Fisher Services Ltd.

MGF (K Series) Head Gasket Developments

(Credit: MG Owners Club website)                                                                                                                                                                                   Return to MGF Head Gasket page

"I’m thinking about buying an MG TF but friends have said that they always are blowing head gaskets and engines and I should steer clear?’ This is a not uncommon question from non-members and surprisingly members too. Surprising because many members adhere to that simplistic and inaccurate general view, when the facts are far more complex.

For the benefit of those with little experience or contact with as many MGF, TFs or K series engines as we, have let me run through some of the Club’s experiences. MGFs made during the first three years production have had a dreadful rate of failures, easing noticeably from 1999, and then improving further in 2001 for the last few months of MGF production. Overall I can account for direct communication and conversations with over 1000 MGF owners whose cars have suffered head gasket failure (HGF), most of these pre 1999 T registration in the UK cars. Total production of the MGF was 77,269 cars.


Payen head saver shim - a used one in this image


The TF arrived in January 2002 and ran until April 2005, although a few examples unfinished and trapped on the production lines were finished up to June 2005, and of course we await restart of TF production during 2007. During this production period 39,249 examples were made and to date (August 2006) I have recorded just 44 failures through the same method of communications and conversations. Not only is that a very small number and considerably different to the MGF failure rate, but significantly 43 of those 44 failures are cars made before the summer of 2003, a point of significance, as I shall explain later.

Those plain observations tell a quite different story to the bland ‘all K series engines blow head gaskets’ view. In addition there is also a common statement that the K series head gasket is ‘rubbish’ and like the other bland statement this is far from true. There are reasons why there is such variation which will become clearer as we negotiate the engine changes.

The K series was conceived with a maximum displacement of 1400cc, but through necessity and clever engineering this was raised to 1600 and 1800 cc. Indeed 1900 and 2000 litre capacities have been around for a while now in competition and one off modified form, but not as a factory produced engine. These larger capacities do add stress, both mechanical and heat and partly why it is the larger capacity engines that suffer the greater proportion of HGFs.

The K series has a sandwich construction, as I have mentioned in previous features, with very long through bolts going from the top to the bottom of the engine and clamping it together using the stretch properties of these bolts. The engine construction is mainly alloy with iron liners. The liners and engine castings do have a different expansion rate and so the design takes this into account and is why the head gasket is actually a very clever design to accommodate these variable expansion rates, yet retain a fluid seal. Well most do but as can be seen from the failure pattern some do not!

There are two common points of failure of the standard K series gasket and both involve displacement of the sealing bead leading to coolant loss, one externally and the other internally. External loss is seen either with coolant dripping under the car or with plenty of steam from evaporation on the exhaust. Internal loss is mostly hidden and can lead to a severe overheating problem and dilution of oil before it is noticed. This type of failure is usually the one that leads to serious engine damage.

The reasons for this displacement are not simple and DO NOT just relate to the gluing of the sealing beads to the stainless steel shim section of the gasket, but a combination of conditions all contributing to the failure.

Firstly the bead displacement is almost always on the exhaust side and towards the front of the engine. This is of course the hottest part of the engine short of being in the combustion chamber or exhaust port. My understanding is that the critical temperature for the bead material is around 250 deg C and that independent testing of the temps around the edge of the cylinder head can reach temperatures over 200 deg C which are uncomfortably close to that critical temp. Note that these are localised temps and not average temps and these have no direct connection to the coolant temp in an engine operating normally. It follows then that when an engine suffers a problem within the cooling system or loss of coolant the ability to develop localised overheating can allow that specific area to reach even higher temps and this can damage the gasket bead.

Immediate failure of the seal is not always experienced except in severe overheating cases. Often the damage is done and the seal remains functional for a varying period. It is also possible to have a coolant loss at that time but when refilled and bled there appears to be no problem, often for considerable periods, until there is a sudden return of coolant loss and other associated HGF tell tale signs.

The reason for this variable condition is again related to other factors. Firstly the gasket is under a degree of compression so the sealing beads are being squeezed all the time. The degree of squeeze varies with temperature changes and different expansion rates within the engine. The iron liners are always fitted to be proud of the top of the block by between two and four thousands of an inch, so that the compression rings of the gasket have a good clamping load all the time and that causes the variation that the sealing beads need to be flexible enough to cater for and retain seal. It does appear that this liner protrusion may actually have been greater in the 1996 to 1998 period and this may point to one reason why this was also the worst failure rate period with better returns from 1999 on engines.

The next main influence comes from head to block movement, known as ‘head shuffle’ and this is further influenced by the individual and variable maturing stresses that have been present in the engine since new. Interestingly the full clamping load of the engine construction is not seen until the engine has undergone a full thermal cycle of reaching normal working temp and cooled off again. With new engines there is also the maturing process that adds this variable additional internal stress and perhaps this provides a clue why so many engines have quite early failures within the 20 to 30 thousand-mile band, but repeat failures are very uncommon for considerably higher mileages.


K series head gasket 2001 to 2004, then resurfaced in later 2005 to date when bought in an MG Rover box. Note also the same part number as the reinforced 2004 and early 2005 gasket image


MG Rover supplied K Series head gasket 2004 into 2005 note part number identical to earlier, but different in detail gasket


Image taken in 2001, on the right two MG Rover gaskets, top - pattern used 2001 - 2004. On the Left is a standard Payen aftermarket replacement - note small sealing track differences


This head shuffle condition is another clue to the failure rate changes that saw a marked further reduction in engines made from the spring of 2001. Here the change is one that is the simplest and easiest retrospective change for all engines and which is a standard additional part with every head gasket supplier I know on the market. It is, as many know, the steel dowels that replaced the nylon dowels used up to that point. The rigidity of the steel dowel is obviously better than the nylon and this single move has had a significant impact on further reducing the possibility of repeat failures through reducing the amount of head shuffle that can occur.

Such a simple change without further explanation brings with it loud criticism on the basis of, ‘that’s obvious, why didn’t they use steel before?’ The reasons are mainly historically based and come from huge problems of earlier alloy heads with steel dowels corroding into the alloy and being difficult and sometimes impossible to remove without damage to the head. Changing to nylon for these engines solved the problem with no side effects. Changing to steel for the K series will in future years see that corrosion issue return, but on balance it is a problem worth putting up with for the benefits gained in normal use. I have for some time used a simple approach to providing future proofing for easier removal of these dowels by running an 8mm tap through them so a M6 bolt, nut and a few flat washers make a simple but effective extractor.

The introduction dates for the changes listed so far also coincides with the step changes in the reduction of HGF rates with the MGF and now we move into the MG TF period. The TF started with a revised airflow through the engine bay leading to 20% lower average temps, a useful help, but otherwise the engine was much the same as the last MGF.

The next step change occurred in the summer of 2003 for the 2004 model year cars and involved the move of the thermostat from the original position at the inlet back into the engine, to much closer to the outlet, and thus become more responsive to coolant temperature changes. This is known as a Pressure Relief Thermostat (PRT) and is an apple sized plastic component with three full sized hose connections that sits just in front to the block to sump flange on the exhaust side of the engine.

The PRT provides several advantages and these are seen in quicker response to coolant change coming from the engine, greater smoothing control of coolant temperatures when the thermostat operates and improved pump performance by using pressure differentials to advantage, rather than disadvantage.

About the same time was a small change in the detail make up of the gasket with a series of additional pieces of sealing bead, which were clearly meant to act as strengthening buffers. (Image) Oddly this change did not see any change in the gasket part number and following the demise of MG Rover the replacement gaskets supplied have returned to the earlier design.

The final change of note was during 2004 when the header tank adopted an integral coolant level sensor and this was connected to a warning lamp on the instruments.

Once again it is interesting to see correlation of the PRT changes with a change in the HGF rates of the TF and why I see the use of the PRT as a fundamental advantage for earlier cars just as the adoption of steel location dowels has been. However, costs, complications and currently parts supply delays means it is not currently viable to rush out and order all the parts needed, and yes we are looking at creating a kit.

Having reached this point there may be a feeling that all is rosy and we can all drive off into the sunset, but whilst the rate of failures has been drastically cut, to what should have been a normal rate from the start incidentally, there is still further room for improvement as can be seen from my but localised, operating temps and the apparent gasket bead material ceiling temps. In addition these localised hot spots occasionally open up the door to another problem.

This additional issue comes in the form of softening areas of the cylinder head face almost always showing up in the form of clear grooves in the head face made by the head gasket’s fire rings and due to the compression between the head and liners. Once again it is almost always concentrated on the exhaust side of the head and has in the past spelt the need for a replacement head.

Payen, the gasket maker and OE supplier, make essentially a head gasket without the silicone beads and fire rings in a plain stainless shim and call it their ‘Head Saver Shim’. It requires the damaged head to be skimmed and then the shim is bonded to the head face and it is refitted to the engine using a standard new gasket. The shim is much denser and harder than the alloy and is able to spread the spot loads seen around the liner fire ring seal and so is able to return damaged heads to serviceability. It is also possible to have damaged head refaced (within limits) and the surface hardened, although this will usually be more costly than the shim route.


Common failure point for the original gasket design when oil and water mix inside the engine


Close up of the Land Rover head gasket showing the upper two secondary shims and also the lack of vulnerable strips of sealing material as found on the original design of K series gaskets


Lower rails with Land Rover lower. Looking at this and the image showing the other side of these rails the new version’s strengthening is clearly visible


Having now run through the common aspects of HGF as it applies to MGF, MG TF and by inference other K series engined cars as well, it is now interesting to report on continuing developments and a wider experience with the latest developments, which have been briefly mentioned in these pages within recent months. I do now refer to the new K series head gasket being supplied by Land Rover for all their Freelander models when a gasket change is done. Whilst land Rover have interest only in their Freelander, which uses only the 1.8 litre engine in four cylinder format, the commonality of all four cylinder K series engines from 1.1 to 1.8 litre means that the Land Rover parts are suitable for all. MG wise that means for MGF and TF only 1.8 litres, but ZR 105 uses a 1.4 litre and ZS 110 uses a 1.6 litre engine, with both these ranges using 1.8 litre as does the ZT, and I also include the VVC and 1.8T engines in the suitable range.

The gasket is part of a three point approach to what Land Rover sources say will be a solution for the Freelander’s also dreadful rate of HGF. The inference therefore carries over towards other four cylinder K series engines although Land Rover will be quick to distance themselves from giving any such support or comment. The reality is that with common engines in 1.8 litre form there is no reason to doubt it.

The three parts to their solution is the multi layered gasket as shown in the image, a separate thin stainless steel shim, one side of which is pre-treated with a bonding agent to stick to the head face, and thirdly a new lower engine rail that fits to the underside of the engine and which is what the ten long through bolts (usually and incorrectly called head bolts) screw into.


Lower rail before fitting, Land Rover uprated above standard


Land Rover K series head gasket - the contents of the box - bottom is the multi shim gasket, top is the additional shim with bonding agent side shown, and to the left the two standard type steel dowels


Sump removal needed to fit new lower rail, fortunately reasonably straight forward except for 160 models with oil cooler


New rail in place showing refit of oil pick up before sump replacement. Note the threads where the very long through bolts engage to hold the engine sandwich together


The gasket is fitted in a normal way and not only is made up of five separate layers, one main layer and two subordinate on top and two below, it has a different sealing arrangement so that there is no vulnerable sealing track to be displaced. Then there is a separate shim that sticks to the head face using the bonding agent which is activated when the engine warms up.

There is an interesting suggestion that carefully locating this shim onto the head face and evenly preheating it to activate the bonding agent, before fitting back on the engine, will remove any potential for coolant to permeate between the head face and shim during the rebuilding of the engine before it is first run to normal working temperature to otherwise activate the bonding agent. I would be interested in any comment on that front from anyone with better contacts within Land Rover than I have.

The last element is the new lower rail that is clearly very different and obviously stiffer than the original. It is also 20% heavier and some of this is due to the additional strengthening webs visible. Quite how this affects the overall stiffness of the engine is still difficult to judge when the main through bolts are still the same spec bolts that have always been used along with the same tightening torque and procedure.

Since this gasket arrived late in 2005 there have been a number of member’s cars fitted by MGOC Workshop and I have used them as well, most recently in a new build 1900cc significantly uprated engine. Whilst I believe that these gaskets offer a clear benefit to the K series engine and consider that Land Rover, who have been stung by the HGF issues with Freelander, would not be pursuing this route if it were not proven in their eyes to give an acceptable solution. The definitive proof though is very likely not going to be around for another five years when enough longer term experience will provide the answer. In the meantime look at the images and decide for yourself. If you want to go down this route then the Land Rover part numbers are LVB500190 for the head gasket and LCN000140L for the lower rail, priced at £23 and £29 respectively retail and of course you can always ask MGOC Spares to supply these."

Roger Parker from MG Owners club October 2006 Issue.