The RED Subaru

Cylinder Heads Refurbishment

When I first took the Engine apart I was delighted that it seemed to be in such good condition after 165,000 miles of abuse.

Below, from a distance and from above the cylinder heads looked to be in fine shape…

(Above) one of the first jobs was to carefully file the ragged edges of the lash adjuster buckets where the adjusters’ retaining rings had turned the edges of the buckets inwards.

But, after the initial cleaning and a more detailed inspection, it was clear that not all of the components had faired well – the cylinder heads had cracked. Head 1/3 was cracked in two places between the valve seats and head 2/4 was cracked between the seats of pot number 4.

(Above) Pot 2 was undamaged and cleaned up well.

(Above) The other pots looked like this. Having spent a great deal of time on the internet reading as much as possible about this issue, I decided that this problem should be properly fixed. While it was very unlikely that the crack would reach the water jacket (though that can happen), there was the very real possibility that the valve seats would suffer. Furthermore, as the cracks spread they might cause the heads to warp and the inevitable gasket problems would cause overheating and further cracking. It is certainly possible to run these engines without fixing the cracks but I would rather do a good job and make this engine reliable and trouble free in the future.

(Above) In this shot you can see the tell-tale discolouration around the crack in Pot 1 as it descends below the valve seat into the port.

(Above) Here again in Pot3 the crack descends into the ports.

As Fields Engineering in Layton had done such a lovely job on the crank, I took the heads to show Mike hoping they could undertake the welding, the fitting of a new set of valve seats and a head-skim to regain flatness. But, sadly, their aluminium welding expert had recently retired and they were unable to do that part of the job.

After a fair bit of time on the net and on the phone, it became clear that most machine shops were not interested. I was tempted to buy a spool-gun for my MIG welder and have a go myself but I could not find out enough about the pre-heat procedure for these particular heads (various automotive net gurus could not agree even about the principle of pre-heating, never mind temperatures and times!) and, more to the point, I did not want to make a mess of the job - I have welded aluminium before but not a cylinder head.

So after some more searches I contacted Guy at Wow Automotive in Leicester. Not only did he specialise in welding aluminium heads, he had done a lot of Subaru heads. He advocated no pre-heat. The idea was that he’d grind out the cracked area, weld them up, heat-treat, crack-test, fit new valve seats, pressure test and skim if required. In late November I dropped the heads off. Below is a selection from the progress pictures Guy sent in December.

(above) Pots 1,3 and 4 have had their valve seats removed and the metal around the cracks have been ground down. While not reaching the water jackets they are very deep and would have got worse.

(above) In this shot of Pot 4, you can see how thin the divide is between the ports due to the angle of the exhaust port towards the centre; I suspect that the heat gradient caused by this asymmetry is the much written about ‘design fault‘ that causes cracking when the engine overheats. But, in my view, the problem is the over-heating itself largely caused by coolant hoses not being replaced until they fail.

(above) Here is Pot 4 after several beads of aluminium have been laid down with, I think Guy said, a TIG welder. With the other pots welded up, Guy was then delayed by the non-appearance of the ‘Cosworth’ valve seats. I use parenthesis because it was the firm that supplies Cosworth rather than Cosworth itself. I’m not sure why Guy did not use some standard ‘near-as-damn-it’ valve seats but, after waiting for weeks, I decided not to wait any longer and collected the heads from Guy, sans seats, in December. To his very great credit, Guy refused any payment because of the delays and would not accept the cash I offered him.

(Above) Having returned from a trip to Leicester to collect the heads, I gave them a good power-wash. Guy had done a first-rate job and had machined the welded areas under-size so that seats could be fitted later.

(above) here is one of the pots after welding, thread re-cutting and under-size machining. The next step was to get the heads back to Fields Engineering in Layton.

(Above) Over the run up to Christmas, Mike and the Fields crew fitted the new valve seats, cut the seat angles to the valves I provided, pressure tested the seats and, last week, skimmed the heads. Apart from a tiny bit of damage caused by Guy’s grinding stone on the edge of the combustion chamber of Pot 4, the heads look pretty good.

(Above) here is a machine line in the combustion chamber where the cutter has gone a tad deep over the weld.

By the way; after grinding in the valves, and putting the valves aside for later, the heads went in to the degreasing tank again to get rid of any oil-born grinding paste, swarf or particles of blast media. After the degreasing, the heads went through another pressure-washing session and then, finally, a going over with compressed air. In each case I paid attention to all the galleries and nooks and crannies and ensured they were cleaned really thoroughly. The valves, having been similarly cleaned went back into their marked plastic boxes while the heads were put into new plastic bags straight away to ensure a clean fit-up the following day.

(above) after fitting the valves with springs, collets and retainers in their original matched positions, I fitted the heads to the block. The rockers were also kept in their matched boxes but that’s for a later posting.

NEXT: I’ll write up the final part of the engine rebuild soon.