Head Gasket Failure

[Colin Morgan]

Having returned the car to the road yesterday following head gasket failure a week ago (when I was kindly rescued), I am wondering if there might be anything to be learnt from this experience.  Most failures are down to 'operator error' and this obviously wasn't an original Austin component but one fitted by me only 3000 miles ago in 2017.

The copper-asbestos type gasket that failed had a smear of orange Hylomar beneath it because the block is a little marked in places (though not near the failure).  It was torqued to 18 ft/lbs.  The gasket failed between the 2nd and 3rd cylinders - a piece about 3/8" in length had blown out in a direct line between the two exhaust valves.  There was no other damage.  Significant cooling water had escaped into the oil (mostly after the engine was stopped, I think).  

I wondered if that bit of the gasket was running hot - perhaps the cooling isn't all it might be in that area?  Or whether the relatively low torque used had contributed to the failure?  Or if the layer of Hylomar had contributed?  Or a bit of everything.  

On assembly this time I used a smear of copper-slip underneath the gasket, and torqued to 18 ft/lbs again.  All fine after three runs out.

[Chris KC]

This is a topic which gets aired fairly frequently.
18 or 20lb.ft should be perfectly OK, better still if you tighten in small increments following the recommended sequence. 
Personally I never use any sealant, just a smear of grease.
The nuts should be torqued up again after initial running.
Always tighten with the engine stone cold.
Otherwise the main thing to consider is whether your head and block upper face are truly flat. Cylinder heads are often slightly concave in the middle. I prefer the old-fashioned remedy of lapping out any hollows on a surface plate (milling or grinding takes off too much metal and hikes the compression ratio).
If these surfaces are not flat within a thou or two the gasket will struggle to seal.

[Andy Bennett]

And there is the 'standard' torque ti down, leave it for a few hours, torgue again, run it for a couple of hundred miles and torgue again.

[frequentflyer]

Spray hylomar on the gasket both sides, smear some grease on the threads and washers as any friction eats up your torque and is removed from the clamping force between head and block. Retighten several times.

[dickie65]

If you lubricate threads you have to reduce the torque applied.
By Torquing the bolts / studs you are stretching them slightly.
There are charts on the web that show lubricated and unlubricated torque settings as well as differerent grades and material fastenings.
We use Torque Charts a lot at work.

[David Cochrane]

From what I have been told by a professional mechanical/automotive engineer, original Austin 7 nuts & bolts/studs should be torqued up dry, i.e. no oil or grease on the threads. The torque reading relies on the friction between the threads, and as mentioned above, lubricant reduces this friction.