Cross over Backpressure?
06-08-2015, 08:20 AM
#32
Rennlist Member
Thread Starter
Did a little surfing and found that a 150hp 4 cylinder has a peak cylinder pressure of 600psi. So approx multiply that by four and we have 2400psi. Makes me wonder how 45psi of backpressure can have such a deleterious effect?
Harvard did this paper on cylinder pressure, it's an NA Renault engine. Its peak is 735psi.
http://www.hcs.harvard.edu/~jus/0303/kuo.pdf
ps I don't for one minute pretend to understand all that, just thought I'd put it there.
Harvard did this paper on cylinder pressure, it's an NA Renault engine. Its peak is 735psi.
http://www.hcs.harvard.edu/~jus/0303/kuo.pdf
ps I don't for one minute pretend to understand all that, just thought I'd put it there.
06-08-2015, 08:47 AM
#33
Three Wheelin'
One thing is combustion pressure, but during exhaust stroke the pressure is very low in cylinder and exhaust pressure pushes exhaust gases to cylinder and thn during intake stroke there is exhaust gas residues which make combustion hotter and leave less room for air.
06-08-2015, 09:20 AM
#34
Nordschleife Master
Let's not try to re-invent the wheel here with moving the turbo or going VGT. I would never ever consider a VGT turbo until they are available in the correct size and proven for a petrol engine in a race environment.
The 0.63 housing is a bit mismatched on the large GTX3582R. The simple facts are 1) Trying to get over 600 crank hp requires a somewhat large turbo. 2) Large overlap cams will hurt more with more backpressue = engine designed for large, low back pressure, turbo. 3) 600+ hp from a 2.5l 8v engine will need high rpms and will always be a bit laggy in the low end.
Trying to stay with the same power goals while fixing low end torque and response will end up in a non optimal compromise. IMHO there are only 2 ways to go on this engine. A) Lower power expectations and run a smaller turbo. B) Optimize for top end power and increase powerband upwards.
The 0.63 housing is a bit mismatched on the large GTX3582R. The simple facts are 1) Trying to get over 600 crank hp requires a somewhat large turbo. 2) Large overlap cams will hurt more with more backpressue = engine designed for large, low back pressure, turbo. 3) 600+ hp from a 2.5l 8v engine will need high rpms and will always be a bit laggy in the low end.
Trying to stay with the same power goals while fixing low end torque and response will end up in a non optimal compromise. IMHO there are only 2 ways to go on this engine. A) Lower power expectations and run a smaller turbo. B) Optimize for top end power and increase powerband upwards.
06-08-2015, 09:46 AM
#35
You have a big cam and big valves etc for the purpose of increasing VE, especially for higher RPM so you can generate higher HP.
However, if at the same time you dramatically increase back pressure, your cam and head can actually work against you.
Peak cylinder pressure typically occurs just after TDC on the power stroke whereas the exhaust valve starts to open just towards the end of that stroke so those peak pressure are quite different to what you would expect to see as back pressure and you would not add them together as the cylinders are not firing simultaneously and the exhaust gases are (hopefully) moving.
It can be advantageous for there to be some overlap where the exhaust valve is still partially open as the intake valve is beginning to open as this can assist is the expulsion of the exhaust gases meaning more space is available for fresh intake air and fuel and ultimately higher VE. The bigger cam likely has more overlap to enhance VE higher in the rpm range, this is also assisted by bigger valves. Some of the intake air may also actually flow through the exhaust valves helping to cool them.
However, if your back pressure is high, the reverse happens... you have reversion and the exhaust gases actually push back your intake air not only limiting cylinder filling but also resulting in a much higher temperature within the cylinder with all it's associated problems - potentially, hot spots and irregular burn etc which could be the culprit in your head lift issue since there was no mention of this with the .82 housing which would have had lower back pressure. If you look at your TPS data plot, it seems to swing violently from around 100% to 60% as back pressure increases - I question whether this may be physical evidence of reversion at work? If so, it certainly looks quite dramatic.
I think you have to take into account what the actual design parameters of this engine... in particular, your initial post says:
The data shows that the head won't go into choke up to and possibly over 8000rpm where our last chart shows how it fell over at just over 6000rpm.
It seems that a good deal of your lag issues may have been addressed independently of the change to the .63 hot side whereas the data presented seems to suggest the .63 is not well matched to the design parameters and is really asking too much of the engine.
I suspect that going back to the .82 would probably get you much closer to the power you had with the earlier engine and probably without too much extra lag than you currently have now that the other issues have been addressed. Perhaps the .82 hot side was more of a limiting factor than the head/cam combination in that engine? I'm guessing you'd probably need the 1.06 if you wanted to get closer to 600.
However, if at the same time you dramatically increase back pressure, your cam and head can actually work against you.
Peak cylinder pressure typically occurs just after TDC on the power stroke whereas the exhaust valve starts to open just towards the end of that stroke so those peak pressure are quite different to what you would expect to see as back pressure and you would not add them together as the cylinders are not firing simultaneously and the exhaust gases are (hopefully) moving.
It can be advantageous for there to be some overlap where the exhaust valve is still partially open as the intake valve is beginning to open as this can assist is the expulsion of the exhaust gases meaning more space is available for fresh intake air and fuel and ultimately higher VE. The bigger cam likely has more overlap to enhance VE higher in the rpm range, this is also assisted by bigger valves. Some of the intake air may also actually flow through the exhaust valves helping to cool them.
However, if your back pressure is high, the reverse happens... you have reversion and the exhaust gases actually push back your intake air not only limiting cylinder filling but also resulting in a much higher temperature within the cylinder with all it's associated problems - potentially, hot spots and irregular burn etc which could be the culprit in your head lift issue since there was no mention of this with the .82 housing which would have had lower back pressure. If you look at your TPS data plot, it seems to swing violently from around 100% to 60% as back pressure increases - I question whether this may be physical evidence of reversion at work? If so, it certainly looks quite dramatic.
I think you have to take into account what the actual design parameters of this engine... in particular, your initial post says:
The data shows that the head won't go into choke up to and possibly over 8000rpm where our last chart shows how it fell over at just over 6000rpm.
It seems that a good deal of your lag issues may have been addressed independently of the change to the .63 hot side whereas the data presented seems to suggest the .63 is not well matched to the design parameters and is really asking too much of the engine.
I suspect that going back to the .82 would probably get you much closer to the power you had with the earlier engine and probably without too much extra lag than you currently have now that the other issues have been addressed. Perhaps the .82 hot side was more of a limiting factor than the head/cam combination in that engine? I'm guessing you'd probably need the 1.06 if you wanted to get closer to 600.
06-08-2015, 02:12 PM
#36
Rennlist Member
Patrick,
You have a good feel for the issues. And your question as to how much impact the 40, or so PSI bp impacts the cycle is directionally correct.
To understand the impact of exhaust pressure, overlay the position of the piston within the bore, and the valve positions. It becomes obvious that the BP is working on a small volume as the piston is near the top of the exhaust stroke. And the intake boost overpressures the entire down stroke of the intake cycle.
As other have said, matching the flow map of the turbo is desired to have the turbo at the peak efficiency for desired power (CFM flow). Yielding the best boost versus BP ratio.
You have a good feel for the issues. And your question as to how much impact the 40, or so PSI bp impacts the cycle is directionally correct.
To understand the impact of exhaust pressure, overlay the position of the piston within the bore, and the valve positions. It becomes obvious that the BP is working on a small volume as the piston is near the top of the exhaust stroke. And the intake boost overpressures the entire down stroke of the intake cycle.
As other have said, matching the flow map of the turbo is desired to have the turbo at the peak efficiency for desired power (CFM flow). Yielding the best boost versus BP ratio.
06-09-2015, 01:30 AM
#37
Burning Brakes
"Makes me wonder how 45psi of backpressure can have such a deleterious effect?"
All good answers above, but I like to look at it this way; Exhaust backpressure is a way to measure how much exhaust gas remains in the cylinder after the exhaust valve closes. It would be best to remove all residual exhaust gas at the end of the exhaust stroke (100% VE) but there's always some leftover. Simply lowering the exhaust BP will improve VE, assuming all else is equal.
On another note, having high exhaust BP leads to an EGR effect at WOT. The charge temps go up due to mixing with hot exhaust gas, so the engine has a greater tendency to knock. But since exhaust gas is inert, cannot burn, and since it displaces the fresh air/fuel charge it actually lowers the temperature of the burn after being ignited, leading to slower burn rates and reduced power. Another oddity is that an engine with high exhaust BP and/or poor exhaust scavenging will make better power with high timing. (sound familiar?)
All good answers above, but I like to look at it this way; Exhaust backpressure is a way to measure how much exhaust gas remains in the cylinder after the exhaust valve closes. It would be best to remove all residual exhaust gas at the end of the exhaust stroke (100% VE) but there's always some leftover. Simply lowering the exhaust BP will improve VE, assuming all else is equal.
On another note, having high exhaust BP leads to an EGR effect at WOT. The charge temps go up due to mixing with hot exhaust gas, so the engine has a greater tendency to knock. But since exhaust gas is inert, cannot burn, and since it displaces the fresh air/fuel charge it actually lowers the temperature of the burn after being ignited, leading to slower burn rates and reduced power. Another oddity is that an engine with high exhaust BP and/or poor exhaust scavenging will make better power with high timing. (sound familiar?)
06-09-2015, 10:44 AM
#38
Rennlist Member
I like to evaluate any problem from two independent directions to see if the result compares favorably. So to verify cylinder pressure ranges we look to output.
So let's start at the crankshaft, we make say, 350 ft pounds of torque. For each revolution of the crank, there are two power strokes. Too make it simple; use 4" diameter piston and 3" stroke, 1/4 ft.
So each piston makes half of the torque, 175 ft lbs, at 1/4 ft length, then average pressure is 700 pounds.
The piston is roughly 12.5 sq inches, so 700/12.5=56 PSI average. The effective portion of the power stroke is from about 30 degrees, peaking at 90 degrees and on back to 150 degrees (crankshaft degrees).
The peak pressure at 90 degrees crank is about .7 piston pressure due to connecting rod angles. So for the full 360 degrees the net average is around .1 of piston pressure contributes to torque= 56/.1=560 PSI. That is output pressure to torque. Allow 40 percent internal losses and 560/.6=933 PSI.
Working back from the crank at 350 Ft Lbs, we need around 1,000 PSI peak cylinder pressure. So the BP of 40-45 is acceptable, less would be great.
So let's start at the crankshaft, we make say, 350 ft pounds of torque. For each revolution of the crank, there are two power strokes. Too make it simple; use 4" diameter piston and 3" stroke, 1/4 ft.
So each piston makes half of the torque, 175 ft lbs, at 1/4 ft length, then average pressure is 700 pounds.
The piston is roughly 12.5 sq inches, so 700/12.5=56 PSI average. The effective portion of the power stroke is from about 30 degrees, peaking at 90 degrees and on back to 150 degrees (crankshaft degrees).
The peak pressure at 90 degrees crank is about .7 piston pressure due to connecting rod angles. So for the full 360 degrees the net average is around .1 of piston pressure contributes to torque= 56/.1=560 PSI. That is output pressure to torque. Allow 40 percent internal losses and 560/.6=933 PSI.
Working back from the crank at 350 Ft Lbs, we need around 1,000 PSI peak cylinder pressure. So the BP of 40-45 is acceptable, less would be great.
06-09-2015, 04:39 PM
#39
Rennlist Member
Thread Starter
Our rods are longer and pistons shorter Alan. That is their motion would have more 'sympathetic' path than stock. Plus we have over 400ft/lbs tq to the wheels. Still the same BP. If we're running significantly more combustion pressure 45psi vs 2400psi I am still unsure how it can have an overwhelming effect. Also our hp didn't drop dramatically from the previous setup which included a much larger Xover and the next size up hot housing.
06-09-2015, 05:32 PM
#40
I am curious about the 2:1 ratio suggested. I was under the impression that the closer to 1:1 ratio one can get the better regardless of application.
Hopefully someone who knows could explain more on the subject.
I recently and finally read a book I bought on high performance engine building that had some informative information on peak cylinder pressure and back pressure.
Talked about building your bottom end to withstand X amount cylinder pressure.
My lower back pressure number is due to the design of the hot housing on my car. This is what I understand from the reading I have found so far. Tangential housings apparently typically have this characteristic.
I put down 444.89 ft/lbs at 15 psi base line with the Mafterburner piggyback on pump gas and between 17 to 18psi back pressure.
Will be going to the same dyno when I get the Rogue DME and tuner/logger.
Kinda apples to oranges compared to your setup but thought I would share
Good topic you brought up.
Hopefully someone who knows could explain more on the subject.
I recently and finally read a book I bought on high performance engine building that had some informative information on peak cylinder pressure and back pressure.
Talked about building your bottom end to withstand X amount cylinder pressure.
My lower back pressure number is due to the design of the hot housing on my car. This is what I understand from the reading I have found so far. Tangential housings apparently typically have this characteristic.
I put down 444.89 ft/lbs at 15 psi base line with the Mafterburner piggyback on pump gas and between 17 to 18psi back pressure.
Will be going to the same dyno when I get the Rogue DME and tuner/logger.
Kinda apples to oranges compared to your setup but thought I would share
Good topic you brought up.
06-09-2015, 05:35 PM
#41
Rennlist Member
Thread Starter
Almost 450ft/lbs at the wheels at one bar on pump? Wow! What are the motor's specs again?
06-09-2015, 06:04 PM
#42
Nordschleife Master
Guys, all valves are obviously closed during combustion phase
So back pressure does not work against peak pressure.
The issue is with backpressure and lots of overlap on the intake and exhaust cam.
Old school turbo engines (and factory engines) have high BP and narrow cams with basically no overlap. Modern, fully built, turbo engines can be built more like a traditional high hp N/A engine. In a successful N/A engine the pulses in the headers can help draw fresh air into the cylinders and they can reach above 100% VE. In those engines high overlap is part of the function.
Now in Patricks engine there's a very, very, aggressive cam with lots of overlap. This means that a lot of backpressure will for a period of time work against the filling of the cylinders. I would be careful of running this cam with the reported back pressure. Much higher risk of detonation.
So back pressure does not work against peak pressure.
The issue is with backpressure and lots of overlap on the intake and exhaust cam.
Old school turbo engines (and factory engines) have high BP and narrow cams with basically no overlap. Modern, fully built, turbo engines can be built more like a traditional high hp N/A engine. In a successful N/A engine the pulses in the headers can help draw fresh air into the cylinders and they can reach above 100% VE. In those engines high overlap is part of the function.
Now in Patricks engine there's a very, very, aggressive cam with lots of overlap. This means that a lot of backpressure will for a period of time work against the filling of the cylinders. I would be careful of running this cam with the reported back pressure. Much higher risk of detonation.
06-09-2015, 06:05 PM
#43
edit: Oh yes and MILLERS 20/50 Nano oil
I guess I should have said 3.0L and 94 octane.
5 inch exhaust dump to atmosphere, LR level 3 head with HV exhaust valves, Super 75 with T4 and I can't remember if it's a .58 or .68.
extrude hone intake, stage 2 intercooler. Bigger web cam. I've posted the specs on here. It's not as big a cam as yours. Solid lifters. Pauter rods and JE pistons.
Fully lightened crank and the lightest SPEC setup you can do with the metallic clutch disk. Underdrive pulleys and no AC. Evans coolant and the wizard combo rad.
My oil temps were high at the time at about 215 running 3 fans you see on my build post. Intercooler was also a bit heat soaked as we were not running any large fans in the shop.
It was a dyno day with the Porsche club. The numbers for everyone that ran were on the conservative side. Eg one of the Ferrari's there had dyno numbers that were lower than what you see posted on the web for the same Ferrari.
I felt that the numbers were more realistic than what many dyno shops claim. As in Joe Shmo claims they built car XYZ with 100 more horsepower than anyone else can do
I guess I should have said 3.0L and 94 octane.
5 inch exhaust dump to atmosphere, LR level 3 head with HV exhaust valves, Super 75 with T4 and I can't remember if it's a .58 or .68.
extrude hone intake, stage 2 intercooler. Bigger web cam. I've posted the specs on here. It's not as big a cam as yours. Solid lifters. Pauter rods and JE pistons.
Fully lightened crank and the lightest SPEC setup you can do with the metallic clutch disk. Underdrive pulleys and no AC. Evans coolant and the wizard combo rad.
My oil temps were high at the time at about 215 running 3 fans you see on my build post. Intercooler was also a bit heat soaked as we were not running any large fans in the shop.
It was a dyno day with the Porsche club. The numbers for everyone that ran were on the conservative side. Eg one of the Ferrari's there had dyno numbers that were lower than what you see posted on the web for the same Ferrari.
I felt that the numbers were more realistic than what many dyno shops claim. As in Joe Shmo claims they built car XYZ with 100 more horsepower than anyone else can do
06-09-2015, 06:16 PM
#44
I'm glad you posted this. I read the same thing in some of the books I have looked at. Too high Peak cylinder pressure is what breaks rods and destroys rod bearings and blows pistons apart according the authors......hence the cylinder would have to be closed off to the outside and finds the path of least resistance.
Guys, all valves are obviously closed during combustion phase
So back pressure does not work against peak pressure.
The issue is with backpressure and lots of overlap on the intake and exhaust cam.
Old school turbo engines (and factory engines) have high BP and narrow cams with basically no overlap. Modern, fully built, turbo engines can be built more like a traditional high hp N/A engine. In a successful N/A engine the pulses in the headers can help draw fresh air into the cylinders and they can reach above 100% VE. In those engines high overlap is part of the function.
Now in Patricks engine there's a very, very, aggressive cam with lots of overlap. This means that a lot of backpressure will for a period of time work against the filling of the cylinders. I would be careful of running this cam with the reported back pressure. Much higher risk of detonation.
So back pressure does not work against peak pressure.
The issue is with backpressure and lots of overlap on the intake and exhaust cam.
Old school turbo engines (and factory engines) have high BP and narrow cams with basically no overlap. Modern, fully built, turbo engines can be built more like a traditional high hp N/A engine. In a successful N/A engine the pulses in the headers can help draw fresh air into the cylinders and they can reach above 100% VE. In those engines high overlap is part of the function.
Now in Patricks engine there's a very, very, aggressive cam with lots of overlap. This means that a lot of backpressure will for a period of time work against the filling of the cylinders. I would be careful of running this cam with the reported back pressure. Much higher risk of detonation.
06-09-2015, 07:07 PM
#45
Rennlist Member
Hi Guys,
The numbers I choose were for example only, to show you can work the crank numbers back to combustion pressure.
I do not think we are at 2,400 PSI combustion, probably half, or less, of that, even for 450 TQ to the wheels.
Easy enough to develop cyl pressure, as we know how much air is pushed in the cylinder per unit time, and how much fuel is there also (think AFR). So now we look at flame travel to determine where the piston is in the cylinder at peak flame spread, to determine the theoretical CC of the cylinder at that precise moment. Given we are firing at around 25 degrees BTDC. But the cylinder pressure is of most use when the crank is past 0, better at 15+ degrees. But the lever arm at 15 degrees is Sine of 15 degrees or .25 of our stroke say .75".
The cylinder pressure needs more angle to drive TQ. By 30 degrees we have a 1.5" level arm and .866 of cyl PSI working. So peak cyl pressure is of most use around 30 degrees ATDC, and beyond to max at 90 degrees. After 90 degrees our lever starts to shorten, but we do have better rod and piston angle.
On the question of BP, these are fans (turbine) driving fans (Compressor). As I said in the earlier post mass flow is driving the blades (yes volume, but modified by density, hense Mass). So with fuel you have 1.1+ on the exhaust mass flow side. There is no perpetual motion here. The turbine needs work plus losses to provide work to the compressor. Even at 1.1 BP to boost the turbo is approaching 80% efficient. Older designs 60-75% efficient.
The numbers I choose were for example only, to show you can work the crank numbers back to combustion pressure.
I do not think we are at 2,400 PSI combustion, probably half, or less, of that, even for 450 TQ to the wheels.
Easy enough to develop cyl pressure, as we know how much air is pushed in the cylinder per unit time, and how much fuel is there also (think AFR). So now we look at flame travel to determine where the piston is in the cylinder at peak flame spread, to determine the theoretical CC of the cylinder at that precise moment. Given we are firing at around 25 degrees BTDC. But the cylinder pressure is of most use when the crank is past 0, better at 15+ degrees. But the lever arm at 15 degrees is Sine of 15 degrees or .25 of our stroke say .75".
The cylinder pressure needs more angle to drive TQ. By 30 degrees we have a 1.5" level arm and .866 of cyl PSI working. So peak cyl pressure is of most use around 30 degrees ATDC, and beyond to max at 90 degrees. After 90 degrees our lever starts to shorten, but we do have better rod and piston angle.
On the question of BP, these are fans (turbine) driving fans (Compressor). As I said in the earlier post mass flow is driving the blades (yes volume, but modified by density, hense Mass). So with fuel you have 1.1+ on the exhaust mass flow side. There is no perpetual motion here. The turbine needs work plus losses to provide work to the compressor. Even at 1.1 BP to boost the turbo is approaching 80% efficient. Older designs 60-75% efficient.