Cylinder heads & head gaskets: living with the alloy V8

The root cause: two metals, one cooling system

The Stag V8 pairs an iron block with aluminium cylinder heads: without a corrosion-inhibiting coolant kept in good condition, that pairing sets up electrolytic corrosion inside the coolant passages, narrowing them, restricting flow, and letting the heads run hotter than they should. Everything in this article follows from that one fact.

How heads warp

An overheating Stag will, over time, end up with a warped cylinder head, the aluminium moves under heat in a way the design didn't allow much margin for, and the Stag gained a reputation for this happening at low mileages, sometimes as little as around 25,000 miles, on cars whose cooling wasn't kept in proper order. A warped head seats a gasket unevenly, and a blown gasket usually follows.

The skim limit: 0.010 inch, and no more

When a warped head has blown its gasket, skimming (resurfacing) the head face can sometimes rescue it: but only within a firm limit: no more than 0.010 inch (ten thousandths) can safely be removed before the head must be replaced rather than machined. This is often not enough to fully correct a badly warped head, which is why prevention matters so much more than cure on this engine. If a head is skimmed, it must also be checked and straightened first if needed, because machining the face can leave the camshaft bore misaligned if the head wasn't true to begin with, a Stag-specific complication worth knowing about before you agree to a skim.

Seized studs: the other alloy-head problem

Aluminium heads on steel studs set up their own electrochemical reaction over time, and it is common, not exceptional, for head studs to seize solid in the block. Owners doing their own head removal should expect this as a real possibility, not a surprise: some studs come out fine, others snap. The practical response, at reassembly, is to dress the threads and use a good anti-seize compound (copper- or aluminium-based) on the studs before refitting, and, as cheap ongoing insurance, to check and re-torque the head fasteners periodically rather than assume they'll stay put for decades untouched. The figures and the sequence are set out below, with their sources. Check them against the manual for your own engine number and market before you rely on them.

The manual adds one instruction that is easy to miss and directly relevant here. Cylinder head studs are to be fitted to the full depth of the thread, finger tight. A stud that has not gone fully home carries its load on fewer threads, in aluminium, on an engine already notorious for seizing them.

Working on the valve gear

Three figures and one identification point, from the manual, for anyone going further than a gasket.

• Camshaft bearing cap nuts: 10 to 14 lbf ft (1.4 to 1.9 kgf m), tightened evenly and progressively so the shaft settles properly.
• Camshaft cover slotted nuts and screws: 1 to 2 lbf ft (0.14 to 0.30 kgf m). That is very low, and camshaft covers are exactly the fastening people overtighten before wondering why the gasket weeps.
• The left hand bank camshaft carries an annular groove around the periphery of the camshaft gear flange. That is how the two are told apart, and on an interference engine fitting them the wrong way round is an expensive mistake.
• Valve springs are fitted with the closer coils to the head, and tappets and pallets are assembled dry. The manual separately instructs that the tappets and cams are oiled before the covers go back on, which is a later stage of the same job rather than a contradiction.

Valve guides are reamed with a 0.3125 in reamer after fitting. The valve seat angle is 89 degrees total, with the land on the seat face 0.062 to 0.063 in on the inlet and 0.077 to 0.078 in on the exhaust. Seat part numbers are 150862 inlet and 150863 exhaust, and the head is heated evenly to 180 degrees C maximum when pressing new seats in.

One inconsistency in the manual is worth recording, because anybody machining a seat needs to know about it. The inlet valve seat bore is given as 1.519 to 1.570 in (33.58 to 33.88 mm). Those two do not correspond: 33.58 to 33.88 mm converts to 1.322 to 1.334 in. The exhaust figure immediately above it, 1.329 to 1.330 in (33.76 to 33.78 mm), converts correctly. The archive does not assert which of the inlet figures is the good one, but the millimetre range is the one consistent with the exhaust dimension. Work from your own measurement of the head and a second source rather than from either figure alone.

Source: Repair Operation Manual, operations 12.29.18, 12.29.27, 12.29.34, 12.29.48 and 12.29.76, read at source.

Symptoms worth acting on

A failing head gasket on any engine tends to show some combination of: a persistently creeping temperature gauge with no obvious external leak, coolant loss you can't account for, a sweet smell from the exhaust, white or "mayonnaise" contamination where oil and coolant have mixed (in the header tank or on the dipstick), or one bank running noticeably hotter than the other. Our cooling-upgrades guide already makes the point that a creeping gauge should never be ignored on a Stag, this is exactly the failure that advice is protecting you from. Catching it early is the difference between a gasket job and a head replacement.

Living with it: what actually prevents this

There is no trick beyond the fundamentals, and the fundamentals work:

• Correct corrosion-inhibiting coolant, kept in condition and flushed annually, see cooling-upgrades for the full picture.
• A properly working thermostat and the right pressure cap, so the system runs at the temperature and pressure it was designed for.
• Don't ignore a creeping gauge or a rattle from the front of the engine (see our timing chains piece, both failures share the same root cause of a neglected, overheating engine, even though the mechanisms are different).
• If you're buying a Stag, run it up to temperature and check both banks read evenly and the radiator is as hot at the bottom as the top, an easy pre-purchase check that can catch a problem before you own it.

Torque figures and sequence

This archive did not previously publish these figures. That was a mistake. Withholding them does not stop anyone torquing a head, it just sends them to a forum, where the figure arrives without a source and without the caveats. What follows is sourced, and where sources disagree the disagreement is set out rather than resolved.

The sequence

The manual gives the order as nuts first, A, B, C, D and E, then bolts, F, G, H, I and J, and that is now confirmed from the manual itself at operation 12.29.27 rather than from a transcription. Nuts and bolts are separate stages, not interleaved. The lettering refers to the manual's own diagram, which this archive does not reproduce. If you do not have that diagram, get it before you start, because the order matters more than the number. The manual prefaces the whole operation with a caution in bold: to avoid cylinder head distortion the sequence must be followed.

The thousand mile re-torque

After 1,000 miles, or 1,600 kilometres, the manual instructs that the head fixings are rechecked. The method is not simply to pull them tighter. Each fastener is slackened by approximately one flat and then re-tightened to the correct figure.

Where the sources disagree

Two disagreements are worth knowing about.

The Triumph 2997cc V8 Engine Service Training Notes give the head figure as 50 to 55 lbf ft rather than a flat 55. The difference is small and the Repair Operation Manual figure is the one this archive publishes, but an owner working to 50 is not working to a wrong number.

More substantially, a technical article published by the Triumph Stag Club of the Americas argues that the Repair Operation Manual torque specifications were estimates based on typical practice for the thread size rather than figures derived for this engine, and suggests that a slightly higher figure on the first re-torque, of the order of five pound feet above the manual, may be closer to correct. The same article recommends approaching the final figure in three or four stages with the engine left to rest between them, and re-torquing within 100 to 500 miles of a new gasket rather than waiting for 1,000. The archive records that view without endorsing it. It is one experienced practitioner disagreeing with the factory, which is not the same as the factory being wrong.

Figures this archive has not been able to corroborate

The following come from a single source, the Service Training Notes, in a transcription this archive could not independently check. They are set down because they are otherwise hard to find, and they are fenced off because one source is not enough for a fastener that holds an engine together. Confirm every one of them against a manual before use.

Main bearings, 55 to 60 lbf ft. Big end bearings, 38 to 42 lbf ft. Flywheel, 38 to 45 lbf ft.

Sources: the Triumph Stag Repair Operation Manual, whose tightening sequence at operation 12.29.27 has now been read at source, with its torque data still taken by way of the Vintage Triumph Register's published transcription; the Triumph 2997cc V8 Engine Service Training Notes; and a technical article on cylinder head fastening published by the Triumph Stag Club of the Americas.