one part of the seat first which increases wear.

Valve springs are probably the most boring part of the head but they are as critical a part as any other. The spring is responsible for returning the valve to it seat after the cam has finished lifting the valve off the seat. About 1 in 10,000 new springs fails which mean they must be tested like any other part. Used parts must definitely be tested.

Head Gaskets:

This head related topic probably stirs up more argument that any other. In the 1990's VW released a tech bulletin that among other things recommended the removal of the shim gasket between the head and the cylinder. This gasket is actually made from several thinner gaskets pressed together and when one layer shrinks and burns through a gap forms that allows the combustion gases to escape resulting in eventual head failure.

head gaskets

The consensus on this issue, after much photographic evidence depicting burns in this area, is to remove the head gasket and lap the head to the cylinder and rely upon a carbon build-up to help maintain the seal. Now there is no gasket to fail because there is one less sealing surface but it means you have to get the contact surface absolutely flat because there is no gasket to crush and make up for inconsistencies. It also means that you need to recalculate your compression ratio but that is a necessary part of an engine rebuild anyway.

The gasket (~0.80mm before crush) in not the same thickness as the 1.6mm spacer VW recommends usingL

VW Cylinder Head Gasket Tech Bulletin

This begs the question: has VW also recommended lowering the compression ratio or were there further instructions? Well, first consider there is was also a paper gasket at the base of the cylinder and if that measured 0.80mm also then there should be no change to the CR but Wilson says the paper gasket should be 0.20mm thick and so is the metal base gasket included with the rebuild kits so it appears VW did recommend a slight reduction to the compression because overall the deck height has increased by 0.6mm (approximately 7.3:1 to 7.0:1).

AMC step AMC warning

AMC has solved the problem of the weak head gasket by machining a step into the head in place of the original gasket. This serves to provide an edge and mating surface that obstructs any escaping gases after sufficient carbon buildup which is promoted by the gasket sealer used. The step is about 0.80mm thick (incorrectly shown as 1mm in left photo) which is about the same as the gasket that was removed. The head combustion cc is about 50cc on a 2.0L AMC head below the step. Above the step it's about 57cc.

To summarize, if you elect to remove the gasket you've got some machining to do to your head and some math in order to add the step. If you have an AMC head you don't need the head gasket (AMC even gives you a warning notice with your head). Electing to use the gasket is an acceptable practice on a VW head if you cannot guarantee the seal but only if there is no step cut in place.

The Rebuild Epidemic:

Now that you understand how the cylinder head is constructed it will become clear why some last longer than others. It's not clear to me whether or not head rebuilders do poor quality work to a) make a quick buck b) they don't know any better or c) they are hoping to sell you another head or d) they are trying to up-sell your their various services. From what I can ascertain, the competition for head rebuilding has driven the price down to the $200 level and the quality along with it. A good rebuilder would take a day to redo a head and is this not cost effective enough for most VW owners so they take their chances with a budget rebuild. The business opportunity to provide properly rebuilt heads is limited. Of course, when the head fails and you have to replace the heads and R&R the engine, the money that could have been invested in better heads has been blown even before the repair bill is tallied.

If you read Bentley, VW's recommendation is to simply buy a new cylinder head rather than attempt a repair. This advice was provided a long time ago when VW solds new heads for $125 in the late 70s and early 80s. Today, these components are NLA so you are forced to rebuild them. Rebuilding as a practice was invented by racers who were forced to modify their parts but as the parts became unavailable these techniques were adopted to provide repair services for stock heads. Luckily VW provided the specifications so there is no guesswork required.

A number of years ago, AMC of Spain (Amadeo Marti Carbonell S.A.) decided to cast new heads because VW stopped. Unfortunately although the castings are new, the valve job that is done is not nearly as good by VW factory standards. I've also been told that the valves used are not as long lasting as the originals and there is some evidence to show this. The result is a new head with a reduced lifespan compared to a VW factory head. AMC heads cost $400 and opinion is divided whether or not to improve the machine work done to the head after the casting has been imported at greater costs vs. rebuilding a VW head to the same standards. Because of the cost of heads, warranties and time invested engine rebuilders are modifying the AMC heads for extra insurance because they are stating that AMC heads out of the box last 60k miles max (still pretty good but 100k+ would be nicer).

When rebuilding a VW casting, it's generally thought best to start with an old VW head that has already been heat cycled many times over. A used head can be had for $30 or less depending on its condition. If you've ever broken a exhaust stud boss or stripped the spark plug threads, you might want to invest the repair cost into purchasing AMC heads because these repairs can be expensive.

Of course a used head maybe cracked, have broken studs, etc so they have to be inspected and possibly repaired before the valve job is performed. Bus heads have a tendency to crack in various locations when they are abused. Finding a crack free head for $40 is a large savings over a $30 head with a known crack because welding repairs greatly increase the rebuild cost.

A rebuilt head can be bought for $200. Cheap heads exist for the same reason that cheap engines and paint jobs exist: price competition which also allows unscrupulous owners to sell their vehicle more easily. Unfortunately what you get for your money isn't very good value if it fails prematurely. You could replace the heads without removing the engine but it requires so much disassembly in a confined space the process will take much longer overall. If you are paying somebody to do this work, they will undoubtedly remove the engine first.

For $200 this is what you'll get from the average head rebuilder:

Let's address the problems with each of these items to demonstrate why you should demand more from your head builder:

  1. Single-angle valve job

    This is probably the #1 quality issue. Automotive engineers have been using 3-angle valve job for decades. VW is no exception to this requirement but the low budget head rebuilders skip this step. Do not be led into believing that 3 angles is only for high performance engines. The angles are chosen to increase the flow of air through the head and you may have heard of a flow tester that rates heads by how many cubic feet per minute (CFM) of air will flow through. The amount of air that can flow is related to the lift of the valve, the geometry of the intake/exhaust port and the angles on the valve seat. With a stock engine the valve angles are the only variable during a rebuild (same port geometry, same cam). Some high performance heads use 4 or 5 angles at the expense of head life.

    The volumetric efficiency (VE) of the head is a percentage of the air that can flow through the head at any given RPM. For example, when the piston of a 2L bus engine (493cc) completes its intake stroke it will draw in of volume of air that is somewhat less than the cylinder displacement because it is obstructed by the shape of the intake throttle, manifolds, ports, etc. The RPM that maximizes the VE is the speed at which the maximum torque is produced.

    So what does all of this have to do with the head rebuild? VW matched the profile of the cam and the ignition timing to the design of the head. Alter the way that the head flows (reduce it) and you've just lost a bit of the precious little power that VW engineered in unless you are modifying the rest of the engine. If you don't think this is a big deal just drive a 2L with a low budget rebuild and one with a factory head with excellent compression: there is a big difference in power felt and gas mileage achieved. The 3-angle valve job is critical to engine performance and economy. When head rebuilders ramp up their production they seem to favor skipping the 3-angle job.

  2. Valve Seat Concentricity

    In order to speed the reconditioning turn around time (aka profit) the low budget rebuilders use grinding stones to cut the angles in the valve seat. If they use 1-angle, they use one stone. 3-angles requires the use of three stones. The tool that is typically used is a large drill with an angled head and as the stone grinds the seat the sparks fly. The drill is guided by a pilot that uses the valve guide for support. If the guide is worn out then the pilot will not keep the stone in position during grinding. Similarly, if the stone is worn out, improperly dressed or the operator makes an error the result is a seat that is not concentric with the valve guide and thus not concentric with the valve head. Since the valve's ability to cool depends on its contact with the seat, poor concentricity results in overheated valves that stretch, burn and sometimes break.

    The rebuilders would like to have you believe that your head failure was your fault, that you overheated the engine, tuned it incorrectly or otherwise assembled it wrong. They put heat tabs on the head that fall off so they can invalidate your warranty but the fact is that non-concentric valves will overheat from the first moment the engine fires and then it's a slow path to failure which probably occurs right after the warranty expires. Why don't they care? They can sell you another head can't they? Of course, not all head rebuilders are this devious but the auto repair industry hasn't exactly been know for its honesty over the decades. In this case, you simply get what you pay for.

    The rebuilders use a dye (Dykem is one brand) to paint the seat so they can see how much they are grinding away. Often when the seat is not concentric they simply keep on grinding. One of the side effects of this is that the more they grind the deeper the valve sits in the head and this affects the combustion chamber volume and valve lift. When this becomes unbalanced, so does the compression ratio and the smoothness of the engine.

  3. Valve seat seal
    Another important reason for a 3-angle valve seat is to increase the contact pressure between the valve and the seat for a good seal. With one large angle the concentricity of the seat becomes even more critical.

    The second problem with grinding stones is that they leave a rough surface behind. The stones have to be continually dressed and used with a compound as they cut and they wear during grinding. This leaves a rough surface. The valves were probably cut to the specified angle with a machine at the factory and so you have a very smooth surface trying to mate with a rough surface. The result is that the valve fails to make a perfect seal which you can witness if you pour some solvent into the combustion chamber and watch it leak into the intake and exhaust ports. The rebuilders are hoping that the action of the valve hitting the seat in the first few minutes of operation will beat the seat into shape. Lapping the valves is a common practice but it cannot cure the roughness left by the stones without a lot of work.

    A better way to cut the seat is with a carbide cutter. Most high performance VW shops and any reputable Porsche head rebuilder will tell you that stones are only good for old American cars and they are right. The only way to get a long lasting seal is to use carbide tungsten based cutters. Some cutters do one angle at a time, some all 3 at once. Although the cutter is inexpensive it requires a expensive machinery to hold the cutting blade and it takes time to setup and align the tooling to the valve guide pilot which is why the low budget rebuilders avoid it.

    One advantage of the 3 angles at once method is that it cuts the seats to the prescribed width evenly. The contact area is critical to the flow and cooling characteristics of the valve. Exhaust valves typically have a wider contact seat to increase cooling capacity. When a monkey uses the stones and rushes the job, the amount of contact surface can be too little resulting in overheated valves or too much resulting in a poor contact, again overheating the valves.

    Another problem with the stones and even the carbide cutter is that the shop you use may not be setup to cut the angles required. On a Type 4 head, VW specified 15/30/75 for the intake valve seat and 15/45/75 for the exhaust valve. You might end up with a 20/45/60 grind because those stones are more readily available for their tooling. The seat has to be cut to the prescribed angles and seat width and this process takes time: something the low budget shops cannot do and still charge you $200.

  4. New and used parts

    Any rebuilder re-using exhaust valves should be avoided like the plague. The exhaust valve is the first part that fails, no doubt from the previously poor rebuild job. The factory heads last 100k miles or more.

    You should insist on both new intake and exhaust valves in your rebuild. Exhaust valves simply warp or break. Intake valves wear along their stem and have to be replaced but sometimes the rebuilder simply regrinds the face of the head leaving too little material behind. This affects the combustion chamber cc's as well as the strength of the head which can lead to cracks. Some valve heads are stellite coated and you can only grind them so thin before they are unusable. Stainless steel valves are inexpensive and popular these days so there is no reason to re-use valves.

    Valve seats should also be replaced. Old seats can fall out of their pocket at high temperatures and newer material with the proper hardness should be installed (old seats soften with age).

    Valve guides have to be replaced during each rebuild. They are made from a bronze alloy and wear out. The result is that the valves rock in the guides and that increases their wear. When the valve rocks it does not hit the seat concentrically. More importantly since the valve guide is what holds the valve seat pilot in place, you will never achieve an in spec grind on the seat unless the guide is within spec. Because each guide compresses a little as its installed, they must be reamed to size and honed for smoothness. Again, the budget rebuilders either skip or rush critical steps for long head life. When the stem to guide clearance is too large, the valve is not lubricated properly and the oil consumption increases because the oil under the valve cover is drawn through the guide by the airflow on the other side.

  5. Re-used hardware (avoid)

    Springs are pretty foolproof but they do wear out so they must be tested. New springs are a much better option for a rebuild but they must also be tested. One sign that the springs are old is that they have been shimmed. Shimming is sometimes used to maintain the correct spring tension when the valve seat has been cut deeply but for the budget rebuilder it's often a workaround. The problem with using shims is that the push rod doesn't know there is a shim there and if the lift is sufficient the coils of the spring may bind before the maximum lift is achieved.

    Exhaust studs often have terrible threads and sometimes pull. Because of the difficulty of the repair while the engine is in the bus you should have them replaced (at least the questionable ones). The tricky part of this repair is that the studs must be perfectly aligned otherwise they will never allow the manifold of the heat exchanger to fit. Be sure to take along part of your exhaust to the head/machine shop to ensure the studs are correctly installed. Intake studs are usually ok unless the threads have been stripped but they can break off later and you'll pull your hair out trying to remove them. Sometimes you'll be forced to pull the engine to perform the repair.

    Since you typically keep the rocker arms, rocker shaft, etc before you send in the head for repair, it's up to you to determine if they are worn. If you have bought a longblock then these items will be included so you should definitely check them over. Typically they will only have been degreased and bead blasted to clean them and often that's all that's necessary. Check the fit of the rockers on the shaft and make sure no hardware is missing: count the washers and check with Bentley. When replacing the engine, compare the parts from your core to the longblock and send back the worst ones with the core.

    Valve adjusting screws wear out. If you screws are not nicely radiused you will have difficult performing your valve adjustments. The rebuilders simply give you whatever screws they have. Check them and replace them with Genuine VW screws only: they are the only ones with the correct shape and hardness.

    Look out for missing wavy washers and replace rounded nuts.

  6. Spark plug threads

    Because the head is aluminum and the spark plug threads are nickel plated steel, the threads in the head can easily strip if someone was clumsy. Before you install the head, make sure the spark plug can thread in easily. If it cannot then you must chase the threads. If the threads are in poor condition they may not be able to hold the spark plug under the compression causing the plug to shoot out. In that case the hole has to be Timesert or Helicoil repaired. Check your rebuild for a previous repair of this kind.

  7. Misc welding

    Often a used head will have been welded to repair a broken fin or crack in the combustion chamber. Typically these repairs will go unnoticeable but there is one area you must inspect. Any welding in the exhaust port must be low profile because a sloppy thick weld may prevent the exhaust manifold from seating. Again, check this before you install the head. Also inspect the seat under the copper ring in the exhaust port for large nicks.


So there you have it: based on the criteria I've laid out, head inspection and selecting a rebuilder is an involved process. If you want to guarantee a long lasting head you must go to the lengths I describe in the article.

Your biggest challenge will be finding a head rebuilder that meets these standards at a reasonable price. Often with VWs, we tend to base most of our purchasing decisions on price alone but remember these vehicles are 30+ years old now. They are classics that are coming of antique age and as such repair gives way to restoration as fewer and fewer people offer services to properly repair them. The price of quality repairs is only going to increase over time so if you want to enjoy your bus for the next 25 years, spend you money wisely now and get the job done right.

Gallery of Shame:

broken valve ruined head
This exhaust valve stretched until the head broke off... ...the head was bouncing around the combustion chamber...
piston damage valve face contaminated
...and it beat a hole in the head of the piston before flying into the crankcase. The larger intake valve (left) was bead blasted and re-used but the rebuilder didn't care that it still had something foreign welded to its face.
spring shim stem wear
These springs have been shimmed on one head because the springs have worn out or the seats were cut too deep. Valve stems have galled after 700 miles because of worn out valve guides.
#2 valve seat no bevel
Valve seat is not concentric with valve head. The valve didn't seal so it overheated and stretched. The blue is a machinists grease called prussian blue and is used to check mating surfaces. There should be an even width blue ring all the way around. The tell-tale look of an el-cheapo valve job: ONE large angle. This head will not breathe properly and the valves will not seat properly. You can see the rough surface left by the grinding stones in the enlarged photo.



02/29/05 - Created
08/06/05 - AMC photos courtesy of Matthew Brooke
09/07/11 - Fixed broken photos, added translate button, updated footer
07/15/19 - Google update: new adsense code, removed defunt translate button
11/17/20 - Updated links in references