996 3.4 Engine - 2016 view?
Discussion
I've been looking at early 996s recently - especially the 3.4 - and I stumbled upon this article written for Total 911 back in July 2006:
http://www.autofarm.co.uk/pdf/Total911_July06.pdf
"First, it’s important to point out that it is the early 3.4- litre engines, made from 1998 to 2000, that are most susceptible to major failure."
Really?
Perhaps I'm getting my M96 problems mixed up here, but doesn't the 3.4 have a dual-row IMS, and furthermore, better cylinder wall lining material to ward off bore score?
How perceptions change.
http://www.autofarm.co.uk/pdf/Total911_July06.pdf
"First, it’s important to point out that it is the early 3.4- litre engines, made from 1998 to 2000, that are most susceptible to major failure."
Really?
Perhaps I'm getting my M96 problems mixed up here, but doesn't the 3.4 have a dual-row IMS, and furthermore, better cylinder wall lining material to ward off bore score?
How perceptions change.
Bore-scoring was more prevalent in the 3.6
The 3.4 was still susceptible to IMS failure but again, not as much as the 3.6
http://www.pistonheads.com/gassing/topic.asp?t=136...
The 3.4 was still susceptible to IMS failure but again, not as much as the 3.6
http://www.pistonheads.com/gassing/topic.asp?t=136...
fot0 said:
At the time of writing I guess the 996.2 had not had significant failures for it to be relevant.
This. Plus some 3.4s had faulty castings that showed themselves early on leading to a bad rep (D chunk cylinders / head cracks). Any with a few miles on should have shown themselves by now.Sorry I cannot offer a proven answer to these problems but I can offer the following logical collection of explanations that do fit all the evidence we have.W
hen the first M96 engines were designed (Boxster 2.5, 2.7 and 3.2) the cylinder wall thickness was thicker.
When they squeezed a bigger bore into the very same crankcase (same height and cylinder pitch etc), to make the 3.4 996 - they basically bored out the Boxster S casting another 3mm leaving the cylinder walls thinner (possibly also to maintain the same coolant volume around the cylinder) and some cracked.
This much is certain and indisputable - but we are not privy to what went on next and can only try and put together an explanation for events that followed based on experience of typical failures, published scanty information and try and find an explanation that fits what we have found happens in reality - so what follows is our conclusions but not really verifyable.
Stuck with the general sizes of everything - when they made the later even bigger engines they might have changed the mixture or proportion of the silicon in the Lokasil matrix that the cylinder block is made of because the manufacturers documentation lists Lokasil 1 and Lokasil 2 as different versions - the later one having slightly larger silicon particles embedded in it. The affect of this in a composite matrix would probably be to make the cylinder basically stronger and this makes sense if they were experiencing failures in the first versions.
If this coincided with the manufacture of the later models (which we cannot guarantee) it would explain two other issues - (1) that despite even more pressure on the same thickness of cylinder wall they do not appear to be cracking at such low mileages, (2) they do not seem to go as oval as soon.
However - if this is the case then it also contributes to another issue. Alloy cylinder bores with silicon deposits do inevitably gradually have some silicon particles break lose from the matrix to float or get trapped between the piston and the cylinder wall where they usually find a way out in the 924S, 944 and 968 variants through gravity before impinging into the piston face and before causing any damage. This is why the original Alusil Vega engine and all the original Alusil blocks and the first Lokasil blocks used alloy pistons with a plated ferrous coating that was hard enough to resist penetration and allowed the particles to escape in the oil film (although of course with horizontal cylinders in the M96/7 engines that may take a little longer). This is born out by the visible wear rates between the pistons in the previous engine types lasting longer in the 924S. 944 and 968 variants than the ferrous coated pistons in the M96 engines.
In looks like a greater proportion of silicon particles break lose from Lokasil (where they are enveloped by molten alloy during casting) than Alusil (where they come out of solution more evenly during casting and could be better secured). the distribution of silicon particles in the matrix also seems better in Alusil than Lokasil. The sizes of the silicon particles in Lokasil 2 are also greater than in Alusil and are probably large enough to cause a bigger problem when entrapped between the piston and cylinder wall within the oil film and the thinner that oil film the more problems they can cause.
Now the thickness of the oil film is a function of the viscosity and temperatures. In the 924S, 944 and 968 100% of the coolant passes into the cylinder block first and then into the heads. In the M96/7 engines something under 15% passes into the cylinder block (the rest directly into the heads) so the lower flow rate into the M96/7 cylinders could mean t hat the cooling is likely to be less and the thermostat is higher rated so they run at a higher cylinder temperature anyway.
The change to a softer and less well bonded plastic coating to the pistons in these later engines combined with the likelyhood of larger silicon particles, more torque from later engines creating more pressure between the piston and cylinder walls and higher oil and coolant temperatures in this area (and therefore thinner oil film) - all could contribute to cylinder scoring which is prevalent in the later models.
This is why a LTT is beneficial and why avoiding high torque at low revs (especially after a pause with a hot engine where heat soak increases internal engine temperatures) is good advice.
Nikasil does not have similar problems hence even pistons without any coating can run perfectly without scoring.
Baz
hen the first M96 engines were designed (Boxster 2.5, 2.7 and 3.2) the cylinder wall thickness was thicker.
When they squeezed a bigger bore into the very same crankcase (same height and cylinder pitch etc), to make the 3.4 996 - they basically bored out the Boxster S casting another 3mm leaving the cylinder walls thinner (possibly also to maintain the same coolant volume around the cylinder) and some cracked.
This much is certain and indisputable - but we are not privy to what went on next and can only try and put together an explanation for events that followed based on experience of typical failures, published scanty information and try and find an explanation that fits what we have found happens in reality - so what follows is our conclusions but not really verifyable.
Stuck with the general sizes of everything - when they made the later even bigger engines they might have changed the mixture or proportion of the silicon in the Lokasil matrix that the cylinder block is made of because the manufacturers documentation lists Lokasil 1 and Lokasil 2 as different versions - the later one having slightly larger silicon particles embedded in it. The affect of this in a composite matrix would probably be to make the cylinder basically stronger and this makes sense if they were experiencing failures in the first versions.
If this coincided with the manufacture of the later models (which we cannot guarantee) it would explain two other issues - (1) that despite even more pressure on the same thickness of cylinder wall they do not appear to be cracking at such low mileages, (2) they do not seem to go as oval as soon.
However - if this is the case then it also contributes to another issue. Alloy cylinder bores with silicon deposits do inevitably gradually have some silicon particles break lose from the matrix to float or get trapped between the piston and the cylinder wall where they usually find a way out in the 924S, 944 and 968 variants through gravity before impinging into the piston face and before causing any damage. This is why the original Alusil Vega engine and all the original Alusil blocks and the first Lokasil blocks used alloy pistons with a plated ferrous coating that was hard enough to resist penetration and allowed the particles to escape in the oil film (although of course with horizontal cylinders in the M96/7 engines that may take a little longer). This is born out by the visible wear rates between the pistons in the previous engine types lasting longer in the 924S. 944 and 968 variants than the ferrous coated pistons in the M96 engines.
In looks like a greater proportion of silicon particles break lose from Lokasil (where they are enveloped by molten alloy during casting) than Alusil (where they come out of solution more evenly during casting and could be better secured). the distribution of silicon particles in the matrix also seems better in Alusil than Lokasil. The sizes of the silicon particles in Lokasil 2 are also greater than in Alusil and are probably large enough to cause a bigger problem when entrapped between the piston and cylinder wall within the oil film and the thinner that oil film the more problems they can cause.
Now the thickness of the oil film is a function of the viscosity and temperatures. In the 924S, 944 and 968 100% of the coolant passes into the cylinder block first and then into the heads. In the M96/7 engines something under 15% passes into the cylinder block (the rest directly into the heads) so the lower flow rate into the M96/7 cylinders could mean t hat the cooling is likely to be less and the thermostat is higher rated so they run at a higher cylinder temperature anyway.
The change to a softer and less well bonded plastic coating to the pistons in these later engines combined with the likelyhood of larger silicon particles, more torque from later engines creating more pressure between the piston and cylinder walls and higher oil and coolant temperatures in this area (and therefore thinner oil film) - all could contribute to cylinder scoring which is prevalent in the later models.
This is why a LTT is beneficial and why avoiding high torque at low revs (especially after a pause with a hot engine where heat soak increases internal engine temperatures) is good advice.
Nikasil does not have similar problems hence even pistons without any coating can run perfectly without scoring.
Baz
delays said:
I've been looking at early 996s recently - especially the 3.4 - and I stumbled upon this article written for Total 911 back in July 2006:
http://www.autofarm.co.uk/pdf/Total911_July06.pdf
"First, it’s important to point out that it is the early 3.4- litre engines, made from 1998 to 2000, that are most susceptible to major failure."
Really?
Perhaps I'm getting my M96 problems mixed up here, but doesn't the 3.4 have a dual-row IMS, and furthermore, better cylinder wall lining material to ward off bore score?
How perceptions change.
4th paragraph in:http://www.autofarm.co.uk/pdf/Total911_July06.pdf
"First, it’s important to point out that it is the early 3.4- litre engines, made from 1998 to 2000, that are most susceptible to major failure."
Really?
Perhaps I'm getting my M96 problems mixed up here, but doesn't the 3.4 have a dual-row IMS, and furthermore, better cylinder wall lining material to ward off bore score?
How perceptions change.
"The later 3.6-litre engine is pretty much problem-free."
Tells you all you need to know about the credibility of the article.
The early 3.4's are more susceptible to IMS failures and cracked "D" chunked cylinders (only in small numbers) but not bore scoring.
The early 2000 3.4's. and 3.6's may score bores (although with lower power output than a 3.8 and because they tend to be bought over a 3.8 by those not needing ultimate performance - tend to last longer) and still have an older IMS bearing.
Later 3.6's. 3.8's and Cayman S's have the later more reliable larger IMS bearing but plastic coated pistons and may score bores.
Incidences of failures - while low - are still more than would be expected of the marque - although in common with many (if not most) modern high performance sports car engines manufactured by modern techniques and materials are not as reliable as their grossly over engineered predecessors that created the superb reputation but almost resulted in bankruptcy.
Because they are such brilliant cars that maintain their desirability for decades and eventually appreciate - for all models an engine rebuild at some stage becomes an accepted feature of long term ownership - this model range just a little premature compared to many others.
Baz
The early 2000 3.4's. and 3.6's may score bores (although with lower power output than a 3.8 and because they tend to be bought over a 3.8 by those not needing ultimate performance - tend to last longer) and still have an older IMS bearing.
Later 3.6's. 3.8's and Cayman S's have the later more reliable larger IMS bearing but plastic coated pistons and may score bores.
Incidences of failures - while low - are still more than would be expected of the marque - although in common with many (if not most) modern high performance sports car engines manufactured by modern techniques and materials are not as reliable as their grossly over engineered predecessors that created the superb reputation but almost resulted in bankruptcy.
Because they are such brilliant cars that maintain their desirability for decades and eventually appreciate - for all models an engine rebuild at some stage becomes an accepted feature of long term ownership - this model range just a little premature compared to many others.
Baz
hartech said:
The early 3.4's are more susceptible to IMS failures and cracked "D" chunked cylinders (only in small numbers) but not bore scoring.
The early 2000 3.4's. and 3.6's may score bores (although with lower power output than a 3.8 and because they tend to be bought over a 3.8 by those not needing ultimate performance - tend to last longer) and still have an older IMS bearing.
Later 3.6's. 3.8's and Cayman S's have the later more reliable larger IMS bearing but plastic coated pistons and may score bores.
I have a 2003 C4S 3.6 with 56,000 miles on the clock. I love it and care for it especially when warming it up correctly before giving it the beans. The early 2000 3.4's. and 3.6's may score bores (although with lower power output than a 3.8 and because they tend to be bought over a 3.8 by those not needing ultimate performance - tend to last longer) and still have an older IMS bearing.
Later 3.6's. 3.8's and Cayman S's have the later more reliable larger IMS bearing but plastic coated pistons and may score bores.
This weeks 911uk forum has dented my confidence - three engines in a week have gone. I really don't fancy a £12K rebuild and I want to enjoy what is a fantastic car but I wondering if to have the IMS done and another borescope just in case. The car was a ex 9E car and every engineer I have spoken to who has worked on it from 9E to 911Virgin told me it is a good'un but I am a bit jittery !
The 996 is such an under rated car due to these POTENTIAL issues - it is just such a pity. On the bright side the ones left that are in decent condition, well maintained and not over modded will continue to rise in value.
DeadInside said:
This weeks 911uk forum has dented my confidence - three engines in a week have gone.
Ive seen those three examples too, horrible experience for the owners but i think they are very different again. One of them had a recent questionable rebuild already? The other one was bore-scoring, and other is dodgy valve giving many signs but useless OPC and indies could not figure out, lol(I think?) and three of them over 10 years old, 100k+ cars.These cars values are so low nowadays I personally don't see any reason for a worry.
I should also stop looking at 993s, get a very shagged 997 gen1 and rebuild the engine properly with all upgrades then relax for the next 15 years, really much better value. Who knows!
Gassing Station | 911/Carrera GT | Top of Page | What's New | My Stuff