HEADGASKET dissertation...
09-26-2002, 02:39 AM
#31
Addict
Rennlist Member
Rennlist Member
I agree that water injection in itself won't increase HP. If you are running enough octane that the computer is not retarding the timing due to knock. Then you're good. But, for most of that, octane is hard to come by. Racing gas is $4-$5 per gallon.
I see water injection doing 2 things. Replacing the need for high octane gas (100+) in most cases, thereby keeping the computer from retarding the time, and cooling the intake charge that becomes heat soaked over time.
Also, The water never makes it to the engine (or shouldn't). It vaporizes the moment it hits the intake charge, absorbing the heat/energy and cooling the air/fuel mixture that does go into the cylinder.
Here are a couple sites on water injection.
<a href="http://www.frii.com/~maphill/wi.html" target="_blank">http://www.frii.com/~maphill/wi.html</a>
<a href="http://www.kennedysdynotune.com/water_injection.htm" target="_blank">http://www.kennedysdynotune.com/water_injection.htm</a>
<a href="http://www.racetep.com/wik.html#trbh20" target="_blank">http://www.racetep.com/wik.html#trbh20</a>
<a href="http://www.syclone.freeserve.co.uk/waterinjection.htm" target="_blank">http://www.syclone.freeserve.co.uk/waterinjection.htm</a>
I know that with what I spent on race gas this year, I could have paid for 1/2 a water injection kit.
My 2cents.
I see water injection doing 2 things. Replacing the need for high octane gas (100+) in most cases, thereby keeping the computer from retarding the time, and cooling the intake charge that becomes heat soaked over time.
Also, The water never makes it to the engine (or shouldn't). It vaporizes the moment it hits the intake charge, absorbing the heat/energy and cooling the air/fuel mixture that does go into the cylinder.
Here are a couple sites on water injection.
<a href="http://www.frii.com/~maphill/wi.html" target="_blank">http://www.frii.com/~maphill/wi.html</a>
<a href="http://www.kennedysdynotune.com/water_injection.htm" target="_blank">http://www.kennedysdynotune.com/water_injection.htm</a>
<a href="http://www.racetep.com/wik.html#trbh20" target="_blank">http://www.racetep.com/wik.html#trbh20</a>
<a href="http://www.syclone.freeserve.co.uk/waterinjection.htm" target="_blank">http://www.syclone.freeserve.co.uk/waterinjection.htm</a>
I know that with what I spent on race gas this year, I could have paid for 1/2 a water injection kit.
My 2cents.
09-26-2002, 04:07 PM
#33
Instructor
Join Date: Sep 2002
Location: Selden, NY
Posts: 131
Likes: 0
Received 0 Likes
on
0 Posts
Hey,
One important difference between the WFR and the
stock gasket is in the thickness near the water
passage:
[img]d:\mystuff\hg.jpg[/img]
If you look at the area I circled, you can see
how close the water passage is to the gasket.
On my car, the stock gasket got old and brittle.
It did not burn up (my engine is stock) but
instead got rotten and allowed combustion gases
to enter the water jacket.
The WFR is much thicker in that area and should
last a lot longer.
I have seen three bad head gaskets on 944 turbos,
and all were bad in the same area, #4 cylinder,
right at that water passage.
Barry
One important difference between the WFR and the
stock gasket is in the thickness near the water
passage:
[img]d:\mystuff\hg.jpg[/img]
If you look at the area I circled, you can see
how close the water passage is to the gasket.
On my car, the stock gasket got old and brittle.
It did not burn up (my engine is stock) but
instead got rotten and allowed combustion gases
to enter the water jacket.
The WFR is much thicker in that area and should
last a lot longer.
I have seen three bad head gaskets on 944 turbos,
and all were bad in the same area, #4 cylinder,
right at that water passage.
Barry
09-26-2002, 05:52 PM
#34
Instructor
Join Date: Apr 2002
Location: Stockholm-Sweden
Posts: 192
Likes: 0
Received 0 Likes
on
0 Posts
A very big thank you from Sweden! This is what I Suspected since I blew a head valve on my Previuos 944turbo-91..
I have now runned my car for 5000km and the gasket is still ok. I have run approx 1.3bar=19psi this period.
I tuned my car on the dyno and adjusted my AFM and my fuel pressuare so my A/F seemed ok but a little bit rich!
Regards
Pete
Thank´s again!
I have now runned my car for 5000km and the gasket is still ok. I have run approx 1.3bar=19psi this period.
I tuned my car on the dyno and adjusted my AFM and my fuel pressuare so my A/F seemed ok but a little bit rich!
Regards
Pete
Thank´s again!
09-26-2002, 08:00 PM
#35
Race Director
Thread Starter
"One question: you mention the 3D correction maps that are applied after the initial duty cycle number is computed. What purpose do these maps serve and how are they chosen?"
That's because the amount of air that's actually entering the engine through the intake-valves is not the same amount that's being measured at the air-flow meter. Check out this thread for an air-pressure vs. density map as the air flows through the engine: <a href="http://forums.rennlist.com/forums/ultimatebb.php?ubb=get_topic&f=18&t=002110" target="_blank"> Topic: A bit rich...but my buddy's CBR600 can't keep up!</a>. Some people call this VE-volumetric efficiency which can be measured as a ratio of actual air-volume ingested at STP vs. displacement. This ratio changes based upon intake-tract restrictions, cam-timing & lift, combustion-chamber shape, RPM, headers & exhaust, etc. Depending upon these factors, you may get 90% VE at low-RPMs, but only 60% at high-RPMs (or vice versa), yet the air-flow meter is registering the same flow. So the correction maps are to make adjustments based upon the elasticity of air and how it's stretched and squeezed as it moves through the engine.
" I'm curious why a lot of smart 951 owners and well respected engine builders who are creating big horsepower have gone through the effort and expense of O-ringing if A/F ratios are the critical element for a good head seal. Is this undertaking a safety backup to lean conditions which may cause premature HG failure or is there more of a functional mechanical benefit provided by O-rings in engines running higher boost?"
There's another issue that's part of the headgasket failure equation, but I can't place a percentage of blame on it because I'm still trying to figure out how significant it is. That's flexing of the cylinders due to the open deck. But it does contribute because Porsche changed the block's design later to compensate. Hondas use this style of block and a very common modification is to install a block-brace to connect the tops of the cylinders to the outside block casing. This reduces the flexing of the cylinders when increasing the compression, or converting to forced-induction. While this is such a common mod, that pre-fitted CNC'ed block-braces that pop right in are available for Hondas, our blocks would require a custom-made one welded in like <a href="http://www.928sg.com/block.htm" target="_blank"> on this 928 engine</a>. If you look at the later 3.0L blocks, they have siamesed cylinders with extra material casted in place to connect the tops of the cylinders.
I suspect that there's a certain amount of fretting going on at the headgasket junction. This would cause a rocking back & forth motion that would open up a gap between the headgasket and cylinders. So the role of tongue & groove O-ringing that's done by a lot of the engine-builders is to pin the tops of the cylinders in place.
"It did not burn up (my engine is stock) but instead got rotten and allowed combustion gases to enter the water jacket."
The first thing separating your combustion-chamber from the rest of the headgasket and water-jacket is the steel compression-ring. If combustion gases got past the compression-ring, you've already got a failure. Either the head-studs don't have enough or lost tension, or the compression-ring's seal has been compromised somehow. At that point, it's irrelevant what the condition of the fiber part of the gasket is, because combustion-pressure is way higher than the 15psi or so that the cooling-system is under.
That's because the amount of air that's actually entering the engine through the intake-valves is not the same amount that's being measured at the air-flow meter. Check out this thread for an air-pressure vs. density map as the air flows through the engine: <a href="http://forums.rennlist.com/forums/ultimatebb.php?ubb=get_topic&f=18&t=002110" target="_blank"> Topic: A bit rich...but my buddy's CBR600 can't keep up!</a>. Some people call this VE-volumetric efficiency which can be measured as a ratio of actual air-volume ingested at STP vs. displacement. This ratio changes based upon intake-tract restrictions, cam-timing & lift, combustion-chamber shape, RPM, headers & exhaust, etc. Depending upon these factors, you may get 90% VE at low-RPMs, but only 60% at high-RPMs (or vice versa), yet the air-flow meter is registering the same flow. So the correction maps are to make adjustments based upon the elasticity of air and how it's stretched and squeezed as it moves through the engine.
" I'm curious why a lot of smart 951 owners and well respected engine builders who are creating big horsepower have gone through the effort and expense of O-ringing if A/F ratios are the critical element for a good head seal. Is this undertaking a safety backup to lean conditions which may cause premature HG failure or is there more of a functional mechanical benefit provided by O-rings in engines running higher boost?"
There's another issue that's part of the headgasket failure equation, but I can't place a percentage of blame on it because I'm still trying to figure out how significant it is. That's flexing of the cylinders due to the open deck. But it does contribute because Porsche changed the block's design later to compensate. Hondas use this style of block and a very common modification is to install a block-brace to connect the tops of the cylinders to the outside block casing. This reduces the flexing of the cylinders when increasing the compression, or converting to forced-induction. While this is such a common mod, that pre-fitted CNC'ed block-braces that pop right in are available for Hondas, our blocks would require a custom-made one welded in like <a href="http://www.928sg.com/block.htm" target="_blank"> on this 928 engine</a>. If you look at the later 3.0L blocks, they have siamesed cylinders with extra material casted in place to connect the tops of the cylinders.
I suspect that there's a certain amount of fretting going on at the headgasket junction. This would cause a rocking back & forth motion that would open up a gap between the headgasket and cylinders. So the role of tongue & groove O-ringing that's done by a lot of the engine-builders is to pin the tops of the cylinders in place.
"It did not burn up (my engine is stock) but instead got rotten and allowed combustion gases to enter the water jacket."
The first thing separating your combustion-chamber from the rest of the headgasket and water-jacket is the steel compression-ring. If combustion gases got past the compression-ring, you've already got a failure. Either the head-studs don't have enough or lost tension, or the compression-ring's seal has been compromised somehow. At that point, it's irrelevant what the condition of the fiber part of the gasket is, because combustion-pressure is way higher than the 15psi or so that the cooling-system is under.
10-01-2002, 06:50 AM
#36
Race Director
Thread Starter
Ok, a quick follow-up because someone brought up a good point:
"The WFR gasket is laterally more rigid due to the "u" shape, like a
construction "I" beam, and because it's wider, it can transfer more heat to
the clamping surfaces. Simply put: It will handle pressure and heat better."
That's only true if the heat-transfer is perpendicular to the headgasket. Like from the fiber inside the gasket through the wide clamping surface directly upwards & downwards to the block/head. Even then, only the bottom surface to the block makes a difference since that's the only surface that's different.
However, since the heat is originating from inside the cylinder, it has to travel from the point of the U shape sideways & outwards through the legs before arriving at the block/head surfaces. Since the heat-transfer equation is:
watts = [kA(dT)] / d
where: k is thermal conductivity to transmission medium
A is cross-sectional area of transmission medium
dT is delta of temperature gradient (T1 - T2)
d is distance heat has to travel
Three out of four of these variables are known. We know the thermal conductivity of steel in the headgasket compression-ring. We also know that the thermal conductivity of surrounding aluminum is about 4x faster. We know the delta-T from the heat of combustion and the water-jacket temperature. We also know how the distance from the tip of U on both the standard and wide firering headgaskets fo the head & cylinders. All the above factors are identical in both cases.
What is up to question is what path does the heat take? I assert that they are both the same. Please refer to this <a href="http://boards.rennlist.com/upload/Headgasket-HeatFlow.gif" target="_blank">rather large heat-flow diagram</a> (open in new window).
As I stated earlier, the heat does not flow perpendicular to the headgasket through the entire contact surface at the tops and bottoms (blue lines). But rather it starts at the edge of the compression-ring that faces the combustion chamber and flows sideways (yellow lines). Since the amount of heat transferred is inversely proportional to the distance traveled, the outer edges of the compression-ring gets heated up the least. This is also happens to be the only area that is different between the two headgaskets as well, and it receives the least amount of heat.
Since aluminum conducts heat 4x faster than steel, once the heat reaches the aluminum mating surfaces of the head and cylinder-walls, it's wicked away very,very quickly (also do to much larger cross-sectional area of head/cylinders for conduction path). By the time you get even halfway to the outer edges of either headgasket, the compression-ring will be about the same temperature as the head & cylinders.
Conversely the part that is heated up the most, is the point furthest away from the cool head & cylinder surfaces. This is the center of the U-shaped edge of the compression-ring. Then if you analyze the headgaskets that have 'blown', you'll see that the compression-ring is rarely ever disfigured from its perfectly round shape (with or without O-ringing). But rather, its the very inside edge that's completely burned away leaving two flat rings on the top and bottom; which can still conduct heat to the head and cylinder walls PERFECTLY by the way.
I still say that the best headgasket to use would be a copper headgasket with tongue & groove O-ringing. Copper conducts heat 8x times faster than steel and 2x faster than aluminum. So a cooper headgasket would immediately transfer ALL of the heat it faces to the mating aluminum heat/cylinder surfaces. And the solid copper provides a much thicker conduction path of it's entire thickness 1.5mm vs. 0.1mm for the thin/hollow steel compression-ring.
However durable it may be, a copper headgasket (or wide fire-ring for that matter) can not fix the actual cause of burnt headgaskets either. Which is an incorrect air-fuel ratio (too lean much worse than too rich). In which case, the weakest link in your engine will just move from the headgasket to the rings or valves (much costlier to repair than headgasket).
And another worthwhile mod to investigate would be a block-brace to close up the open-deck. There must be a reason Porsche connected the cylinders in the later 3.0L blocks (non-turbo at that). Anyone had failure-rates of 3.0L Turbo engines vs. 951 engines?
I also agree with Chris that the mating surfaces has to be ABSOLUTELY clean for the best heat conduction possible.
"The WFR gasket is laterally more rigid due to the "u" shape, like a
construction "I" beam, and because it's wider, it can transfer more heat to
the clamping surfaces. Simply put: It will handle pressure and heat better."
That's only true if the heat-transfer is perpendicular to the headgasket. Like from the fiber inside the gasket through the wide clamping surface directly upwards & downwards to the block/head. Even then, only the bottom surface to the block makes a difference since that's the only surface that's different.
However, since the heat is originating from inside the cylinder, it has to travel from the point of the U shape sideways & outwards through the legs before arriving at the block/head surfaces. Since the heat-transfer equation is:
watts = [kA(dT)] / d
where: k is thermal conductivity to transmission medium
A is cross-sectional area of transmission medium
dT is delta of temperature gradient (T1 - T2)
d is distance heat has to travel
Three out of four of these variables are known. We know the thermal conductivity of steel in the headgasket compression-ring. We also know that the thermal conductivity of surrounding aluminum is about 4x faster. We know the delta-T from the heat of combustion and the water-jacket temperature. We also know how the distance from the tip of U on both the standard and wide firering headgaskets fo the head & cylinders. All the above factors are identical in both cases.
What is up to question is what path does the heat take? I assert that they are both the same. Please refer to this <a href="http://boards.rennlist.com/upload/Headgasket-HeatFlow.gif" target="_blank">rather large heat-flow diagram</a> (open in new window).
As I stated earlier, the heat does not flow perpendicular to the headgasket through the entire contact surface at the tops and bottoms (blue lines). But rather it starts at the edge of the compression-ring that faces the combustion chamber and flows sideways (yellow lines). Since the amount of heat transferred is inversely proportional to the distance traveled, the outer edges of the compression-ring gets heated up the least. This is also happens to be the only area that is different between the two headgaskets as well, and it receives the least amount of heat.
Since aluminum conducts heat 4x faster than steel, once the heat reaches the aluminum mating surfaces of the head and cylinder-walls, it's wicked away very,very quickly (also do to much larger cross-sectional area of head/cylinders for conduction path). By the time you get even halfway to the outer edges of either headgasket, the compression-ring will be about the same temperature as the head & cylinders.
Conversely the part that is heated up the most, is the point furthest away from the cool head & cylinder surfaces. This is the center of the U-shaped edge of the compression-ring. Then if you analyze the headgaskets that have 'blown', you'll see that the compression-ring is rarely ever disfigured from its perfectly round shape (with or without O-ringing). But rather, its the very inside edge that's completely burned away leaving two flat rings on the top and bottom; which can still conduct heat to the head and cylinder walls PERFECTLY by the way.
I still say that the best headgasket to use would be a copper headgasket with tongue & groove O-ringing. Copper conducts heat 8x times faster than steel and 2x faster than aluminum. So a cooper headgasket would immediately transfer ALL of the heat it faces to the mating aluminum heat/cylinder surfaces. And the solid copper provides a much thicker conduction path of it's entire thickness 1.5mm vs. 0.1mm for the thin/hollow steel compression-ring.
However durable it may be, a copper headgasket (or wide fire-ring for that matter) can not fix the actual cause of burnt headgaskets either. Which is an incorrect air-fuel ratio (too lean much worse than too rich). In which case, the weakest link in your engine will just move from the headgasket to the rings or valves (much costlier to repair than headgasket).
And another worthwhile mod to investigate would be a block-brace to close up the open-deck. There must be a reason Porsche connected the cylinders in the later 3.0L blocks (non-turbo at that). Anyone had failure-rates of 3.0L Turbo engines vs. 951 engines?
I also agree with Chris that the mating surfaces has to be ABSOLUTELY clean for the best heat conduction possible.
05-20-2013, 03:30 AM
#39
Addict
Lifetime Rennlist
Member
Lifetime Rennlist
Member
