Feedback from Gen 2 engine failures

Feedback from Gen 2 engine failures

Author
Discussion

hartech

Original Poster:

1,929 posts

216 months

Sunday 12th February 2017
quotequote all
Thanks magic919 that actually confirms what we found on Saturday morning from the latest example - that there is a creeping failure mode in one area and a slight tolerance variation in another that both result in the exact same issues and reasons for the actual failures.

Can I ask if you can remember the time of year and distance travelled before you realised it had failed or perhaps if not - your typical distances you travel and if it is in traffic and areas with restricted speeds or you are able to (and enjoy) the freedom to open it up on acceleration a few minutes into your driving?

I appreciate that many owners will want to know more and I am sorry but I am not prepared yet to announce the technicality behind our findings. It has cost us a lot to get this far and our future position in the market is influenced by our confidence in our analysis of - so far - a very small number of failures just beginning to come to light.

I think announcing things too soon could do more harm than good - there is nothing after all that this delay will do to make cars fail or not and I want to be sure of my facts. Hopefully more examples will come to light and help reinforce and not contradict the direction we see things going.

You will not forget that my reasons for posting were entirely to try and gather more evidence and feedback so I could feel confident enough to continue to invest in this area to find causes and develop a solution - and not to spread gossip nor scare people unnecessarily.

I know from experience that it would have been better if we had enjoyed the facility of the Internet many years ago when other problems afflicted different models and indeed would have speeded up solutions if we had more information when the M96/7 engines started going wrong.

It would also help if the manufacturers were more open about such issues. I understand why they would not want to make it public but it would benefit owners enormously if they could confide in a technical exchange way with respected independent providers who have the facilities and scope to find and market solutions that they themselves cannot respond quickly enough to do - but then I see the difficulty they can be in and anyway a pig just flew by.

Baz


Magic919

14,126 posts

200 months

Sunday 12th February 2017
quotequote all
My time with the car was pretty brief. I bought it at the end of Feb 2013 and drove it for a couple of weeks, mostly to work 12 miles away. I went into hospital and didn't drive it for about a month as I was recovering. It got the odd drive, by my wife and son-in-law, mostly 50 mile plus trips.

First or second trip once I resumed driving it failed. This will have been first week April, still pretty chilly that year. We were 20 miles in having done M4 and M25 onto M3. We cleared the 50 restricted part, accelerated up to about 90-100 and it failed shortly after that.

The car had come from a London branch, so I presume did short trips there. Had just under 14k when I bought it and not much more when it failed.

hartech

Original Poster:

1,929 posts

216 months

Sunday 12th February 2017
quotequote all
The old GT3 and turbo's had a separate cylinder block and separate alloy Nikasil plated cylinders fitted into the cylinder blocks.

This is probably the best solution to a variety of design and manufacturing problems that I will reveal when the full technical explanation about the Gen 2 failures is made public.

The Gen 2 (9A1) has an Alusil cylinder block cast with the crankcase in one piece and with more than 13% silicon in the casting mix.

Aluminium cab only absorb silicon up to 13% after which any excess forms silicon nodules evenly distributed throughout the casting. How big they are and how many is connected to the mixture ratio and time maintained at a high temperature. It makes the block more difficult to machine but creates a very strong casting with a surface that etched or exposed back at the cylinder wall creates a microscopically uneven surface of hard silicon spots and valleys for oil retention. - Because of the hard spots the pistons last longest with a hard coating (currently ferrous probably "ferrotec" or similar).

Original Alusil engines (924S. 944, 968) had ferrous coated pistons and lasted very well.

In between - several Alusil engines resorted to plastic coatings that were not as hard (especially at elevated temperatures) and didn't last so long. Since the process of ferrous coating was changed for health and safety reasons - the latest Gen 2 engines have a ferrous coating again (good) but I don't know if it is the same. Either a way was found to return to the original method without health and safety issues or another process was found that was similar.

M96/7 engines had Lokasil cylinder pre-forms cast in that were preformed with silicon is a porous suspension so aluminium could flow into it during the high pressure casting process forming a silicon high surface cast in near the cylinder wall and reducing the machining difficulty (and hence cost) over the remainder of the casting. They were also open deck with a cylinder tube unsupported and held roughly midway down by being cast to the cylinder block leaving a free standing tube sticking out at the top and bottom.

I have not looked into the Cayenne engines yet with great interest but think they are (or were) Alusil with plastic coated pistons but under less individual stress and closed deck construction.

I am happy to inform you all that I do not blame this failure on either the piston coating or the cylinder finish.

In the list of preferred methods of manufacturing cylinders on a quality basis I would put separate alloy Nikasil cylinders top (GT3 and turbo), fixed Nikasil plated wet alloy cylinders next, Dry alloy Nikasil liners next, Alusil closed deck cylinders next (providing some extra features were included), open deck Alusil cylinders next, if the bore was a little smaller - cast in ribbed iron liners next, if the alloy cylinder that a liner was fitted into was if the top deck supported them - iron liners next. I would not list Lokasil as I think the bonding between the silicon and the alloy is not strong enough to survive the high loadings, oil wash and temperatures - long enough. This order would almost reverse on a cost basis.

In terms of piston coatings (where necessary) I would put the original ferrous coating first, the new ferrous coating (or ferrotec) next and plastic coatings last.

With Nikasil plating the surface is so good that there is no need for a coating of any kind however a plastic coating does improve running in by making up a minute difference in expansion during the heating cycle and the inevitable slight difference in individual pistons.

It is impossible/impractical to turn an M96/7 cylinder block into one that has a separate cylinder block so that option is not available. We turn it into the closest possible to a GT3 or Turbo block by fitting an alloy Nikasil plated liner (or more accurately new wet cylinder).

The Gen 2 cylinder is part of the crankcase - very good for production. It is closed deck top and bottom (good for cylinder stability) and the engine has a type of ferrous coated piston - so you would reasonably assume it should be perfectly OK. However there are a few different ways to design these different versions and it can present a small problem in some circumstances - which the evidence has so far shown us is present.

The majority of engines will possibly never go wrong or only after acceptable mileages for a high performance sports car (depending largely on pot luck of the owners driving styles and service quality). Some will go wrong due to the above and others with every possible issue managed perfectly will still fail due to the cause we have uncovered in the engines we have had sight of and investigated.

It is obvious that the ones that failed must have some cause - the difficulty is working out if it will affect all and if so when.

Clearly as contributors to this admittedly small readership are now coming forward we discover some have failed really early as well.

This is why I asked for responses from anyone who has had a failure - to help me work out the answer to what you all want to know (and I do) to judge what the statistics for likely failure are and for me the commercial opportunities and implications.

It is a simple request and so far it has worked very well indeed and sufficiently for me to start investing in a solution.

Whether it will prove worthwhile is something only time will tell.

Any further information is welcomed - thanks


Baz


ChrisW.

6,210 posts

254 months

Sunday 12th February 2017
quotequote all
Very interesting ....

If there is any evidence of a pre-cursor ... we seem to have high oil consumption ...

hartech

Original Poster:

1,929 posts

216 months

Sunday 12th February 2017
quotequote all
Chris - firstly I pass on my hope that your engine is OK. Secondly can I also add (for those that cannot wait to find a way to discredit what we do) that we are perfectly busy enough (flat out in fact) and do not need any more additional engines to rebuild. I rather hope that if this proves to be a problem it emerges in numbers in many months time when M96/7 rebuild numbers start to reduce (if indeed they ever do).

You would need to pay for a boroscope check. If it has seized they usually then squash that piston smaller and carry on running but usually with noises and higher oil consumption.

If it slightly seized (or nipped) it may only show higher oil consumption.

It will probably be cylinder 1 or 4, but it is possible for any cylinder at this stage (they are the only ones we have identified and measured that show the full fault although cylinders 2 and 5 show some signs of the same issue). So far cylinders 3 and 6 look unlikely but that may just be because of the relatively small sample so far.

If it has failed then apart from the higher oil consumption it will eventually start knocking.

I somehow doubt you have a seizure problem - but if it is bothering you it would be a good thing to have it checked. Alusil bores show up better than Nikasil ones as the Nikasil finish reflects light better and can fool anyone inexperienced at looking.

Good luck,

Baz

ChrisW.

6,210 posts

254 months

Sunday 12th February 2017
quotequote all
Too many Chris' in this !

With my CR sold on 34k miles --- some of them fairy testing, and my GT4 now on 8,500 miles --- both of them with far too many oil changes (if that is possible) but no problems so far, if I had a really serious engine problem I would feel very secure seeking a best solution with your Bolton team !

For the moment my average consumption including many track days has been 10,000 miles per litre. In fact, I use so little oil that will only add a maximum of 200 ml which I accurately measure, because of the temperature effect on the oil level measurement ...

I ran both in for 1500 miles with gently increasing load and revs and finished running both in with a track day, and an oil and filter change ...

I even changed the gearbox oil, now there's nerdish smile


Magic919

14,126 posts

200 months

Sunday 12th February 2017
quotequote all
ChrisW. said:
Very interesting ....

If there is any evidence of a pre-cursor ... we seem to have high oil consumption ...
This bit is confusing. It made it sound like you were suffering high oil consumption.

ooid

4,049 posts

99 months

Sunday 12th February 2017
quotequote all
hartech said:
The Gen 2 (9A1) has an Alusil cylinder block cast with the crankcase in one piece and with more than 13% silicon in the casting mix.

Aluminium cab only absorb silicon up to 13% after which any excess forms silicon nodules evenly distributed throughout the casting. How big they are and how many is connected to the mixture ratio and time maintained at a high temperature. It makes the block more difficult to machine but creates a very strong casting with a surface that etched or exposed back at the cylinder wall creates a microscopically uneven surface of hard silicon spots and valleys for oil retention. - Because of the hard spots the pistons last longest with a hard coating (currently ferrous probably "ferrotec" or similar).

Baz

hmmm quick question to Baz;

I read they are supposed to use specific polishing stones to get the surface finished after the etching process. When it is done properly it should be super smooth and quite plain colour. Everytime I see some strange dark or burnt marks on people's engine pictures I assume these just have to be where carbon has stuck to the bores because they have not been finished properly or basically crap surface quality? Whatever the finishing process Porsche is doing (full automation or semi?), there must be a failed process there in order to have so many cars affected because as far as I know Audi R8 engines (4.2 FSI V8)also include Alusil and you do not hear anyone freaking about their bores? Or is it again, "V" configuration engines physically might cause less friction in the area due to their geometry and movement comparing to flat configuration?

Thanks for sharing the all these info and research here again!

beer

Slippydiff

14,742 posts

222 months

Sunday 12th February 2017
quotequote all
ChrisW. said:
Too many Chris' in this !

With my CR sold on 34k miles --- some of them fairy testing, and my GT4 now on 8,500 miles --- both of them with far too many oil changes (if that is possible) but no problems so far, if I had a really serious engine problem I would feel very secure seeking a best solution with your Bolton team !

For the moment my average consumption including many track days has been 10,000 miles per litre. In fact, I use so little oil that will only add a maximum of 200 ml which I accurately measure, because of the temperature effect on the oil level measurement ...

I ran both in for 1500 miles with gently increasing load and revs and finished running both in with a track day, and an oil and filter change ...

I even changed the gearbox oil, now there's nerdish smile
Well if you could keep it in one piece until the RS day it would be appreciated Chris smile I subscribe to your ownership regiment, thrash them mercilessly and service them religiously (if not more).
Stop start, short journeys that rarely or never see the engine at optimum working temperature for extended periods was always the kiss of death for an engine, no reason to think it would be any different now.

GT4P

5,188 posts

184 months

Sunday 12th February 2017
quotequote all
I will follow this post with great interest and I am sure hartech will provide us the best reason, answer and solution.
But interestingly from his first comment he points mainly to the 3.8? or is this all capacities or this a 911 problem and not a cayster problem ie that the 911 block is furthest from rads? But do 991gt3 also suffer?

7and911

51 posts

121 months

Sunday 12th February 2017
quotequote all
Slippydiff said:
I subscribe to your ownership regiment, thrash them mercilessly and service them religiously (if not more).
Stop start, short journeys that rarely or never see the engine at optimum working temperature for extended periods was always the kiss of death for an engine, no reason to think it would be any different now.
So what is the advice for owners who lives in busy cities and their 20 minutes journey to work involves too many stop/start?

I got my 991.1 from new since March 2014. Easy break-in for 1500 miles. Always waited until the rpm comes down before starting driving. RPM less than 3500 until oil temp reached. Never been to track, never redlined and never launch controlled. And here I am, could be heading for a new engine at 5,000 miles.


I have to admit though I didn't change the oil after the brake-in period as the OPC advised me at that time to stick to the 2 yearly maintenance schedule. Yes my journey to work was 8 miles but engine temp always reached by the time I arrived work. Can't believe how unreliable engine (or may be unlucky).

ChrisW.

6,210 posts

254 months

Sunday 12th February 2017
quotequote all
Magic919 said:
This bit is confusing. It made it sound like you were suffering high oil consumption.
I meant the Forum "we" --- not me --- for those that may be concerned ??

ChrisW.

6,210 posts

254 months

Sunday 12th February 2017
quotequote all
7and911 said:
So what is the advice for owners who lives in busy cities and their 20 minutes journey to work involves too many stop/start?

I got my 991.1 from new since March 2014. Easy break-in for 1500 miles. Always waited until the rpm comes down before starting driving. RPM less than 3500 until oil temp reached. Never been to track, never redlined and never launch controlled. And here I am, could be heading for a new engine at 5,000 miles.


I have to admit though I didn't change the oil after the brake-in period as the OPC advised me at that time to stick to the 2 yearly maintenance schedule. Yes my journey to work was 8 miles but engine temp always reached by the time I arrived work. Can't believe how unreliable engine (or may be unlucky).
Understood and if it happens to "you" --- from the you perspective that's a 100% failure rate.

It will be very interesting to see how this develops --- if it does.

My theory is if there are few failures, all manufacturers can afford and should afford the cost of really looking after the customer(s) concerned.

If many failures, then the manufacturer should be in damage limitation mode and at least seek an affordable repair --- damage limitation that's relevant fro both the manufacturer and the customer.

If there is a problem here, Porsche will hopefully be ahead of the curve in solving it ... maybe in my dreams ?? !!

If not, we have Baz smile

hartech

Original Poster:

1,929 posts

216 months

Monday 13th February 2017
quotequote all
GT4P - in answer to your first post - watch this space!

Chris W You confused me too - glad it is OK.

Ooid This is not a surface finish problem. I appreciate that it must be very frustrating for me not to come straight out with my conclusions - but this post was initiated by me to obtain more information about failures so I could investigate thoroughly and only report more solid facts and not speculation. I feel I have a duty to be really sure with such an emotive subject. I have purposely not revealed all yet and only tried to build my information data base on scant responses. I understand that just like the M96/7 engines - there have been some replaced during the first ownership by Porsche and therefore the independent specialists only get to find out if there is a problem some way down the line - and then just trying to find out more invokes all sorts of negative and reactionary responses confusing things - more soon.

GT4P - so far it is only 3.8's reported and I have not yet inspected or stripped 3.6 (but I have bought one and it is in the pipeline as there is something I want to measure).

If Porsche have followed their usual generic policy (and I have already worked out that the cylinder blocks 3.6 and 3.8 are the same height despite different stoke engines corrected via the piton pin offset - as they did with the M96/7 range) then if the cylinder block casting core is the same for both it would mean that the cylinder walls in the 3.6 would be 2mm thicker. If that turns out to be the case it may either delay the problem further or eliminate it all together. Our 3.6 is a high mileage example because we also want to see how other components are lasting (like crankshaft shells and rings) and this is all part of a very serious and professional approach that is typical of how we go about resolving such issues. It unfortunately takes time but those who may have read about our opinions over a decade ago (regarding M96/7 issues) and 944/968 etc years before will know that when we do firm up on a cause it is usually spot on and as a result our solutions have always been top of the market.

Overall - if an engine exhibits this trend - I don't think it can be resolved with replacing the crankcases.

Replacement crankcases may have the same proportion of random ones that tend to fail unless Porsche understood the problem soon enough to change something else to solve it (which I doubt purely on a time scale between design, tooling and manufacturing - if it takes a long time to reveal itself after the start of the new production run). Nothing seemed to have been done about IMS bearings for a long time (and that was an easy potential retrofit) and nothing was done about cylinder cracking in the M96 and scoring in the M97 range at all.

The most important advice I can give at the moment is to be patient from a cold start to delay high power until the engine and oil has fully warmed up (especially when it is cold). If you only undertake short journeys I am sorry to advise avoiding high full throttle power delivery all together - less throttle but still revving it would be OK.

Hopefully you will all appreciate that we are on our own in looking into this - a small business - with many other priorities and this is just the start of trying to find out what sort of problem caused these failures and if it is likely to grow into a more serious issue.

Baz

westwood35

123 posts

182 months

Wednesday 15th February 2017
quotequote all


Are these the cases you are referring to????

hartech

Original Poster:

1,929 posts

216 months

Thursday 16th February 2017
quotequote all
Yes Duncan that looks like the right type.

Baz

hartech

Original Poster:

1,929 posts

216 months

Thursday 16th February 2017
quotequote all
Yet another one with what looks like exactly the same failure in exactly the same place - not a coincidence.

Baz

IMI A

9,410 posts

200 months

Sunday 26th February 2017
quotequote all
Out of interest are the newer DFI engines a completely new engine sharing no parts with the old chocolate engines or just a redesign of the old engine?

hartech

Original Poster:

1,929 posts

216 months

Monday 27th February 2017
quotequote all
A completely new engine with superior features. The number of weak points are minimal and they have proven very reliable.

The identical fault we have analysed was present in all 5 of the examples we have had access to (and others we have heard about or seen pictures of). Some faults require large numbers to analyse and circumstantial feedback to be sure of a cause - but others can be resolved from just one sample (if the mistake is obvious).

We are still testing out our theory about the cause but I am sure there will be no need to see any more to confirm the technical issue and conditions - causing the seizures.

We are presently concentrating on the repair of the ones we have and the design of tooling and some other products that might emerge in the future.

More on this soon - but little to worry about as failures seem low in numbers and then after relatively high mileages.

Still more feedback would be greatly appreciated from anyone who had an engine replaced - mileage covered and ambient conditions at the time would also help.

Thanks,

Baz




IMI A

9,410 posts

200 months

Monday 27th February 2017
quotequote all
hartech said:
A completely new engine with superior features. The number of weak points are minimal and they have proven very reliable.

The identical fault we have analysed was present in all 5 of the examples we have had access to (and others we have heard about or seen pictures of). Some faults require large numbers to analyse and circumstantial feedback to be sure of a cause - but others can be resolved from just one sample (if the mistake is obvious).

We are still testing out our theory about the cause but I am sure there will be no need to see any more to confirm the technical issue and conditions - causing the seizures.

We are presently concentrating on the repair of the ones we have and the design of tooling and some other products that might emerge in the future.

More on this soon - but little to worry about as failures seem low in numbers and then after relatively high mileages.

Still more feedback would be greatly appreciated from anyone who had an engine replaced - mileage covered and ambient conditions at the time would also help.

Thanks,

Baz



Good to know - thanks