Lokasil 1, 2 and Gen 2 9A1 engines
10-31-2017, 08:23 AM
#1
User
Thread Starter
Lokasil 1, 2 and Gen 2 9A1 engines
I have received a private question from this forum about the difference between Lokasil 1, 2 and Alusil (Gen 2 9A1 engines).
According to the manufacturers, Lokasil 1 was a matrix with silicon particles 20 to 70 microns in diameter - so the largest was 0.0028" (in Imperial thous) while Lokasil 2 was 30 to 120 microns - so the largest was 0.0048".
Alusil particles (Gen 2 9A1) were much better bonded into the matrix and the size was the same as Lokasil 1.
Nikasil is electroplated and although the result is a matrix in which the silicon particles are a permanent part of the plated surface - even if one did become loose they are only 1/10 of the size of those in Alusil or Lokasil.
I think you can see where this is going because the particle size of Lokasil 2 is larger than the piston clearance - so a loose particle trapped between the piston and the cylinder wall could be bigger than the space it occupies.
But the larger particle size did seem to stiffen the Lokasil 2 cylinders so we see very few cracking, despite the wall thickness being similar and the piston forces of the 3.6 and 3.8 probably higher.
We do not know when Lokasil 2 replaced Lokasil 1 but from the evidence of scoring and lack of cracking it looks like it might have been after the 3.4 996 engines.
When the bores are honed the particles can have the edge machined smoother so that the thickness is less than the particle size - but when the particle becomes free it can turn and then the height or breadth is the original size.
Rather like concrete or tarmac - when a stone gets loose it tends to rub back and forth in the space it occupied and lengthen it into a furrow that then allows more stones (or particles) to get loose and do the same creating a long groove or scratch.
This is why the damage is only on the thrust side of the piston where the gap between the piston and the cylinder is less due to the power pushing the piston against the cylinder wall and squeezing the oil film more.
All silicon particle based cylinder bores will eventually loose some particles over time and when they do - if the oil film is thick enough to provide a big enough space between the piston and the cylinder wall they can slide out eventually. The thinner the oil and the smaller the clearance - the more the problem can be serious.
Of course not all the particles are the maximum size so it can be a lottery which one becomes loose and how big it is.
The original hypereutectic cylinders (Alusil) were used in conjunction with ferrous coated pistons which had a much harder surface than plastic coatings (especially at elevated temperatures).
Gen 2 (9A1) engines have reverted to Alusil and the pistons have a ferrous coating again (which seems a little different to the original coatings but seems to do a great job).
The cylinder scoring problem we have identified with the Gen 2 engines seems to have nothing to do with the bore type or piston coating - instead - over a long period of time and hot and cold heat cycles, the bottom of the cylinder bores minutely shrink inwards across the thrust diameter - reducing the piston clearance until on a cold day with a rapid heat cycle the piston expands quicker than the bore and that reduced cylinder clearance causes a typical cold seizure all round both sides of the piston.
Baz
According to the manufacturers, Lokasil 1 was a matrix with silicon particles 20 to 70 microns in diameter - so the largest was 0.0028" (in Imperial thous) while Lokasil 2 was 30 to 120 microns - so the largest was 0.0048".
Alusil particles (Gen 2 9A1) were much better bonded into the matrix and the size was the same as Lokasil 1.
Nikasil is electroplated and although the result is a matrix in which the silicon particles are a permanent part of the plated surface - even if one did become loose they are only 1/10 of the size of those in Alusil or Lokasil.
I think you can see where this is going because the particle size of Lokasil 2 is larger than the piston clearance - so a loose particle trapped between the piston and the cylinder wall could be bigger than the space it occupies.
But the larger particle size did seem to stiffen the Lokasil 2 cylinders so we see very few cracking, despite the wall thickness being similar and the piston forces of the 3.6 and 3.8 probably higher.
We do not know when Lokasil 2 replaced Lokasil 1 but from the evidence of scoring and lack of cracking it looks like it might have been after the 3.4 996 engines.
When the bores are honed the particles can have the edge machined smoother so that the thickness is less than the particle size - but when the particle becomes free it can turn and then the height or breadth is the original size.
Rather like concrete or tarmac - when a stone gets loose it tends to rub back and forth in the space it occupied and lengthen it into a furrow that then allows more stones (or particles) to get loose and do the same creating a long groove or scratch.
This is why the damage is only on the thrust side of the piston where the gap between the piston and the cylinder is less due to the power pushing the piston against the cylinder wall and squeezing the oil film more.
All silicon particle based cylinder bores will eventually loose some particles over time and when they do - if the oil film is thick enough to provide a big enough space between the piston and the cylinder wall they can slide out eventually. The thinner the oil and the smaller the clearance - the more the problem can be serious.
Of course not all the particles are the maximum size so it can be a lottery which one becomes loose and how big it is.
The original hypereutectic cylinders (Alusil) were used in conjunction with ferrous coated pistons which had a much harder surface than plastic coatings (especially at elevated temperatures).
Gen 2 (9A1) engines have reverted to Alusil and the pistons have a ferrous coating again (which seems a little different to the original coatings but seems to do a great job).
The cylinder scoring problem we have identified with the Gen 2 engines seems to have nothing to do with the bore type or piston coating - instead - over a long period of time and hot and cold heat cycles, the bottom of the cylinder bores minutely shrink inwards across the thrust diameter - reducing the piston clearance until on a cold day with a rapid heat cycle the piston expands quicker than the bore and that reduced cylinder clearance causes a typical cold seizure all round both sides of the piston.
Baz
The following 3 users liked this post by bazhart:
10-31-2017, 09:38 AM
#2
Rennlist Member
Thank you so much for posting this. Technical research on our cars is usually not accessible for the common consumer. Between this thread and the Chalk C2S track build we’ve had some great content lately.
10-31-2017, 10:17 AM
#3
RL Community Team
Rennlist Member
Rennlist Member
Well thank you for that. I will now cover my engine with an electric heating blanket each evening to keep it all snuggly and happy. I have already been reading bed time stories to it... you know the type.... a little car with a little engine who beats the faster cars through hard work and believing in himself.
So... besides blankets and stories, what can I do to keep my 2009 C2S happy forever?
Peace
Bruce in Philly
So... besides blankets and stories, what can I do to keep my 2009 C2S happy forever?
Peace
Bruce in Philly
10-31-2017, 10:35 AM
#4
Burning Brakes
"The cylinder scoring problem we have identified with the Gen 2 engines seems to have nothing to do with the bore type or piston coating - instead - over a long period of time and hot and cold heat cycles, the bottom of the cylinder bores minutely shrink inwards across the thrust diameter - reducing the piston clearance until on a cold day with a rapid heat cycle the piston expands quicker than the bore and that reduced cylinder clearance causes a typical cold seizure all round both sides of the piston."
Typical cylinder wear is along the thrust axis line to increase that diameter (ovalility). But you are observing the bottom of the cylinder drawing inward along the same thrust axis ?
Typical cylinder wear is along the thrust axis line to increase that diameter (ovalility). But you are observing the bottom of the cylinder drawing inward along the same thrust axis ?
10-31-2017, 03:36 PM
#7
Pro
What would be the minimum with the oil and coolant temperatures before one can start changing gears at 5000 rpms and above?
Trending Topics
11-01-2017, 06:26 AM
#8
User
Thread Starter
When we first became involved in Gen 2 seizures we measured the cylinders and detected a slow creep inwards at the lower area in the unseized cylinders and measured the same trend above the seized area. We then obtained some engines from crashed cars that had not seized and measured the cylinders and found the same trend.
The cylinder at one end of the blocks was OK, the middle on was smaller and the other end cylinder was smallest - in each set - and where the seizures were.
Ironically the much maligned "open deck" design of the M96/7 engines avoids this type of creep dimension change - but if the wall thickness is too thin doesn't prevent stretching the bores oval in the thrust direction. Was good up to and including the 3.2 versions but after than the cylinders were made thinner and hence allowed repeated strain to allow them to change shape.
The Gen 2 (9A1) engines are proper closed deck designs and that stabilises the cylinder bores but because they are now connected to the external cylinder block casting that can have a downside if the rest of the block distorts with time and alters the bore dimensions.
When any engine alloy material is manufactured there is usually temperature involved (high in castings, less in extrusions etc). Upon cooling the outside always cools first while the inside is hotter and has not yet contracted - so as the outside sets the inside continues to contract and leaves a thermal stress inside all parts of the material.
So when you machine castings, extrusions or rolled materials, anything that allows those stresses to be relieved minutely distorts the shape. In some production, materials are "stress relieved" after roughing out before final machining to eliminate the problem.
After the industrial revolution, large machine castings were left outside for a few years to "weather" - basically to allow heat changes to let the material settle down to a final shape before machining.
Since then no one has the time or resources to produce goods that way and as a result - sometimes and unexpected problem emerges.
The Gen 2 block has large thick cast areas below the cylinder bore (but a homogeneous part of it) where the crankshaft shells locate.
Differences in the way the original casting was produced and its cooling rate in different areas will leave stresses inside those areas.
These stresses are small and molecular but every time there is a temperature change cold to hot to cold - the most minute adjustment may take place in the original machined shape.
To reproduce this under test a manufacturer would have to drive several cars for a few hours them leave to cool right down naturally before repeating the process. Then after several months of this - strip and measure everything.
To reproduce a problem that only emerges after say 9 years of daily heat cycles - would take several years and then if a problem was found - it would have to be modified and the whole process repeated before release to the public. It's not going to happen any more as model changes needed to maintain sales demands product changes too fast for such lengthy testing (that may prove unnecessary anyway).
Instead manufacturers have to keep their fingers crossed and it seems that in the case of this engine a minute change has taken place in some engines.
Each time we were involved in this failure mode it had been driven on a cold day without lengthy warm up times before fast driving.
I would say you need to get the oil temperature up to normal running temperature and then drive for another 10 to 20 minutes, gradually increasing throttle opening before giving the car full aggressive throttle.
We do not know if this problem was found early and something done to eliminate it in newer production, or if it only affected some castings or production runs.
Some manufacturers do carry out freezing and heating cycles on some critical parts to speed up the stress relieving process before final machining - so it might be that those owners living in colder climates (if only in winter) who also drive aggressively in short journeys from cold - may promote more shrinkage sooner than say an owner in warmer (or a more stable annual climate) that warms the car up (or drives less aggressively).
All we can do to date is measure the engines that either fail or we get in for rebuild for some other reason and base our findings on those (which have all been identical).
So this may be a problem that goes away or it may remain rare or escalate with time - only time will tell.
Meanwhile, when we release our 4 litre upgrade - we expect to receive a number of engines that we can measure and update our fact file and analysis from there.
So far it seems a rare problem - insufficient to worry about too much but enough to just be patient before flooring the throttle (as really you should do anyway with any sports car).
Baz
The cylinder at one end of the blocks was OK, the middle on was smaller and the other end cylinder was smallest - in each set - and where the seizures were.
Ironically the much maligned "open deck" design of the M96/7 engines avoids this type of creep dimension change - but if the wall thickness is too thin doesn't prevent stretching the bores oval in the thrust direction. Was good up to and including the 3.2 versions but after than the cylinders were made thinner and hence allowed repeated strain to allow them to change shape.
The Gen 2 (9A1) engines are proper closed deck designs and that stabilises the cylinder bores but because they are now connected to the external cylinder block casting that can have a downside if the rest of the block distorts with time and alters the bore dimensions.
When any engine alloy material is manufactured there is usually temperature involved (high in castings, less in extrusions etc). Upon cooling the outside always cools first while the inside is hotter and has not yet contracted - so as the outside sets the inside continues to contract and leaves a thermal stress inside all parts of the material.
So when you machine castings, extrusions or rolled materials, anything that allows those stresses to be relieved minutely distorts the shape. In some production, materials are "stress relieved" after roughing out before final machining to eliminate the problem.
After the industrial revolution, large machine castings were left outside for a few years to "weather" - basically to allow heat changes to let the material settle down to a final shape before machining.
Since then no one has the time or resources to produce goods that way and as a result - sometimes and unexpected problem emerges.
The Gen 2 block has large thick cast areas below the cylinder bore (but a homogeneous part of it) where the crankshaft shells locate.
Differences in the way the original casting was produced and its cooling rate in different areas will leave stresses inside those areas.
These stresses are small and molecular but every time there is a temperature change cold to hot to cold - the most minute adjustment may take place in the original machined shape.
To reproduce this under test a manufacturer would have to drive several cars for a few hours them leave to cool right down naturally before repeating the process. Then after several months of this - strip and measure everything.
To reproduce a problem that only emerges after say 9 years of daily heat cycles - would take several years and then if a problem was found - it would have to be modified and the whole process repeated before release to the public. It's not going to happen any more as model changes needed to maintain sales demands product changes too fast for such lengthy testing (that may prove unnecessary anyway).
Instead manufacturers have to keep their fingers crossed and it seems that in the case of this engine a minute change has taken place in some engines.
Each time we were involved in this failure mode it had been driven on a cold day without lengthy warm up times before fast driving.
I would say you need to get the oil temperature up to normal running temperature and then drive for another 10 to 20 minutes, gradually increasing throttle opening before giving the car full aggressive throttle.
We do not know if this problem was found early and something done to eliminate it in newer production, or if it only affected some castings or production runs.
Some manufacturers do carry out freezing and heating cycles on some critical parts to speed up the stress relieving process before final machining - so it might be that those owners living in colder climates (if only in winter) who also drive aggressively in short journeys from cold - may promote more shrinkage sooner than say an owner in warmer (or a more stable annual climate) that warms the car up (or drives less aggressively).
All we can do to date is measure the engines that either fail or we get in for rebuild for some other reason and base our findings on those (which have all been identical).
So this may be a problem that goes away or it may remain rare or escalate with time - only time will tell.
Meanwhile, when we release our 4 litre upgrade - we expect to receive a number of engines that we can measure and update our fact file and analysis from there.
So far it seems a rare problem - insufficient to worry about too much but enough to just be patient before flooring the throttle (as really you should do anyway with any sports car).
Baz
The following users liked this post:
Saybaz (02-09-2024)
11-01-2017, 06:54 AM
#9
Pro
Thanks for the detailed background info Baz, as always, it's awesome.
I've had my 997.2 for 3 weeks now, so what is the 'normal' oil temp? I haven't driven it for long enough to know what would that be.
I've had my 997.2 for 3 weeks now, so what is the 'normal' oil temp? I haven't driven it for long enough to know what would that be.
11-01-2017, 07:15 AM
#10
Three Wheelin'
Baz - excellent writeup. A few additional questions:
1. For the .2s it sounds like you are saying the catalyst for bore scoring is cold weather operation with improper warm up. How many of the rapid heat cycle events would it likely take for symptoms to show: 1 or 2, or somebody ragging it out everyday in the dead of winter...?
2. For .1s are you saying the problem is more systemic: its not related as much to improper warm up in cold weather operation, its just a nature of the Lokasil 2 particle size...? If yes, what can be done to minimize this?
3. Do you have any guesstimates on how many .1 and .2s this effects? The forum poll on the subject looks to be so statistically inaccurate it does more harm than good.
Keep up the great posts!
1. For the .2s it sounds like you are saying the catalyst for bore scoring is cold weather operation with improper warm up. How many of the rapid heat cycle events would it likely take for symptoms to show: 1 or 2, or somebody ragging it out everyday in the dead of winter...?
2. For .1s are you saying the problem is more systemic: its not related as much to improper warm up in cold weather operation, its just a nature of the Lokasil 2 particle size...? If yes, what can be done to minimize this?
3. Do you have any guesstimates on how many .1 and .2s this effects? The forum poll on the subject looks to be so statistically inaccurate it does more harm than good.
Keep up the great posts!
11-01-2017, 09:15 AM
#11
User
Thread Starter
There are so may variables in climate, oils, etc that the best advice is to run the car for about half an hour and see what the oil temp is (usually around 10 degrees hotter than the coolant but no problem if slightly different to that).
I find it takes about half an hour to properly warm up an engine.
The cylinders we measured had shrunk to within a few tenths of a thousandth of an inch of the piston diameter.
Pistons heat up quicker than cylinder blocks and therefore expand sooner.
The failures we know about are all among the first Gen 2's manufactured and if an owner uses their car twice a day for say 6 days of the week - that is about 12 heat/cool cycles/week. They are 7 years old now and people go on holiday etc - so I guess most cars in daily use of the 2010 era will have had around 4,000 heat cool cycles.
The ambient temperature the car was at cold will influence the effect of the cycle as the greater the temperature difference the more the material will stress relieve.
Cars in warmer climates and used less often will have had fewer heat cycles and therefore - if this is a common problem that was not modified by the manufacturer, most engines will have had less cycles than this and therefore the problem will not emerge in any quantity until some years ahead.
Furthermore, delaying booting the car until it is properly hot will extend the life by allowing the block to expand with the piston rather than after it.
Pistons do distort and under load they can be pressed inwards by the thrust pressure and accommodate quite large variations in clearance but in cold weather - going flat out from cold in a gen 2 that is several years old and has had frequent daily use - will probably result in earlier failure than those that have shrunk just as much but are warmed up properly first.
Other than this problem (that in all fairness it is impossible for manufacturers to test for and alter before production) the Gen 2 engines should be very reliable and last for a very long time.
The Lokasil bore engines (M96./7) are different.
The influence on scoring will depend on a lot more variable factors - listed below.
(1) particle size, particle distribution, bonding strength, honing influence (all out of the publics control or influence).
(2) Bonding quality of the plastic piston coating (also out of the publics control or influence).
The following can be influenced by owners
(3) Thrust loads (highest when the torque is high and the revs moderate - better to operate at peak revs on full load than full throttle at low revs in low gears).
(4) Thickness and viscosity of the oil film that would separate the piston from being too close to the cylinder bore - influenced by oil choice, viscosity, and the running temperature of the coolant (hence a good idea to fit a low temperature thermostat).
Hence an owner from new that never ragged the car too much, fitted a LTT, used a good quality oil of thicker viscosity as the engines wear and who warmed it up and only drove fast at highish revs and was lucky enough to own a car that had good silicon particle distribution, bonding strength and size and had well bonded plastic coatings on the pistons - may get well of 100K before problems.
In contrast any owner (during a multiple owner period) that used thin oils, used high throttle openings from low revs (high torque typical pulling away in 2nd with a tiptronic), who allowed radiators to get clogged and ran with high coolant temperatures and happened to get an engine with poor silicon distribution and/or bonding and poor piston coating, may have a score after 20K.
In between are all the variations and permutations that make predicting failure and advising owners almost impossible except how to minimise the potential problem accepting that despite that an engine with less than perfect internals may still fail prematurely.
We find the most typical failure scores occur between 40 and 100K and probably the most common 60 to 80.
Baz
I find it takes about half an hour to properly warm up an engine.
The cylinders we measured had shrunk to within a few tenths of a thousandth of an inch of the piston diameter.
Pistons heat up quicker than cylinder blocks and therefore expand sooner.
The failures we know about are all among the first Gen 2's manufactured and if an owner uses their car twice a day for say 6 days of the week - that is about 12 heat/cool cycles/week. They are 7 years old now and people go on holiday etc - so I guess most cars in daily use of the 2010 era will have had around 4,000 heat cool cycles.
The ambient temperature the car was at cold will influence the effect of the cycle as the greater the temperature difference the more the material will stress relieve.
Cars in warmer climates and used less often will have had fewer heat cycles and therefore - if this is a common problem that was not modified by the manufacturer, most engines will have had less cycles than this and therefore the problem will not emerge in any quantity until some years ahead.
Furthermore, delaying booting the car until it is properly hot will extend the life by allowing the block to expand with the piston rather than after it.
Pistons do distort and under load they can be pressed inwards by the thrust pressure and accommodate quite large variations in clearance but in cold weather - going flat out from cold in a gen 2 that is several years old and has had frequent daily use - will probably result in earlier failure than those that have shrunk just as much but are warmed up properly first.
Other than this problem (that in all fairness it is impossible for manufacturers to test for and alter before production) the Gen 2 engines should be very reliable and last for a very long time.
The Lokasil bore engines (M96./7) are different.
The influence on scoring will depend on a lot more variable factors - listed below.
(1) particle size, particle distribution, bonding strength, honing influence (all out of the publics control or influence).
(2) Bonding quality of the plastic piston coating (also out of the publics control or influence).
The following can be influenced by owners
(3) Thrust loads (highest when the torque is high and the revs moderate - better to operate at peak revs on full load than full throttle at low revs in low gears).
(4) Thickness and viscosity of the oil film that would separate the piston from being too close to the cylinder bore - influenced by oil choice, viscosity, and the running temperature of the coolant (hence a good idea to fit a low temperature thermostat).
Hence an owner from new that never ragged the car too much, fitted a LTT, used a good quality oil of thicker viscosity as the engines wear and who warmed it up and only drove fast at highish revs and was lucky enough to own a car that had good silicon particle distribution, bonding strength and size and had well bonded plastic coatings on the pistons - may get well of 100K before problems.
In contrast any owner (during a multiple owner period) that used thin oils, used high throttle openings from low revs (high torque typical pulling away in 2nd with a tiptronic), who allowed radiators to get clogged and ran with high coolant temperatures and happened to get an engine with poor silicon distribution and/or bonding and poor piston coating, may have a score after 20K.
In between are all the variations and permutations that make predicting failure and advising owners almost impossible except how to minimise the potential problem accepting that despite that an engine with less than perfect internals may still fail prematurely.
We find the most typical failure scores occur between 40 and 100K and probably the most common 60 to 80.
Baz
The following users liked this post:
Saybaz (02-09-2024)
11-01-2017, 09:30 AM
#12
Nordschleife Master
Join Date: Mar 2003
Location: Destin, Nashville, In a 458 Challenge
Posts: 5,128
Received 904 Likes
on
532 Posts
So essentially, there is no guarantee that similar long term problems will not arise with the rebuilt motors by flat 6 or your company because long term testing is not economical or feasible.
These are startling and disappointing results with the 997.2 DFI engine. All the more reason to remain in new 911s covered by solid warranties and one of the primary reasons I used turbos as my daily drivers when the turbos had the Mezger/GT1 case. The risk of catastrophic failure in post air cooled NA 911s may reach the point where these cars are just not worth it. The 60k top end rebuilds on the air cooleds were a bit annoying, but that is like a 60k service charge compared to the $25k motor replacement in the 996 and 997s.
These are startling and disappointing results with the 997.2 DFI engine. All the more reason to remain in new 911s covered by solid warranties and one of the primary reasons I used turbos as my daily drivers when the turbos had the Mezger/GT1 case. The risk of catastrophic failure in post air cooled NA 911s may reach the point where these cars are just not worth it. The 60k top end rebuilds on the air cooleds were a bit annoying, but that is like a 60k service charge compared to the $25k motor replacement in the 996 and 997s.
11-01-2017, 09:56 AM
#13
These are startling and disappointing results with the 997.2 DFI engine.
11-01-2017, 10:05 AM
#14
Nordschleife Master
Join Date: Mar 2003
Location: Destin, Nashville, In a 458 Challenge
Posts: 5,128
Received 904 Likes
on
532 Posts
Its them drumming up business for the 9A1. Pretty sure we would have all heard about seized 9A1 engines on this forum if it was happening.....If one exists I would like to read it. All things can fail, even jet engines and bridges designed to the .999 MTBF. Thee best way to never find out is to never actually use it. I would not call this startling, the 9A1 has been out over 10 years now in the 997.2 and the 991.1. Its failure rate is low.
11-01-2017, 10:19 AM
#15
If 9A1 engines were seizing due to the lower part of the bore shrinking or something like that, this forum would be lit up like a Xmas tree. it not. Any mechanical part can fail, even more so if you don't let it warm up. But seizing, C'mon man. Their core business on the M96/97 rebuilds are dropping off as cars with them get wadded, sit in garages, and get off the road in general. They have done a great service for those guys, but there is less of those new guys to get business from as those cars with those engines fade out. in 10, 15 or more years you will see 996 and some 997.1 cars with blown engines on BAT or even barn finds needing a new engine and the owners just want to get rid of them. At some point the price of the M96/97 rebuild will far exceed the value of the whole car. Then it not worth it to most people. Not there yet, but its creeping up.