Doc,

You're right! However, is it worth spending £1200+ on an MG TF? You could drop in a second hand engine for a lot less than that. Sometimes the Bodgit and Scarper approach is appropriate!

As I said aobve, a similar analogy can be made with my engine (yes, I know the head bolts don't run through the block but still!). The Renault K7M block is a basically an overbored K7J. Thankfully, unlike the K7J, it doesn't use wet liners. The K7M is notorious for blowing head gaskets-but it tends to be only occur as a consequence of overheating. When the gasket went, it went badly (the engine hydraulically locked upon cranking because cylinder 2 was full of coolant. Head came off, measured with a straight edge and feeler gauge-no measurable warping. Head went straight back on the cheapest pattern gasket available, with the old head bolts and has survived for four years with plenty of full throttle runs.

No, it's not the way you should do it but I just couldn't countenance spending £1250 on a TF-or a Renault

You have to look at expected values.

If a thorough repair costs £1250 and the probability of the gasket failing again is essentially zero, then your expected cost is basically £1250.

If the £320 bodgit and scarper approach is used, the probability of it holding out might be 50%. If it fails, you have to do it properly at £1250.

(50%*£320) + (50%*£1250) =£785

£785 is a lot less than £1250...

This is how all professional engineers work. If the best available approach was used for every component, a basic car would cost £30,000, not £10,000.

Edited by unthrottled on 25/04/2012 at 16:54

My Motto is

"if your going to do a job, do it once & do it right "

I'm not interested in may be's or ifs, because what ever you offer the customer ( ie cheap job) they always come running back crying when it goes wrong !

Most people that buy a TF buy to keep for a few years, yeah we know they are bad news but having said that every H/G job i ever did on MGF's I never ever had one come back failing after 3-4or even 5 years !

If you dont skim a Rover head your asking for big trouble, the casting is very weak & soft & the head gasket sealing rings embed them selves in the head causing an uneven surface in just one fit.

madf is 100% correct, without a skim and a cheap job its garranteed to fail again, most likley very very quickly. Its not a question of probabilities with an MG

The OP's car has had the hg done: once. And it has failed again.

So the chances of a cheap and nasty failing again are in my view 100%.

Head gaskets seldom 'go' randomly. The reason the head gasket has failed again (assuming the head was not damaged and it was torqued up correctly) is that the root cause of the head gasket failure was not determined and resolved. Sorting out any cooling irregularities is much more important than skimming 5 thou off the head whether it needs it or not, because it makes people feel they've done it properly.

You see the same hand wringing about running old cars on unleaded.

The nay sayers shake their heads and say that it'll detonate itself to pieces or that the exhaust valves will recede. Most run absolutely fine

Or how about accidentally putting some petrol into a diesel tank? The 'proper' way to do it, is to have it towed to the dealership (who'll void any remaining warranty) and charge you a fortune to have the fuel system steam cleaned. Most people top up with diesel and, for the most part, get away with it.

My car has been bodgit and scarpered for most of its 16 years-and it has recently sailed through its MOT. I'll let you go back to painting your exhaust system with hammerite, or putting underseal on the undercarriage. Total waste of time and money. :-)

Doc-I'd be interested to hear what those bulletins said. Presumably the engines survived full power runs and shock cooling on the engine test bed. So there must be a reason that they go (or was it just shoddy Rover build quality?!)

madf-most of those faults are generic and would apply to any engine. Virtually any engine will crush its head gasket if the cylinder head is subject to localised or global overheating from insufficient cooling or a faulty thermostat.

I'm still puzzled as to why the K series is so prone to HG failure (uneven clamping from headbolts/uneven cooling etc).

Myy dad worked as a self-employed mechanic many moons ago, and his policy regarding HG failure was: no detectable warpage-as measured with a straight edge=no skim. He did a lot of HGs-and didn't get many people coming back wanting a refund. He put his efforts into sorting out cooling problems which were endemic in those days.

"I belive bleeding the cooling system of a MG TF of air is a work of art..."

100% correct, recommended to jack the rear of the car up high in the air, & vacum fill the cooling system ! & leave for 12 hrs !

The cylinder head is so fragile that one air lock is enough to kill it ! & as you say there is no construction to the head, its more like chocolate

"He did a lot of HGs-and didn't get many people coming back wanting a refund" blimey unthrottled sounds worrying ,

Edited by injection doc on 26/04/2012 at 10:59

In the interests of cost, and with better casting and metallurgy, cylinder heads have become lighter. And in the interests of rapid warm up, cooling systems have lesser volumes.

I agree entirely. But one of the biggest advantages modern cylinder heads have over their old ones comes from advancved CFD allowing a good understanding of heat flow in the head and coolant. Cylinder heads have much more even cooling than they did in the old days, so they don't need as much volume of water and are less prone to warping.

How many systems today canbe easily refilled if they run low by a litre of coolant?

Probably most! Any air locks tend to form in the primary cooling circuit, not the secondary one, so topping up should not be a problem.

A few observations based on a few years of my career spent in engine testing:

If an engine has a weakness identified in development it will be 10x worse in service where the real world of skimped maintenance and abuse applies. Production assembly may not be as precise as that of development prototypes.

All of the engines I have tested which have had cooling system problems were lacking an essential feature - degas pipes to the header tank. If an engine has bleed screws fitted to the primary circuit of its cooling system it is liable to failure because it will not be self-venting. A bleed screw for the heater matrix is acceptable on most installations because an air lock here does not usually jeopardise the cooling efficiency of the engine.

On a vehicle installation, degas pipes can be recognised by their narrow bore and their routing to the air space at the top of the header tank. All engines will develop air/vapour locks due to compression leakage, water pump cavitation and micro-leaks as well as that caused by dissolved gases in the initial fill. The lack of a degas route will inevitably cause trouble.

The engine under discussion is notorious for these problems and I found the development expertise of the team responsible for it to be sadly lacking.

Only someone with an unreasonable and unwarranted desire to own this engine would ever dream of buying one. There are plenty of well engineered alternatives from other makers.

659.

Edited by 659FBE on 27/04/2012 at 16:22

When the engine is turned off when hot and allowed to cool, the temperature spikes in the head (due to the low coolant capacity offering an insufficient heatsink capability)

I've been thinking about this. When the engine is switched off, the temperature of the coolant in the head will indeed rise. But the temperature rise is limited by localised boiling. The heat of vaporisation of water is vast-the vaporiasation of water requires abut five times as much energy as simply heating the liquid from 0 to 100C. If the water in the head is boiling, then the stat must be open so the vapor should be able to escape and be replaced by more water, no?

In engines with marginal cooling capacity, the problem can often be solved by running virtually straight water with only a hint of antifreeze. In summer, I run an 80/20 water/antifreeze mix in my engine for this very reason. This allows me to tape over about 3/4 of the front grille with zero cooling problems. In winter, I reverse the ratio-and that's the only time I experience high coolant temperatures. All the organic coolants have cooling properties that are grossly inferior to plain old tap water!

Or expansion bottle relief) opens, whilst the radiator cap remains shut any steam formed will be at the system pressure and circulation will be limited at best, this will tend to excarbate local differences in temperature leading to the causes of CHG failure discussed above.

At ~2000KJ/Kg for latent heat of vaporisation you don't need to move much water to cool the hot part of the head, since the specific heat capacity for aluminium is only around 0.9KJ/Kg/K.

The exhaust ports are largely cooled by nucleate boiling when the engine is running since they are being constantly blasted by 800C+ exhaust gas temperatures, so the temperature gradient of the aluminium from the gas side to the water gallery side is quite high.

As for thermal expansion, the cylinder head cycles through ~80-100 degrees every time you start from cold, yet it survives!

Shock cooling? Happens every time you grab a downshift, takes a few seconds for the 'stat to close after the temperature of the water in the head drops. Again, the head survives. I just don't see how shutting down hot is going to put undue stress on the head that it wouldn't experience during normal operation.