The Engaged Driver

Think Outside The Gasket!

Do I really need them?

Here’s why not:

• Golf Mk 2 exhaust manifold gaskets deteriorate in no time.

• The cause is a design flaw within the whole assembly.

• The remedy is omission of all gasketry.

The most unfortunate victim of this malaise is the Turbodiesel.

I should mention that though I live, for the most part, in New York, I am driving my Mk 2 GTD in Europe, especially in Germany. My GTD is a 1986 with some 110,000 miles on the clock. Its modest engine capacity is put to the test day in and day out on long stretches of Autobahn under full-load condition, regardless of traffic. Engine and tranny are still original. Not so the turbocharger. The Garrett’s bearings failed prematurely, turning the unit’s sound into something resembling a circular saw.

Here is the Problem

These engines have no torque stress bar between the manifold/turbo assembly and the engine block like e.g. the Mk 4 TDI’s. So eight M8 studs have to carry the manifold with the turbo on top, the cast iron 90° elbow with swivel flange and then the down pipe / exhaust system till the rubber mounts halfway down the line step in. But, as if that weren’t enough, you have the transverse-mounted engine rocking back and forth at every load change, which means a heavy lateral push/pull motion at the studs each and every time you accelerate or coast.

Next comes the impact of engine reveberations, internal exhaust pressure, and most important thermal stress/expansion.

The manifold design doesn’t guide the expansion well either. In essence it can creep all over the place, with (gradual) misalignment at the exhaust ports almost guaranteed.

For expansion purposes, only its bolt holes for cylinder no. 3 are M8 sized (though not a snug fit bore). All the others are oversized. Since the studs of no. 3 don’t fit tight, the manifold has enough play to move quite a bit vertically, whether bolted on correctly according to specifications or not. This is particularly pronounced on cylinder no. 1’s side (given the longer leverage). The same is true horizontically speaking, as the individual manifold ports are stiffened by cast-on-bridge segments between them that do not allow for individual thermal movement/compensation. Thus the total movement is cumulative, giving the gaskets at cylinders one and four a particularly hard time (sliding compensation).

This is exactly where the first breach occurs.

To make things worse, the manifold movement is not limited to one pane only. It curves away from the cylinder head and warps. Expansion and warpage together form a type of motion that almost literally wants to slide-press the gaskets out sideways. The forces are so huge that it doesn’t matter that the gaskets have perforated steel platings on both sides. Their bolt holes are becoming elongated with ease, so much that the inner side of the gasket face ultimately migrates like a sliding choke into the gas stream, whereas the outer side opens up the final gap.

VW’s Official Position

The dealerships (at least in Germany) are quite straightforward about the blueprint weakness. But the only cure the VW dealer offers is ordinary gasket replacement over and over again. Their job does not include any resurfacing. The parts are just being “cleaned” which means the worst is simply scraped or sanded off by hand.

Engineers of the regional VW headquarters are more defensive. One pointed out to me that the gasket design had been improved after Mk 2’s first production year. He insisted in case of a repair that the studs holding the manifold had to be replaced at the same time.

Well, I tried it and it didn’t do me any good.

Formerly one-sided, and now with two-sided perforated steel plating [part no. 028 129 589B], the gaskets are still used in today’s TDI’s with distributor-type injection pump and even numerous gas engine applications across all company segments -- Audi, VW, Skoda or Seat.

Now the Solution

My objective was to firm up the mounting as much as possible without the gaskets cracking and sliding out. Two of my Mk 2 GTD’s (the second one still up and about) suffered consistently nasty manifold gasket failures throughout their lifespans. The blowouts started at about 35,000 miles and then reoccurred at some 20,000 mile intervals. You may be aware that the repair is a rather dreadful procedure (and costly if a dealer does it), with four M10 twelve point socket head screws holding the turbo usually being seized as one of the sour highlights.

The idea to eliminate the gaskets grew on me first of all out of the drastic alternative of giving up and tossing the car. I found it striking that the mating faces of manifold and turbo don’t have a gasket to begin with. Yes, the turbine side housing is also cast iron, thus the same thermal properties as the manifold - but still. Besides, no engineer seems to have bothered about the slight blow-by indicated by the carbon lines and patches in the contact area which are obviously considered irrelevant. Engine designs from the U.S. quite frequently have no gasket material between head and manifold at initial assembly. It is only recommended for later service.

So, why not?

I did not attempt to manufacture a bracket between manifold/turbo and engine block. The available mounting areas weren’t close enough for a straight, stiff design and positioning. Besides, the specific type and range of reveberations and possible resonance frequencies have to be accounted for, otherwise you are just causing another set of problems - something practically requiring lab research - which was beyond the scope of this project. My time budget was too limited for going trial and error on this one.

The Repair

I thoroughly inspected the exposed cylinder head mating surface using a machinist’s straight edge and bright light angled painstakingly to trace minute surface irregularities and warpage. The new exhaust manifold I added (to have at least one side “resurfaced” and not thermally stressed out) was used to doublecheck on straightness / flushness.

The results were surprising: the head was as straight as you can trace with the naked eye. Surface indentation and corrosion did, however, occur at the gasket contact areas. The new manifold, shaped at the port areas as an oval ring, would overlap them well into the undamaged area.

I cleaned the head surface around the ports with fine grade sand paper and washed it off and degreased it. Next I ran a tap through the threaded holes of the studs and flushed them out. Yes, studs are originally threaded into the head with locking compound! Thus the need to clean out the remains.

Upon assembly, I decided to tighten the new studs by hand only, “textbook style”, and without applying a locking compound, which I later regretted. It’s not so much about rattling loose, but that locking compound between threads enlarges the load carrying surface. Also, starting the self locking 12mm socket hex nuts is tricky for the studs, as the designed clamping force of the nuts has the effect of high drag. This inevitably tightens the studs beyond “hand tighten”.

The studs were coated with anti seize paste, the manifold aligned carefully and tightened stagewise in a criss cross pattern (from center) to final torque (25Nm{18ft-lb} using also new thick buffer washers under the shouldered nuts). The rest was straightforward. (Except the threads in the turbo housing had to be recut and new twelve point socket head screws had to be used, coated with anti seize paste.) All others fasteners were renewed as well, the steel gasket between turbo and elbow pipe included. Here, however, I refrained from alterations simply because there had never been any problem - not least due to similar mating materials, in this case cast iron.

The Result

Remarkable. Blow-by is astoundingly low with occasional faint traces of carbon at the far ends of the manifold (where it warps first).

Performance is restored to its original - actually a touch more responsive. The engine is no longer sluggish at higher revs. Mind you, this is still an engine layout with turbo lag and no torque for fifth gear city cruising. Its domain is somewhere between 3200 and 4200 rpm. The rest must be done with the stick. Top speed is reached quickly again.

On slopes, even slight inclines, the rpm easily exceeds 4500 (the red zone), which lets you look for the sixth gear!

Yes, doing it without gaskets is the right way - the same is true for current versions of the Golf if you are in dire straits. The only thing I regret is not having done it sooner.