When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Well, I'm late to the game here and it took me a while to get through the whole thread.
To the OP, I'm very sorry for your luck here, but hope in the end you'll be extremely happy with the new build.
So, to ask a couple of "Selfish" questions:
1) Are All 9A1's the same?? Meaning, does my 2012 Turbo S have the same 9A1 but simply a turbo'd version of the OP's car?? Or did Porsche do something different for the Turbos in the 997.2's?
2) With almost 36k miles on the clock, should I immediately swap to the DI40 Oil?? I have no problem doing this, as I'm up for a change soon.
I'm clear on proper "Warm up" of these cars , as I've always done that for Every car / truck I've ever owned. Why would anyone NOT properly warm up and engine before beating on it? SMH.
I think perhaps I should apologise but explain about my long replies.
This is going to be typical so those that don’t like it don’t read on. Being a qualified professional engineer that has designed and manufactured the parts for complete engines and transmissions for over 50 years, I perhaps come to a problem from a different perspective than most others.
When I was involved in motorcycle racing engines in the early ‘70’s top mechanics ran largely Japanese race bikes and had a lot of reliability problems – so to sort it they used analysis – yes – but then (often only) trial and error fitting other parts or changing things – more often than not without success. But because all the top mechanics all over the World were doing it - eventually one would find and improvement and as soon as the word got around – so did everyone else.
Due to the numbers involved, this was often a quicker way to solve a problem than trying to analyse it from scratch properly. However because I had both the basic knowledge base and manufacturing machinery to make anything - I wanted to make something new to solve the problem and so had to understand the issues from more of an engineering point of view because I couldn’t make lots of different parts to try – I had to try and work the problem out and make something right first time. I probably spent longer thinking about the cause but got the right answer sooner (or at least understood it).
Even then I didn’t always get it right first time but it enabled alternatives to be assessed quickly and overall it was successful. Now years later when I examine a Porsche engine problem my first approach is to try and work out why Porsche made it that way, then think about what could influence it and try and link that to the evidence we see and that is reported by others Worldwide. I did (for example) immediately understand why the M96/7 crankcase design was how it was and knew I had to machine out and replace the cracked liner and could fit a simple ring in the top to stabiise the others (something Porsche could have done at minute extra cost during production).
But I didn’t understand the cause of bore scoring in later models for a very long time. I had discussions with KS engineers (who made them)which helped, was slow to link Lokasil (because the KS manual described as the same as Alusil which it isn’t), was slow to link the piston coating to the failures, spent money on a solution that didn’t work and towards the end had pistons coated with a lot of different coatings – fitting different ones in the same engines, ran them, stripped them and analysed the result.
It was the DLC coated piston that revealed the answer with tiny shiny scratch marks clearly visible and from this I researched Lokasil more and found that it did not retain the small silicon particles as well long term as Alusil (or distribute them as evenly) and that soft piston coatings would eventually be worn out by it or flake off exposing aluminium to the particles – which would damage the surfaces beyond repair.
Fortunately throughout this process we were able to provide a perfect solution by fitting the same type of new nikasil cylinders we had for the cracked earlier engines and since Nikasil runs with any piston coating our repairs gained a good reputation. However the poor plastic coatings we tried failed quickly but gave us an opportunity to test other issues and read out the consequences after relatively few miles driving.
We used this process to test other contributory factors. The puzzle was why mainly bank 2. We wondered if the oil from the spray jets fell by gravity to the bottom of the bores and since bank 1 thrust face (the loaded piston face when driving) was at the bottom but bank 2 was at the top – we fitted extra oil jets upwards but it made no difference. We pondered about the piston pin offset issue so experimented with changing it – but it made no difference (and anyway the engines most prone have them the right was round).
We noticed that the cylinder block had very shallow coolant channels and that only about 10% of the coolant was fed through the cylinder block and fitted temperature sensors to reveal differences between bank 1 and 2 and of course that although the coolant entered both banks at the bottom it got hotter towards the top (especially when stopped during a hot run say at lights). With the thrust face on bank 1 being the bottom but the top of bank 2 we realized that the oil on top of bank 2 would not only drain away from the thrust face but that the oil left would get hotter and thinner – whereas the opposite was true of bank 1.
This was then a contributory factor. Having the thermostat at entry to the engine also resulted in a higher temperature gradient bottom to top (especially at colder ambient conditions) so we fitted lower temp thermostats and altered the coolant flow in the block.
On our racing engine we split the coolant between banks and fitted different rated thermostats each side to equalise the thrust face temperatures. Our replacement cylinders had external ribs( to increase cooling rates) and stabilized the cylinders by a top closed deck precision location.
While testing a lot of different piston coatings we had some coated with ferrotec (by Mahle) which had replaced Ferrostan (the original hard iron coating that had been outlawed). Whereas the original was magnetic the new replacement didn’t seem to be and was obviously better than a plastic coating but we found not as hard as the older coatings.
The introduction of the Alusil based 9A1 engines could not be linked with the original (now outlawed) hard iron coated pistons that rendered the 944 and 968’s so robust so we wondered if the newer Ferrotec coatings would last as well – which by and large they seem to have done. We also noticed that Porsche had g one back to a closed deck (as our cylinders create) and deepened the coolant channels.
When we received a few failures we analysed and measured many things. We found the pistons had different profiles to support the loads over a larger area (which we think might be to compensate for the possibly slightly less durable coating). But we also noticed that these were 2 sided seizures (not one sided M96/7 scores) and found the crankcases we had measured had shrunk in on some cylinders to pinch the already tight pistons especially on fast warm up on a cold day.
Other engines with different failures had not shrunk the bores. On analyzing the shrinkage we did see - the adjacent large section main bearing part of the casting was exactly where they had pinched in and we assumed from that it was caused by slow and gradual stress relieving as a result of hot cold cycles and therefore inevitably would only reveal itself after several years – which also fitted the present scenario.
So when someone asks “what cause bore scoring” you can hopefully see why a long answer is unavoidable and why without it owners could be mislead by different explanations. Bore scoring in M96/7 is caused by surface peaks of silicon particles wearing out the piston coating, loose silicon particles impacting on the piston coating and the piston coating flaking off all together. The pistons run hot and the oil film protecting them is often compromised but change the bore to Nikasil – problem solved. 9A1 seizures we have seen are caused by aged stress relieving and too fast warm up rates.
There are other issues with the M96/7 engines that could be improved upon and differences between the UK and the USA that can introduce different failure causes in both that model and the 9A1 engines requiring different changes.
Other solutions can also work and other changes might suit some locations more than others. I agree totally that video is better than words but I could not find time to pre-make videos to cover every question people ask and I have to be thorough in my written answers to avoid lengthy follow-on responses that are white elephants. You can already see from Jakes responses how difficult it has become (and soul destroying it can be) to explain things but without someone trying - the public are more at risk from innocently based or deliberate exploitation.
If it benefits some readers to receive quality well based professional explanations to the questions they raise – it would be much easier if those that cannot cope with long answers simply don’t read them rather than to irritate and disillusion the very people taking a huge amount of their valuable time trying to help by criticising them.
The pistons run hot and the oil film protecting them is often compromised but change the bore to Nikasil – problem solved. 9A1 seizures we have seen are caused by aged stress relieving and too fast warm up rates.
1) Are All 9A1's the same?? Meaning, does my 2012 Turbo S have the same 9A1 but simply a turbo'd version of the OP's car?? Or did Porsche do something different for the Turbos in the 997.2's? And change again for 991's?
If it benefits some readers to receive quality well based professional explanations to the questions they raise – it would be much easier if those that cannot cope with long answers simply don’t read them rather than to irritate and disillusion the very people taking a huge amount of their valuable time trying to help by criticising them.
I think perhaps I should apologise but explain about my long replies.
No need to apologise for anything as you provide a free, valuable consultant services for this forum. Same for Jake. We own you guys big time.
It was the DLC coated piston that revealed the answer with tiny shiny scratch marks clearly visible and from this I researched Lokasil more and found that it did not retain the small silicon particles as well long term as Alusil (or distribute them as evenly) and that soft piston coatings would eventually be worn out by it or flake off exposing aluminium to the particles – which would damage the surfaces beyond repair.
WHAT’S THE COATING ON A PISTON?
There are several different types of piston coatings, but the one that most people are familiar with is called a “break-in coating.”
This Teflon-like coating covers the skirt area of the piston to reduce friction and ease the break-in period. It will normally wear off after one race. CAN PISTONS BE HARD-ANODIZED?
Yes. There are coatings for the piston domes that provide an additional barrier for heat dissipation and reduce accumulated burnt carbon.
Additionally, the ring land and pin area can be hard-anodized to prevent the hard metal rings from eating into the softer aluminum.
Piston skirts cannot be hard-anodized. The hard coating with the most promise is a DLC (diamond-like coating) on the wrist pin.
The wrist pin is very busy, and reducing friction and wear pays dividends, but the DLC costs more than the pin itself. WHY AREN’T THE SKIRTS HARDENED?
Piston rock is to blame. A hardened skirt wouldn’t have enough give to take the constant rocking.
If the skirt of the piston is too hard, it will rip apart the Nikasil liner.
Humm, "If the skirt of the piston is too hard, it will rip apart the Nikasil liner." This is interesting info that I did not know. Can you comment on this?
If it benefits some readers to receive quality well based professional explanations to the questions they raise – it would be much easier if those that cannot cope with long answers simply don’t read them rather than to irritate and disillusion the very people taking a huge amount of their valuable time trying to help by criticising them.
Baz
Well said. I think there is a term for that, "troll" and there is a solution for it, "ignore the troll".
The only engines with alloy Nikasil closed deck liners in the M96/7 range were the GT3 and Turbo models (where they could afford the extra quality through higher retail pricess).
The 9A1 Gen 2 3.6 and 3.8 N/A engines are Alusil. Agreeing that a picture is worth a thousand words there are lots of pictures and descriptions in our buyers guides (relevant section probably 3 and or 4) but they were written perhaps a decade ago. However regarding the specific issue of bore scoring in M96/7 engines - I have attached two photos.
The 3 liners/cylinders reveal the importance of the temperature cooling capacity of the cylinder blocks. The left hand cylinder was a spare from a 3 litre 944 turbo engine I built many years ago but the 2 ribbed areas show the depth of the coolant in the block. Furthermore 100% of the coolant from the pump was fed into the block and around those cylinders before it passed into the cylinder head. In contrast the cylinder on the right hand side is from an M96/7 engine showing the very much reduced coolant depth (the ribbed area).
But this is not the only issue influencing the amount of coolant in the block to cool the cylinders because the second picture shows how the coolant enters the engine from a gallery and the round hole is where it flows up directly into the cylinder head for each cylinder. The small slot at the top shows the area through which a small proportion of that coolant can get into the cylinder block to cool it which is about 10% of the total (or 10% of the amount the 944 and 944 turbo flowed). So a tiny fraction of the coolant that used to go into a 944 can get into the M96/7 block- which then has a much reduced area to remove heat from the cylinders = greater temperature rise.
The cylinder in the centre of the picture is from a batch made recently to repair a 9A1 Gen 2 seized bore and you can see from this that it has increased the coolant depth to something similar to the older 944 which also had an Alusil bore. This is why we fitted a low temperature thermostat and increased the proportion of coolant passing into the block when we fit our Nikasil cylinders and the small amount of coolant in the block combined with shallow depth and the fact that it enters both banks at the bottom increases the temperature at the top and hence thins the oil. But on bank 1 the thrust face (where they score) is at the cool side and on ban 2 it is the hot side.
This issue is a small percentage of the work we carried out over a decade ago – into problems with these engines and I have no problem in admitting that most of our research we retain as our intellectual property and have not yet released to the public.
This means we know very much more about these engines and their weaknesses than we make public and carry out many other changes to minimise their influence when we rebuild engines.
It costs a small fortune to carry out this type of research and I don’t see why we should then make the results available to our competitor for free. It is only after our record of many years and hundreds of successful engine rebuilds has resulted in a superb reputation and the volumes we repair are so large that our investment that made that efficient has made it unlikely that anyone else could justify the viability of trying to reproduce a similar set-up in opposition (with what could only ever be less numbers now) - that we are prepared to offer the advice we do especially when it may help owners seeking answers to avoid exploitation and make more informed choices.
So whereas I am happy to advise as much as possible when asked – there are still limits to how much we are prepared to divulge especially when competitors offer alternative explanations and our responses may show them where they might be going wrong and enable them to compete without putting in the time or absorbing the costs of our research that found the right answers in the first place.
I will also try and attach the photo of the DLC coated piston that revealed the tiny scratches - although it doesn't look as obvious in the photo as it does in real life.
Baz
3 cylinders left to right 944 turbo, 9A1 and M96/7 The round hole is where the coolant flows to the head and the slot is what enters the cylinder block DLC coated test piston with small scratches A small selection of pistons with different coatings we tested.
Pistons are complicated things. Assuming in my description mentioning verticality - and that the the bore and piston run vertically in an upright engine - for racing (where the piston only has to work flat out and in a small rev band) they tend to have more ovality (so they reduce the surface area where they touch the cylinder bore and hence friction) but the vertical face that touches the bore is often flat (or straight) so that the thrust touches the bore in a long vertical patch but narrow circular patch. This makes them more noisy as the piston passes over TDC and BDC because the piston slaps from one face to the other.
Pistons for fast road use tend to have that vertical face slightly curved so that the piston rolls rather than rocks and the impact at TDC and BDC reduces and with it the noise, but this also reduces the vertical surface area so to reduce wear they tend to have less ovality and spread the thrust forces across a wider area of the circumference.
Pistons also distort under load and flex and bend - so there is a lot more to consider in their design than many realise.
Most specialist piston manufacturers are less familiar with the shapes that are best for a road going sports car and apply racing profiles to all their products.
Here is a view of three different profiles we tested with especially soft coatings (so we could easily see the differences).
Left - widest contact patch, Right typical racing profile, centre something in-between
These were for one for our oversized engines - and you can see the different wear patches demonstrating the different contact patches.
The taper is also another feature designed so that at running temperatures the hotter top expands the right amount to create a good fit in the bore. Most racing engines have more taper (and the taper can also be barrel shaped).
In my experience - a piston skirt will only impact on Nikasil if the piston is very tapered (towards the top) and not curved in the vertical plane and it has too much clearance. Nikasil is in my experience virtually indestructible.
The information I sent privately to Jake includes video of "our 9A1 failures" I am sure he will result in him agreeing with our findings over here as they show the centre to bottom of the bore and piston seized both sides - not scoring). I also sent an analysis of Gen 2 and M96/7 piston profiles showing that the Gen 2 has less ovality and rod/stroke analysis of different models suggesting this is not a significant issue.
The difference in ovality of 9A1 pisotn is relevant and important because a piston has two webs running across it that are joined to the curved part of the piston at the edges whereas the centre (where the thrust loads that push against the cylinder wall and in turn push the rod down to rotate the crankshaft) is a curved piece of aluminium that is unsupported.
If a piston has too much ovality then if the piston gets too hot and expands too much (to cause a 2 sided seizure) it will usually be the centre of the piston faces that squeezes too hard against the cylinder wall and the seizure will be in the centre (which is most common).
Even if the cylinder stretches oval (as in M96/7 open deck cylinders) as long as the piston has enough ovality, it will have too much clearance to expand enough to seize both sides but the extra clearance will make it easier to squeeze out the oil film at lower rev high torque situations and rub the face on the thrust side against the cylinder wall, wearing out a soft coating and entrapping released silicon particles to damage the surface - one side only = bore scoring. Anyone that has used wheel plates (2 flat metal plates with oil between them you sit your wheels on when doing wheel alignment) will know that the loads are not enough to squeeze out all the oil and they freely move about as you adjust the geometry. However if you used too thin an oil or too heavy a car and left the weight on the plates long enough the oil would eventually all squeeze out and the plates would bind together) and this is similar to the oil film between piston and bore under thrust loads driving the car.
A common phenomenon in larger engines is like M96/7 scoring but they call it "scuffing" (because it is on the thrust side of the piston) and the people analysing those failures have worked out it occurs at low revs high torque scenarios - and this is because of the time factor the piston is trying to squeeze out the oil film between the piston and the cylinder wall. At 2000 rpm there is 3.5 times longer that the piston is squeezing the oil film thinner during one stroke than at 7000 rpm. If ever the metal of the piston face rubs against the cylinder wall under load without an oil film between them - it will quickly over heat and score, scuff or seize. this can partly explain why we have seen a higher proportion of M96/7 tiptronic engines score (over here) than manuals - because they normally accelerate from standstill in 2nd and with high torque compared to a manual in 1st (and hence another piece of our UK advice has been to select 1st in a tiptronic if you intend a fast take off).
The DLC coated pistons we tried in the M96/7 engines showed very interesting results. The surface that was beautifully shinny from new was still just as unmarked and shinny all over the thrust face on strip down except for thin score lines where a loose piece of silicon had broken free from the cylinder wall matrix and got entrapped between the piston face and the cylinder wall - so although the oil film was sufficient to keep the shinny DLC coating away from metal to metal contact at the cylinder wall - it was a small piece of silicon that still damaged the surfaces.
This helped us understand what was going on with the Lokasil engines and why the manufacturers stated that they needed a hard coated piston to survive (which incidentally they also supplied with the cylinder blocks). The change from cast to forged pistons coincided with the change from hard to soft piston coatings (and piston supplier) so although that could lead to a conclusion that it was the differences between cast and forged pistons that were significant - we thought it was the change from a hard iron coated piston to a soft plastic coating that was more relevant.
Similarly although too much petrol in the bore will dilute any oil there and make it less effective - it will (as confirmed by a boroscope) always settle in the bottom of the bore due to gravity in a horizontally apposed engine. This would make the underside of the pistons more likely to score if that was the cause - yet bank 2 (which almost all the scoring takes place) scores on the top where the thrust face is while bank 1 (that should suffer most if fuelling was significant) has the thrust face at the bottom (where is pools) yet rarely scores.
From this and the fact that the top of the cylinders are the hotter part of the cylinder (so oil films will be thinner) linked issues relating to the forces between the piston and Lokasil bore as the most significant issues influencing bore scoring and because Nikasil does not release loose silicon particles (and can run with any or no piston coating as a result), it solves the problem totally for us over here in the UK.
However the 9A1 Gen 2 engines are quite different. Because the pistons are less oval, bore clearances smaller and piston coating are harder than plastic and running in more solid Alusil that releases silicon particles at an insignificant rate (and they are smaller), so there are 2 causes that could lead to damage.
(1) This is what we have seen where both sides of the piston and bore are damaged in a seizure (not scoring one side). Because the piston in our scenarios is not hotter than it was designed to be (i.e. running OK) - but if the cylinder shrinks into it at the sides (due to ages stress relieving where the 2 sections of the casting at the bottom are pulling it inwards) it is also in line with the webs supporting the gudgeon pin bores (that prevent the piston skirt at the sides from flexing inwards) and so the pressure between the piston and bore at the sides is sufficient to squeeze out the oil film and result in a siezure (especially when too fast a warm up allows the piston to grow faster than the bore expands).
(2) This scenario is what could make scoring occur over in the USA where the bore is still round but something else has made the piston expand more than it was designed too do because it has become too hot. Because the piston has been designed with less ovality the sides "stick out" (for want of a better description) more than the centre in M96/7 pistons and as the sides are supported by the 2 webs where the gudgeon pin fits (like all pistons) - so if they got too hot the sides would expand more but resist being flexed inward under load and this could make the contact area that squeezes out the oil film until it is too thin or not there at all - be the sides of the pistons instead of the more familiar centres.
So whereas the M96/7 engines have more piston ovality and much bigger bore clearances (especially as they go oval through being unsupported), and are unlikely to experience a piston expanding too much so it is too big for the cylinder diameter and clearances (so unlikely to seize both sides) it could still score the thrust side because of loose silicon particles impinging on the soft plastic piston coating - a 9A1 Gen 2 engine could seize one side or both sides if the piston expanded too much.
I cannot see from the bore scoring video if the Gen 2 piston or bore were scored one side or seized both sides (the video I sent to Jake from our UK examples shows both sides). However IF the failures in the USA are one sided scores the tighter bore clearances combine with the shape of the pistons could cause that if the piston got too hot and therefore any contributing differences (like fuel, injectors, fuel pumps etc) could indeed cause the problem.
Even if the failures in the USA are double sided seizures anything that results in the piston getting hotter than intended could cause them.
It could be that in the UK a small number of engines have failed with scoring but were replaced under warranty in the main dealer network (and hence we have not seen that yet and have no experience of it) whereas the ones we have seen were older (and out of warranty) lasted longer, have had many more heat/cool cycles and were indeed caused by aged related stress relieving (as we first muted).
It could be that in he USA owners have brought failed engines to Jake earlier than the independent network got to see them in the UK.
So I am pleased to work out that for these reasons I can agree with the many posts that have suspected that we have both been right about our different experiences with and causes of 9A1 Gen 2 failures in our different Countries.
I can also now understand why there may be different influences in the USA and UK that affect the failures in M96/7 bore scoring incidents.
Over here although we change many things in the engines when we rebuild them - if we only ever changed the cylinder bore material to our Nikasil alloy cylinders (and adjust the coolant flow and LTT setting) it results in excellent reliability and so we cannot link any other significant issues with bore scoring other than the Lokasil material (since the plastic coated pistons that do not last in Lokasil do in Nikasil - cannot be faulted when running in Nikasil).
However although we are sure it is free silicon particles that damage the bores and scores them in Lokasil engines with plastic coated pistons, anything that reduces the effectiveness of the oil viscosity and lubricity would allow the pistons to run closer to the cylinder walls with thinner oil films and enable free particles to cause damage sooner. So if there is a problem with injectors as a result of fuel differences in the USA (that we don't experience over here), or more owners run with thin oils (as per the handbook) as they age and general clearances increase etc - that could again make us both right about the different contributory factors in our separate Countries.
Currently all the cylinders are being probed and pistons are being 3D scanned at LN.
Once my M96/M97 Engine Rebuild School is completed, we should be able to start sharing some results.
Earlier this week Lake Speed, Charles and I got together to shoot a video series on LSPI. We studied the pistons from Bronz’s engine and came up with a pretty good synopsis. The piston that Baz sent me a picture and some video of his failed pistons, and they are definitely very similar.
As I stated earlier, it will take a long time, with many more failures having to be studied to have a solid understanding of this. It does not come overnight.
The 3D plots and scans should shed some light on the subject that we’ve never shared before.
Really huge thanks to Jake and Baz for sharing so much information on topic that hasnt been discussed much in the past. I know some people dont like seeing this, but its reality. I was one of you, believing that this will be a forever "bulletproof" engine and then reality hit.
I reached out to the previous owner again to see if he still has records for the car. He did respond and gave a little insight on what he was doing to the car. I know this may not be the issue but it will be interesting to see once the injectors come back from the lab. From my understanding using fuel additives could damage some seals over time and possibly cause the injector to malfunction (correct me if im wrong). However I do agree with Baz's theory about expanding pistons and shrinking and ovaling bores and wearing down the piston the way it did in my engine.
Really huge thanks to Jake and Baz for sharing so much information on topic that hasnt been discussed much in the past. I know some people dont like seeing this, but its reality. I was one of you, believing that this will be a forever "bulletproof" engine and then reality hit.
I reached out to the previous owner again to see if he still has records for the car. He did respond and gave a little insight on what he was doing to the car. I know this may not be the issue but it will be interesting to see once the injectors come back from the lab. From my understanding using fuel additives could damage some seals over time and possibly cause the injector to malfunction (correct me if im wrong). However I do agree with Baz's theory about expanding pistons and shrinking and ovaling bores and wearing down the piston the way it did in my engine.
Thanx for this..... I add a bottle of Techron every once in a while in the belief it will clear the injectors. So here is the rub: All "good" gas such as Exxon, Shell, Chevron, use additive packages to "clean" your engine.. that is how they differentiate themselves in the market and justify their higher price over the cheap gas stations. Techron in a bottle is what Chevron already has in it at the pump so my little brain thinks it is an OK additive. So.... let me put this bluntly..... YIKES!