Welcome to our resource for MG Car Information.



MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG MGF Technical - Head gasket post mortem

For the morbidly interested, I have posted a couple of pictures of the cylinder head just after it was removed this afternoon. I couldn't see any evidence of exhaust gas blow past, but quite clearly, the head has seen better days!

The gasket's elastomeric sealant simply fell away from the metal as it was removed from the head - astonishing! But there was absolutely no evidence of water/oil mixing, so this wasn't the mode of failure.

What you can see, is where the liners have seriously marked the cylinder head - looks as though the alloy has gone soft? The fire rings on the gasket were pretty manky looking too...

The loss of power is, I suspect unrelated to all this. The spark plug electrodes have been significantly erroded- and indeed show distinct evidence of having been running extremely hot (see middle picture - the electrode ceramic insulator has a small fissure evident). Plus, look at the colour of the exhaust valves!

Which suggests the engine has been running extremely hot. Question is, why? If lean running, I'd be surprised. The last time I was up at Emerald for a RR, we checked the mixture. Running slightly rich. The only thing that has changed since is the new Powerspeed 4-2-1 manifold (which was superb BTW) - but whether that is enough to tip the fueling equation the wrong way, I have no idea?

Thoughts and suggestions appreciated. New DVA ported head going on tomorrow... I would rather that didn't suffer the same fate as the original cylinder head!
Rob Bell


I have justed ordered the same complete exhaust system from Powerspeed, Dave there did say it would need to go back on the RR to be re-mapped but I assume he is just talking about tweaking to get the best output. I'm going for the Emerald ECU.

It's all going on a brand new engine too, so I'll be very interested in your findings..


Stu Dickens

Rob, You appear to have a failure sequence of a coolant leak, low coolant level overheating (boiling)with the resultant pressure releasing at the weakest point. In many instances, the HG is the weakest point but your pipe failed first.
After rebuild, you need to find that leak.
Geoff F.
G. Farthing

Geoff - I found the leak after all the coolant was lost - a split upper hose. I tried to take a picture at the time, but at midnight, it was too dark for my camera phone!

I've now uploaded the picture of the split hose I took yesterday: I had to perform a temporary repair that involved cutting about a centimetre of hose off the end to move the split section beyond the flaring at the end of the stub pipe. As the picture shows, the split was pretty massive - about 3/4 inch long! :o(
Rob Bell

The valves look ok to me - (but I have a dark monitor at work)
As for the engine getting very hot - you do race it, how does it compare with others techspeed see from their race cars (actually I guess there not campariable as they come apart every few k miles)
Will Munns

True Will. Plus the last thing that my car had done half a dozen laps of Castle Combe before I retired it and drove the F gently home to Warminster to await a low-loader.

You're right, exhaust valves are often lighter-coloured than the inlets (the inlet valves are cooled by the air/fuel mixture) - but what struck me was *how* light. The heads I've seen to date (mainly on Dave Andrews' benches) have shown exhaust valves that are darker coloured than these - and that includes Peter Carmichael's old 1700 K-series cylinder head - and he races (raced!) that one too. Trouble is, my experience at looking at the inside of K-series combustion chambers is limited. Perhaps there is a problem. Perhaps there is none - it would be useful to know either way!
Rob Bell

check the liners height before fitting the new head and get the latest Rover (Payen) gasket with steel dowels and new bolts.
I'm aware you already knew that, though! ;-)

IMHO valves and spark plug colours are fine (plugs are worn and damaged, but this is another story)
if the elastomer was loose, it is possible you also had a tiny leak just above the alternator (if that's where the rubber was detached).

Hi Rob,

I think the markings of the gasket firerings on the head is _the feature_ of a high mileage K-engine.
Not the uneven though, but markings in any way.
I remember discussions with the fellow MG mechanic know from the 1999 Dormagen Treffen with Dirkies appearance. He predicted the micro-hammering of the fire rings onto the head would lead to such markings and a blow out earlier or later.
Happend to you more or less earlier. So what, it's normal, isn't it.

I wonder about the high temperatur of the pale brown spark.
Could this be due to a problem mith the affiliated injector ?

Didn't get what makeyear the other stub reffers to
Is it from a new engine ??


I know this is pretty obvious - Robs first picture of the Stub connection is typical of the one piece casting that was used and does eventually corode, losing its "lip" as shown.
The second picture shows a modern Stub connection where an insert, s.s or m.s. has been inserted to avoid that problem.
On some of my earlier "Classics", the one piece stub connection has literally fallen apart when removing the hose. Often manufacturers will attach the hose using a pinch clip rather than a Jubilee clip just so that you can't overtighten it and cause more of a leak than you started with.
Sorry to state the obvious to those that knew.

Thanks John - well worth stating! No idea when the stub design was changed though? Dieter, I don't know the engine number relevant to that particular casting, but the new head is from a TF135 - so pretty recent.

Carlo - feeling suitably weak of knee waiting to hear whether the liner heights pass muster. A quick thumbnail test was encouraging though. No external water leaks were evident at all - but I was shocked at the condition of the gasket once the head was removed. Practically ALL of the elastomeric beading simply fell off the gasket!!! Will be replacing with the latest version, and will be replacing the original plastic dowels with the steel upgraded ones (this is the first time in 10years that the head has come off!!!)

I've also posted this story on SELOC. Dave Andrews mentions that water leakage into the combustion chamber can result in 'white' spark plugs and exhaust valves. That is a possibility here I guess?
Rob Bell

If all of the plugs and valves are the same colour, it would suggest that if it were water leakage, it would be seeping into all cylinders equally... which I would think to be unlikely.

I agree with Will - with the same proviso regarding monitors - the valves don't look too badly discoloured to me.

Again, as per Will, have Techspeed made any comment about the condition of the head?


Looks as though there is a consensus growing - DVA reckons that the discolouration looks normal too - so I think that I am worrying about nothing. Good to have my mind set at ease on that count! :o)

Mike - good point regarding the question of water entry. All the exhaust valves and spark plugs look the same. There is, however, heavy indentation from the liners around all four cylinders too. Unfortunately, I suspect that this head could be scrap :o( Some of the indents are very deep.

Which brings me round to the next question: how long has this cylinder liner marking of the cylinder head been going on for? Years? Or has it all happened as a consequence of loosing coolant?
Rob Bell

T J Nicolson

Geoff is 100% correct. The coolant level on the F is critical. If it gets too low, (and it doesn't have to be all that low) then an air lock develops which stops the coolant circulating through the radiator - the engine can be imitating a kettle, pushing coolant out everywhere yet the radiator remains stone cold as no coolant is going through it.
The presurising in such cases is simply good old fashioned boiling. Heaps of pressure, coolant coming out everywhere, but just driven by boiling rather than compression pressure.

If caught soon enough, no damage occurs - simply top up, bleed the air out of the system and off you go.
We've done this to successfully to a few cars now which we were confident had not gone past the point of no return in coolant loss/temperature rise and have monitored the resulting job over a significant period time. No further problems have evolved.

Unfortunately most people don't notice the high temperature in time, not the least because the engine is behind you so there is no telltale trail of coolant coming out of the bonnet onto the windscreen. The result is all/most of the coolant get boiled off, and the head overheats and softens, leading to the indentations from the fire ring. No headgasket can cope with a collapsing head.

The give-away is the fact that the indentation is worst round the exhaust valve, where it is hottest. If you want to confirm, get the hardness of the head checked. Starts off 100+, servicable above 90, I'd expect that head to be down in the 50 range in the worst spots, possibly lower.

Dieter, I disagree about it being mileage related. We have checked out quite a number of heads now, and have yet to find the indentations on any head still retaining its correct hardness, regardless of milage. The softer the area of the head is, the deeper the indentation goes. And the only thing that causes the softness is serious overheating.

Giving the car a good thrashing round a track won't cause a problem so long as the cooling sytem works properly and carts the heat away as it should - the villain is coolant loss. Without coolant, nothing stops the head from reaching seriously unpleasant temperatures.

We have succesfully rehardened such heads here, the guinea pig head being the current one on our K midget which has done a couple of years really hard graft with no ill effects. (There was nothing wrong with the old head, just wanted a good test bed to verify if we could sell such a process to our customers with confidence, and the K Midget works harder and is easier to work on than our F)

Paul Walbran

Thanks TJ. Hopefully you've got a reply; email is playing up for some reason... argh!
Rob Bell

Paul, unfortunately, I think you're right here. Amazingly, and unhelpfully, the temperature gauges all read normal, even after the engine stalled and the coolant had all bled away.


I didn't get any indication whatsoever that the coolant level was low until the heater went cold, until I slowed down and could smell the sweat-coolant smell of doom :o(

My excuse for not noticing was that it was dark! Nothing to see!

Still, have bought one of Brown and Gammons low-coolant level kits. At least this way I'll have some kind of warning if this happens again!
Rob Bell

Paul ponted out what I refrained from saying. It looks like the pipe burst before any HGF damage and the strip condition is no worse that can be expected from an older, well used engine. The overheating caused the pipe to burst, but this may not have been the original location of the leak.
A new pipe, some OAT and a little water together with subsequent monitoring of level may well have been all that was needed to recover the car.
You have various rebuild options, albeit you seem to have taken the most expensive.
The engine worked OK for a long time whilst the head fretting occured with the old gasket. A slight skim with some residual fretting and a new gasket would outlast your use of the car.
Your option to repair to as good as if not better than new is admirable for an enthusiast but not cost effective against the residual value of the car.
As it happens, I cannot disagree with your choice as I demand 100% reliability from my MY2000 and had TS rebuild and upgrade the top end of the engine this year at 22,000 miles before the Highlander and a 2,600 mile European Tour.
Geoff F.
G. Farthing

My car is at the age where doing anything at all becomes 'cost ineffective' Geoff; this car is enjoyed and lovingly abused - and to that end, it is easier to justify what I've done to myself. ;o) That I happened to have a 'spare head', was planning to swap the cams etc means that this wasn't a wasted opportunity to have a little bit of 'interesting' work done :o)

Regarding the markings on the cylinder head mating face - what I don't know is how quickly this appears. Is it something that appears over a very prolonged period of time - or is it something that can occur quickly, if the alloy head is overheated and thus goes soft? One thing is for sure, the marks are pretty deep - I doubt that one would be able to skim that much material off without seriously increasing the CR!

How much does some metalurgic hardness testing cost? If not that expensive, it may be interesting to see whether the alloy has gone soft or not...
Rob Bell


thank you. The hardness, *Bingo*, I think that's it with overheated head conditions.
Please add some details about the hardness tests you did.

Easy tests are possible with craftsman tools. Need to look up in English. German is *Poldihammer* for Brinell like hardness.

The question is. What kind of Aluminum (AlMgSi ??) is the head made of. This details are required to find the parameters for hardening the material. (Not so easy like steel hardening. Temperatures need to be held very correct to get good results)

@ Rob, do they have a sensitive look at the block for loose or dropped liners ?

Had a look again to the _new_ stub with insert picture.
Hmm, may be they found the tooling for the casting worn and decided to go the insert way to overcome this problem (may be temporary) ?
(Quality issues and temporary _improvements_ never led to p/n changes. Remember the white dot to the header tank cap.)

>low-coolant level kits
Errrr, I'll get my own DIY version in ASAP :)

Someone told me the material specification, but sorry, for the life of me, I can't remember what it was :o(

Regarding liner height, this is something that I discussed with them - and I am sure that it will be looked at. A thumbnail test I did on the block was encouraging - but that's hardly the same as measuring it! ROFL

Regarding the stub pipe - I reckon this could be a permanent change rather than something temporary - albeit not one that resulted in a different part number. The cylinder head casting is slightly different too - insofar as one of the coolant passages has been deleted, yet the casting number is identical. I suspect that the stub pipe alteration is in response to a recognised problem of corrosion/wear resulting in poor sealing and potential coolant leakage (as I found).
Rob Bell


A number of split hoses are triggered by overheating arising from gradual cooling loss from a slow leak, leading to air lock, circulation failure, boiling, excess pressure and thus a tired hose gives up.

However from what you have described, I'd say that the failed hose caused the overheating rather than vice-versa. That the temperature gauge showed normal indicates sudden coolant loss, as the gauge monitors the coolant temp. With the coolant gone, it has to wait for conducted heat from the head, and there is usually a lag in this showing up.
The gauge could of course be faulty, but the lag in reading is very characteristic of sudden coolant loss.

(I might be showing my age, but the first time I experienced this was when I was stranded in the dark out in the countryside in my old Morris Oxford van when the heater suddenly when cold while the temp gauge still went normal ... )

In these circumstances, I'd expect the head to soften substantially and to collapse around the fire ring of the gasket very quickly - so I think it appeared quite quickly.

As I noted before Ive seen high milage heads which show no problem, and every head I've seen with the problem has been hot at some stage. A lesser degree of cook-up seems to result in the damage taking longer to appear, but it's always instant when the overheating is major.

A good guide as to how hot the head got is to check the inner cam belt cover - if it has melted (and I've seen that on a quite a few!) then the head got seriously hot.

Paul Walbran

Sounds like a coolant pressure sensor is required as well as coolant level detector. But where to measure it is the question. Shades of Concorde air intake pressure. So lots of sensors neede.
Ken Waring

That's very interesting, thanks Paul...

>> I might be showing my age, but the first time I experienced this was when I was stranded in the dark out in the countryside in my old Morris Oxford van when the heater suddenly when cold while the temp gauge still went normal ... <<

Near identical circumstances in my case!

I didn't inspect the cam cover all that closely - grossly, it was not badly melted. However, up close, I have no idea.

Something to look at when I go and collect the car (hopefully tomorrow)
Rob Bell

Just to clarify, I was refering to the plastic cover between the cam gear and the head. When it melts, there is no doubt - it looks very soggy, all sagged with a gaping gap.

Dieter - yes, it is the Brinell scale I was referring to, measured by an electronic box of tricks. Gasket makers reckon that you can get by at 75 but I prefer 90.
We don't do the hardening process ourselves, leave that to a local heat treatment company which has had years of experience reclaiming Japanese heads. It has proved to be a reasonably economic process here, but that's probably because there are not many second hand heads here so their price tends to be a bit high.

Paul Walbran

Paul, that's the one I meant too.
Rob Bell

The 2005 TF has an idiot warning light for low coolant level. It seems to be pretty conservative too - it comes on even if the level drops only slightly. This, I suspect, is a good thing!
Mister T

Thank You !

>local heat treatment company
Yes, indeed, a lot of should exist in the UK also.

Will discuss this point of hardness soon with a German Tuner and get back with his experiance .. if he tells me ).
I've a new exhaust on order there and take the option to speaking with the guy (ROTEC, Bernd Rossol)

Dieter <<--- off now for some days vacancies :)

Heat treatment is too late for my old head. Anyone got instructions on how to turn an old K-series cylinder head into a coffee table? ;o)
Rob Bell

For this you will need 2cm thick glass sheets, 50cm x 70cm x 2, 50cm x 30cm x 2, 70cm x 30cm x2 + legs to suit.

You will also need engine oil (new) to fill the table, and 1 litre of bright green coolant to pour over the floor....
Will Munns

Instructions and pictures on line Will?
Rob Bell

>Anyone got instructions on how to turn an old K-series

Ask Mike S., I think the designer of the Head gasket wine rack will get some ideas.

This thread was discussed between 10/08/2005 and 13/08/2005

MG MGF Technical index

This thread is from the archive. The Live MG MGF Technical BBS is active now.