To be honest - I am amazed and disappointed at the way this post has deteriorated especially since the 15th when I last posted. I only posted the findings because we are continuing our own development and it was several years ago we tested the first engines fitted with temperature sensors and the bore scoring has become an issue since and I wanted to re-visit the problem and understand it more in the light of more failures and a better understanding internally.
I also got the impression (rightly or wrongly) that the third radiator was being proposed as the ultimate answer whereas the LTT was much more significant in my view and I simply wanted to express my opinion on that. I was also careful (if you read it all or follow it) to express appreciation for GT4's contributions and play down the mistake people easily make about engine cooling.
However I must still take issue with a few factual points.
(1) It is wrong to say no one else has raised the issues of unbalanced cylinder block cooling between M96 and M97 engines as our Internet Buyers Guide sections 4 and 5 goes into great detail including photos and has been around for almost a decade.
(2) I am not sure exactly when or why a piston scores the cylinder and as it takes place in 2 or 3 seconds during somewhere usually between 20 and 60 K (for a small number of owners). I am not sure how anyone could possibly pin point the exact moment it occurs and therefore I very much doubt anyone will be able to prove it or not - especially as the first signs are not noticeable anyway for several thousand miles.
I think this means the best we can do is to use all the powers of intellect, experience, statistical analysis, common sense, engineering knowledge and YES debate and argument - to see if some answers point in one direction or another.
(3) You are wrong to say there are only 2 ways to increase cooling - the third is increasing coolant flow speed and it is this that most systems use via the thermostat and have done for years (although there is a limit to this of the basic capacity of the system to cool the engine).
(4) I am not sure if a "third radiator is the best answer if the engine usually runs at over 95 degrees" because we do not know for sure what circumstances result in bore scoring but it doesn't result in scoring in bank 1.
(5) I am not sure why it is my fault that I "took two and a half years to update GT4's findings". There is some hypocracy if someone that explains "he is not a free consultancy" in this thread expects me to monitor and test every ones theory and provide an answer within a short space of time.
(6) I do not have access to figures about numbers of models sold or affected - nor those already repaired here and do not have time to research this when I am also trying to do the manufacturers job in finding and testing solutions. I do have an impression of those answers from the very large numbers we repair weekly.
Now a few other points
(7) Well done Mister Corn a concise and clear brief explanation of the main issues.
(8) We only started building our first race car 2 years ago and it was a major project without any prior experience. We fitted brake cooling air scoops inside the air inlets for the radiators and this reduced our radiator intake area by about a quarter and so fitted a third radiator to all the cars partly to compensate and on the same assumption that GT4 made that if an X51 etc needed it so would we. However we found out too late that a replacement front spoiler fitted after a "racing incident" didn't have the centre intake cut out - but to our surprise it made no difference. Given a few seconds of thought I too would have expected a 3rd radiator to offer more cooling by way of a larger temperature drop inside the engine (which it does). I didn't even consider that with the thermostat on the inlet this means it would raise the exit temperature - but then racing engines never idle and then get high torque at low revs in racing - so the conditions may not ever be met.
(9) When I was struggling to get our 996 to handle I informed those interested (on the three main internet sites I communicate with) and when I realised I had made a stupid schoolboy error (in just one cell of a 10,000 cell spreadsheet I had used successfully for the Boxsters) that resulted in the rear suspension being too light) - I admitted it. I am not bothered about being wrong and only seek the truth but this problem is a really serious one for many owners and we owe it to them to try and find the best solutions possible whoever it may upset or implicate.
(10) All I can do to find a solution to bore scoring is to look at all the evidence and see what links may coincide with good engineering practice and technical explanations and if possible test out the theories.
Let's review things here once more.
(a) Only bank 2 is affected (by any standards of statistics).
(b) It is odd that only one side of the piston scores while the other side does not and this can only be explained by lack of oil, or too high a piston face temperature. This is generally described Internationally as "piston scuffing". We tried doubling the oil spray jets and directing them exactly on the scoring position and it made no difference so I am discounting this suggestion.
(c) We looked at the system to see if there was a difference in the temperature of the thrust face on bank 1 and bank 2 and found that bank 2 would always run hotter due to the opposite coolant flow and that bank 1 had classic "contra flow" (which technically increases overall cooling rates) but bank 2 did not. Bank 2 also receives the hotter coolant from the oil cooler position. The Tiptronic gearboxes may result in more temperature for the system to cope with than the third radiator cools - I have not tested this yet.
(d) We researched "piston scuffing" and the only engines we could find that suffered from similar problems did so through too high a piston face load at low revs and high torque. The M97 engines produce higher torque at low revs (exacerbated by the longer stroke combined with the same crankcase block height which increases the rod angle).
(e) We think we see more tiptronics with bore scoring than manuals compared to the numbers in circulation but have no evidence to prove this.
(f) We notice that racing engines or customers who regularly speed round long circuits like Spa and Nurbergring do not seem to suffer scuffing neither did our race cars.
(g) We notice that a typical failure is often driven by average modest drivers who does not race or dramatically exceed speed limits but - like most of us - do enjoy pulling away smartly up to the sort of speeds everyone else is driving at - which creates our limits.
(h) When a car is driven at a constant speed (even if that is higher than 70 mph) the temperature rise in the system between the coolant entering the engine and leaving it is usually only around 2 to 4 degrees - because the throttle opening at those speeds is only relatively small and hence the heat generated is relatively low and the system can easily cope with it. The only time a lot more heat is put into the system is when the throttle is fully (or almost fully) open for longer repeated periods.
This cannot be sustained indefinitely as the car quickly reaches 170mph and stays there (unless the drive is on some back road which is similar to a race track - which is the other scenario where a lot of full throttle opening occurs). In our experience - on a track - unless the car is stuck right up the bumper of a car in front - the car is going fast enough for the airflow to still cool the engine sufficiently - but at lower top speeds (like an owner driving fast on a Country route with lots of corners on a hot day) I can see that the conditions may be met to overcome the capacity of the system - especially if the air con is on full but we have not managed to reproduce this although we have tried.
(i) I find it hard to believe that the small temperature difference between bank 1 and bank 2 in normal running explains why all the failures are on bank 2. Furthermore it is clear that even though some cars are running at over 100 degrees sometimes - bank 1 still generally survives OK - so there must be some other circumstances in which the temperatures at the thrust face of bank 2 exceed those by a much higher margin. It was to re-test this in the later engines that we recently built one with another set of sensors inside the engine to test and record the temperatures where we have interest and we found (as I previously theorised) that the highest temperatures by far were reached during a heat soak in bank 2 while the engine was switched off or ticking over for a short while following some spirited driving. This creates condition where the reduced coolant flow and air flow combined with the heat soaking back to the coolant (from driving faster just before that) raise the temperature on bank 2 thrust face much more than bank 1 and when you set off they cool much more quickly in bank 1 than bank 2. Combining this with the theory of low revs high torque causing scuffing - it seems reasonable to me to conclude that this is the most likely scenario to cause bore scoring - but I cannot prove it.
(j) I also have considered that when the engine has stopped the oil film on bank 1 and 2 runs down through gravity to the bottom of the bore which is the thrust face on bank 1 but leaves the thrust face short of oil on bank 2. However our tests of additional oil spray jets seem to negate this. However as bore go oval the clearance between the piston and the thrust face at rest on bank2 will be higher preventing the piston cooling there through contact with the oil and through that to the cylinder wall whereas in bank 1 the thrust face sits on the bottom of the cylinder where it is the coolest and this could also lead to much higher piston face temperatures and very little very thin oil for a short while.
(k) The pistons that suffer have a new coating (yes also on bank one I agree) but it seems that the thin coating wears or detaches in small pieces more quickly on bank two than bank one. I have previously made an analogy between this and a coated frying pan because the coating on them is strong and hard for many heat cycles but then becomes easy to pick off in lumps as the bond between the coating and the metal weakens and the coating softens. I imagine this would speed up the failure if the temperature the pan is exposed to is high. Similarly I think that over time excessive temperatures could damage the bond between the piston coating and the piston face leading to the premature peeling or wear on bank 2 pistons.
(l) We think the silicon particles are not as secure in the Lokasil composite as they are in Alusil and may degrade quicker and that they may be bigger than the space left between the hot piston and the thin oil film and cylinder wall causing them to impinge on the coating too hard under heavy acceleration on bank 2.
(m) We know that a ferrous coating was perfectly OK on less torquey versions and has a harder surface hardness than the various plastic coatings we have tried but this doesn't explain why they seem OK on bank 1.
My conclusion from all this is that despite various changes making some of the components less likely to last as long - bank 1 still has an acceptable lifespan but bank 2 does not - so there is something going on at bank 2 that is different to bank 1 that is significant.
(o) The only significant difference we can trace (and have proven) is that bank 2 runs a little hotter on the scoring face in normal driving and a lot hotter during idling or after a short switch off. This is not picked up on the normal temperature gauge because there is only about 6 or 7% of the coolant going through bank 2 cylinder block and the temperature rise when mixed is insignificant.
(p) it seems to me that if we could get the thrust face of bank 2 at the top of the internal cylinder block run at the same temperature bank 1 normally does at the thrust face on the bottom - then even if ambient or particular driving conditions raised the overall temperature of the coolant - the engines would probably survive for an acceptable mileage.
(q) Unfortunately the only easy, quick and inexpensive solution I could think of is the LTT because it does lower the running temperature on bank 2 significantly. However this has a down side because although it does lower that temperature (and that of the oil) it also uses up something like 93 or 94% of the cooling capacity while it is lowering the temperature of bank 1 cylinder block and bank 1 and 2 cylinder heads lower than they need to be because that did not cause a problem to those areas at the original temperature and this runs the risk of reaching the overall capacity limit sooner.
(r) To solve this we are presently experimenting with a solution to this that should work theoretically but has yet to be proven. We were spot on with our theory about the effect of fitting an over capacity radiator to an engine with an entry thermostat and I cannot envisage any reason why the "new solution" should not work although it could suffer from "hunting" if I cannot get it balanced as it is a more complex control system.
If I can get this to work it could release enough capacity to negate the need for a third radiator but we will not know for sure until a lot of different drivers try it out. Just in case extra cooling capacity is still needed there is another component in the plan to allow that to be used without resulting in higher temperature rises inside the engine if the overall capacity still proves inadequate for some drivers in circumstances we have failed to reproduce ourselves.
Now some general personal points.
In the last two and a half years we have built our first race cars, raced 4 in the team event and supported them at the track to beat all comers who have decades of experience. We have developed several solutions to all alternative IMS failures and designed and proven a new version to fit the larger bearing to all earlier types. We have tested 6 different piston coatings all in different engines involving test mileages, stripping, analysing etc to try and find a way to re-bore scored bores successfully (reducing rebuild costs).
We have built our own rolling road test bay. We have installed computerised machinery to reduce production rebuild costs (and even had to design and manufacture our own tooling). We have identified the cause of a mysterious ticking noise that has afflicted perfectly sound engines (and mislead a lot of people to assume they had piston slap through bore scoring Worldwide) - which was a very poorly designed original component - and found a way to improve the situation.
Unfortunately the component in question would cost too much to completely re-design for the small numbers that have this symptom. I don't think any other specialist comes close to this commitment.
The issues of the Internet have both a good and bad side. I am not sure I would have thought out the consequences of the third radiator if I had not read about it being such an important solution and wanted to check it out and I am not even sure if I would have got round to working out this "new solution" if I had not followed that thought process to a new idea - so in many ways the exposure to public scrutiny has been a good thing.
Often I notice that if I challenge someone on their post over a technical issue I know they are wrong about and misleading readers over - they ignore responding (because they realise they are wrong) but come back with a completely different slant on things to give the impression they know what they are talking about. There are also those that trawl through every little piece of minutia (probably written by me in haste) that I add into more important discussions (trying desperately to find a weakness they can expose) and then challenge me over them to provide answers and responses hoping they will discredit my position and elevate their own.
I have made loads of mistakes in my lifetime and know I will again. I am not sure if the "new solution" will work or backfire (and I am waking up early in the morning going over the sequences of the "new solution" in my head to try and be sure they are right - but they are so complicated I resign myself to the need to test it out and see) but I don't try and argue my way out of it if I have made a mistake - I am only interested in reality and the truth. At an age when most engineers are retired - I also don't care what people think of me personally - but I do care about all the younger people associated with and working with us at Hartech, who have many working years left in them and the image we project and the respect with which we are jointly held.
We spend massively more time, energy and finances than anyone else (and it seems more than what was one of the wealthiest manufacturers in recent times who should be doing so) trying to solve the various problems theM96 and M97 engines encompass partly for our own benefit but also very much for the benefit owners and enthusiasts Worldwide who have been let down by the product failing to live up to expectations in a devastatingly expensive way - and I don't think we have done anything to deserve any criticism over this issue. I also think we were right to question the conclusion some others have reached about the benefits of the third radiator in the present system, and to publicise our views having proven them .
I am really sorry if in doing so we have upset anyone (especially if they are usually valuable contributors) but sometimes there are more important technical issues to expose than personal feelings - as I think in this case.
Although in volume production the "new solution" would be relatively inexpensive to produce - to manufacture test prototypes out of solid material and then build and test the results is a very time consuming process and I think it would be more in readers interests to allow us the time to do so and report back on the results than bicker over small understandable errors some contributors may have made.
The problem is that if I ignore the nonsense some others contribute it appears to those unable to work out the truth that they are right and that I am keeping quiet to "get away with my mistakes" - it kind of locks me into needing to respond even though I don't want to waste time doing so (and many will have noticed my posts are done in my own time at home rather than delay important work at the factory).
In a nutshell - having set up and managed to rebuild more reliable engines and solve more problems than anyone else in recent years it would be nice to receive a little more respect for the work we do - so we can get on with working on the solutions that are more important than Internet chatter rather than divert too much valuable time arguing about things that we are going to test and prove or perhaps disprove anyway!
Baz