Cylinder Heads - What Matters Most?
#21
TECH Senior Member
I think that makes total sense, though I am NOT well-versed in this subject matter. It speaks to making the setup WAY more flexible and a better all-around performer.
#22
TECH Fanatic
Join Date: Aug 2009
Location: Coast of San Mateo County Between Pacifica & HMB
Posts: 1,820
Received 220 Likes
on
131 Posts
Knowing (Hopefully LOL) what I know now, I would choose based on the following criteria:
Intake Valve Diameter as a % of Bore Diameter, ideally ~52% (got that from an article by Darrin Morgan, maybe his website? been awhile).
Coefficient of Discharge for the Valve Lifts, RPM, & CFM targets for the Engine Goals - this I got from Tony Mamo.
CSA is heavily influenced if not determined by the first 2 criteria.
Casting - has to be quality, US made and with the material/pads to allow the springs & rockers of choice to be used without modding.
Intake Valve Diameter as a % of Bore Diameter, ideally ~52% (got that from an article by Darrin Morgan, maybe his website? been awhile).
Coefficient of Discharge for the Valve Lifts, RPM, & CFM targets for the Engine Goals - this I got from Tony Mamo.
CSA is heavily influenced if not determined by the first 2 criteria.
Casting - has to be quality, US made and with the material/pads to allow the springs & rockers of choice to be used without modding.
#23
ModSquad
iTrader: (6)
Well I had a freaking page typed and my iPad decided to have a seizure...yay.
For me, it boils down to what am I building. Street or race? Two. Different. Builds.
Lots of parameters make up data that end up telling a head guy what he has. What is a head guy trying to do? Move air the most efficiently through the port in such a way that fuel molecules don’t try to separate themselves from the airmass carrying them when the whole thing gets to the valve head. DI changes this drastically, but that’s another topic. Air and fuel mixed, going through a tunnel...up comes a turn super quickly, and the air has no problem turning, yet the fuel molecules want to stay straight. Simple physics. Not as big a deal with port injection, but carbs are still being utilized, as well as upstream injection, which makes power...and makes that tunnel example reality.
CSA is being brought up on the site a lot lately, by a few people. Most don’t know or even care what this is. Simply put, its a measurement of area at a certain point in the port. Can CSA be measured at different point in the port(s)? Yes. It’s a great data point to know when developing ports. It helps a port designer understand how he/she got a specific CFM. But so does many other factors involved. VCA, or valve curtain area, is a data point used to help porters understand airflow, although this is concentrated in the seat and chamber, it is still important. Port velocity can be determined from a formula, and is important as well to a head porter. Port velocity will aid cylinder fill at low lifts, and makes CFM increase as well. Who cares about this stuff, right? After all, CFM sells heads. Well sure it does, but CFM is the by-product of these mentioned factors, and many more others as well. Another way to look at CFM of a cylinder head is liken to LSA when looking at camshafts. LSA of a camshaft is basically a by-product of multiple valve events. It is what it is. Can you change it? Yes, by moving valve events around. CFM of a given cylinder head is achieved by many other “events” or dimensions of the port. Can you change CFM? Yes, by adjusting dimensions and angles, as well as textures and many other things in the port, and below the port (valve seat, chamber, etc). Head porters chase numbers, and the number that sells heads is CFM. More is better, because air makes power. Less air, less power. More air, more power.
I look at port velocity for a street build, and I look at overall port size for a race build. A port with a higher velocity will move more air at lower lifts, which makes a street engine build more drivable. More enjoyable to drive. Less crap you have to “settle” with just to have power on the street. A bigger port will move more air, but it’s lazier at low lifts which makes it a better choice for a race build, because a race build will spend its life at 6500 rpm and north. Who cares how it drives?
For both builds the valvejob is UBER important. I want my valves to sit ON the seat, not IN the seat. Makes a yuge difference across the board. I want my valve seat perfectly blended into the bowl. A lot of heads come with this all jacked up. There’s a lot of power to be had in this area, street build or race build.
I also pay attention to what particular intake manifold is available for a particular cylinder head I’m wanting to run, but that seems to changing a lot lately with the aftermarket finally paying attention to what engine builders are looking for in an intake manifold, such as plenum volume, runner lengths, etc.
I like this discussion, and it’s refreshing to see actual tech getting discussed here on Tech.
For me, it boils down to what am I building. Street or race? Two. Different. Builds.
Lots of parameters make up data that end up telling a head guy what he has. What is a head guy trying to do? Move air the most efficiently through the port in such a way that fuel molecules don’t try to separate themselves from the airmass carrying them when the whole thing gets to the valve head. DI changes this drastically, but that’s another topic. Air and fuel mixed, going through a tunnel...up comes a turn super quickly, and the air has no problem turning, yet the fuel molecules want to stay straight. Simple physics. Not as big a deal with port injection, but carbs are still being utilized, as well as upstream injection, which makes power...and makes that tunnel example reality.
CSA is being brought up on the site a lot lately, by a few people. Most don’t know or even care what this is. Simply put, its a measurement of area at a certain point in the port. Can CSA be measured at different point in the port(s)? Yes. It’s a great data point to know when developing ports. It helps a port designer understand how he/she got a specific CFM. But so does many other factors involved. VCA, or valve curtain area, is a data point used to help porters understand airflow, although this is concentrated in the seat and chamber, it is still important. Port velocity can be determined from a formula, and is important as well to a head porter. Port velocity will aid cylinder fill at low lifts, and makes CFM increase as well. Who cares about this stuff, right? After all, CFM sells heads. Well sure it does, but CFM is the by-product of these mentioned factors, and many more others as well. Another way to look at CFM of a cylinder head is liken to LSA when looking at camshafts. LSA of a camshaft is basically a by-product of multiple valve events. It is what it is. Can you change it? Yes, by moving valve events around. CFM of a given cylinder head is achieved by many other “events” or dimensions of the port. Can you change CFM? Yes, by adjusting dimensions and angles, as well as textures and many other things in the port, and below the port (valve seat, chamber, etc). Head porters chase numbers, and the number that sells heads is CFM. More is better, because air makes power. Less air, less power. More air, more power.
I look at port velocity for a street build, and I look at overall port size for a race build. A port with a higher velocity will move more air at lower lifts, which makes a street engine build more drivable. More enjoyable to drive. Less crap you have to “settle” with just to have power on the street. A bigger port will move more air, but it’s lazier at low lifts which makes it a better choice for a race build, because a race build will spend its life at 6500 rpm and north. Who cares how it drives?
For both builds the valvejob is UBER important. I want my valves to sit ON the seat, not IN the seat. Makes a yuge difference across the board. I want my valve seat perfectly blended into the bowl. A lot of heads come with this all jacked up. There’s a lot of power to be had in this area, street build or race build.
I also pay attention to what particular intake manifold is available for a particular cylinder head I’m wanting to run, but that seems to changing a lot lately with the aftermarket finally paying attention to what engine builders are looking for in an intake manifold, such as plenum volume, runner lengths, etc.
I like this discussion, and it’s refreshing to see actual tech getting discussed here on Tech.
#24
TECH Senior Member
VERY well put, Che70velle!
I am learning a BUNCH about head design/tech from you, Darth, Bortous, and whoever else can contribute to this.
Keep it up guys! The learning process continues....
I am learning a BUNCH about head design/tech from you, Darth, Bortous, and whoever else can contribute to this.
Keep it up guys! The learning process continues....
#26
TECH Veteran
Darth, I know you had great success out your LLSR LS1 Mamo build with those heads.
Just wondering have you read an old article from years back in GMHTP where they put small bore Mast LS3 heads on a LS1 with a Mast Motorsports 226/240 cam.... it didnt do all that well. Wonder "what happened"?
Just wondering have you read an old article from years back in GMHTP where they put small bore Mast LS3 heads on a LS1 with a Mast Motorsports 226/240 cam.... it didnt do all that well. Wonder "what happened"?
#27
TECH Fanatic
Join Date: Aug 2009
Location: Coast of San Mateo County Between Pacifica & HMB
Posts: 1,820
Received 220 Likes
on
131 Posts
Darth, I know you had great success out your LLSR LS1 Mamo build with those heads.
Just wondering have you read an old article from years back in GMHTP where they put small bore Mast LS3 heads on a LS1 with a Mast Motorsports 226/240 cam.... it didnt do all that well.
Wonder "what happened"?
Just wondering have you read an old article from years back in GMHTP where they put small bore Mast LS3 heads on a LS1 with a Mast Motorsports 226/240 cam.... it didnt do all that well.
Wonder "what happened"?
#28
Moderator
Thread Starter
iTrader: (4)
Join Date: Jul 2014
Location: My own internal universe
Posts: 10,446
Received 1,838 Likes
on
1,146 Posts
Chevelle- I had never thought of it that way. Mind blown. Look at cfm on heads like LSA on a cam. Thank you for that.
#29
Moderator
Thread Starter
iTrader: (4)
Join Date: Jul 2014
Location: My own internal universe
Posts: 10,446
Received 1,838 Likes
on
1,146 Posts
Tusky, I think a big difference there is cam duration. Mine was 237/245. 653 lift. Second idk what intake. Or the rest of the details of the build. 226 seems small if you are aiming for a record.
That type of question is why I started the thread. I would think they would perform just as well
That type of question is why I started the thread. I would think they would perform just as well
#30
TECH Veteran
#31
TECH Veteran
Tusky, I think a big difference there is cam duration. Mine was 237/245. 653 lift. Second idk what intake. Or the rest of the details of the build. 226 seems small if you are aiming for a record.
That type of question is why I started the thread. I would think they would perform just as well
That type of question is why I started the thread. I would think they would perform just as well
The test car in the article is a 2001 z28 a4
mods:
l76 intake
fast 90mm tb
stock ls1 block
226/240 588/597 114 cam
mast motorsport 61cc ls3 small bore 12 degree heads
rpm built 4l60
vig3600
9" rear
power 423 whp / 405wtq on a dyno jet. 325/326 before heads/cam
Head flow numbers:
lift Intake Exhaust
.300 208 105
.400 266 217
.500 306 222
.600 335 226
.700 351 234
.750 335 238
Like i said its a old article, 2010 to be exact. Dan, i still dont no real MAGIC even if the car was equipped with a ported fast 102mm etc.
#32
TECH Apprentice
iTrader: (2)
Don't mean to Hi-jack the thread and you all have made me feel better about my choice of TF's 220 fast as cast . I chose them because the flow at .500 lift is incredible. compared to 243's, BUT WHAT I NEVER SEE? Is how much AIR/FLOW does say a 4.00 bore and pic a stroke? HOW MUCH AIR FLOW DOES THAT PISTON/STROKE COMBO REALLY NEED? how much can it really suck in? I've never seen a flow chart like that?
#33
TECH Apprentice
iTrader: (2)
Darth, I know you had great success out your LLSR LS1 Mamo build with those heads.
Just wondering have you read an old article from years back in GMHTP where they put small bore Mast LS3 heads on a LS1 with a Mast Motorsports 226/240 cam.... it didnt do all that well. Wonder "what happened"?
Just wondering have you read an old article from years back in GMHTP where they put small bore Mast LS3 heads on a LS1 with a Mast Motorsports 226/240 cam.... it didnt do all that well. Wonder "what happened"?
#34
TECH Veteran
Hard to answer that cause LS3 heads tend to respond to a wide split such as the cam they used in the article. By the head being a aftermarket head vs a factory LS3 head.... the motor might would do better with a 226/236 for example cause the heads SHOULD have a BETTER intake to exhaust relationship over the GM factory LS3 head. Notice some aftermarket heads can make great power with a single pattern cam, reason why is because the intake to exhaust relationship is so damn good that the heads dont need a wide split.
In Tusky opinion i still think you can run "Too much head" for a given cubic inch motor. I may be wrong or get bashed here for that statement but i call it how i see it.
In Tusky opinion i still think you can run "Too much head" for a given cubic inch motor. I may be wrong or get bashed here for that statement but i call it how i see it.
#35
hard to answer that cause ls3 heads tend to respond to a wide split such as the cam they used in the article. By the head being a aftermarket head vs a factory ls3 head.... The motor might would do better with a 226/236 for example cause the heads should have a better intake to exhaust relationship over the gm factory ls3 head. Notice some aftermarket heads can make great power with a single pattern cam, reason why is because the intake to exhaust relationship is so damn good that the heads dont need a wide split.
In tusky opinion i still think you can run "too much head" for a given cubic inch motor. I may be wrong or get bashed here for that statement but i call it how i see it.
In tusky opinion i still think you can run "too much head" for a given cubic inch motor. I may be wrong or get bashed here for that statement but i call it how i see it.
#36
Moderator
Thread Starter
iTrader: (4)
Join Date: Jul 2014
Location: My own internal universe
Posts: 10,446
Received 1,838 Likes
on
1,146 Posts
Tusky I made 470/400 with my heads on my 346 with a 227/235 cam. BUT, key differences:
Fast102
T56
Factory rear (LLSR broke the rear so then went 9 inch)
Ewp
Tested open headers with 18 inch extensions.
That likely makes up the difference.
Fast102
T56
Factory rear (LLSR broke the rear so then went 9 inch)
Ewp
Tested open headers with 18 inch extensions.
That likely makes up the difference.
#37
Moderator
Thread Starter
iTrader: (4)
Join Date: Jul 2014
Location: My own internal universe
Posts: 10,446
Received 1,838 Likes
on
1,146 Posts
Originally Posted by Tuskyz28
Hard to answer that cause LS3 heads tend to respond to a wide split such as the cam they used in the article. By the head being a aftermarket head vs a factory LS3 head.... the motor might would do better with a 226/236 for example cause the heads SHOULD have a BETTER intake to exhaust relationship over the GM factory LS3 head. Notice some aftermarket heads can make great power with a single pattern cam, reason why is because the intake to exhaust relationship is so damn good that the heads dont need a wide split.
In Tusky opinion i still think you can run "Too much head" for a given cubic inch motor. I may be wrong or get bashed here for that statement but i call it how i see it.
In Tusky opinion i still think you can run "Too much head" for a given cubic inch motor. I may be wrong or get bashed here for that statement but i call it how i see it.
Trucks run low compression due to towing and high load so ex needs to open later. Intake closes sooner to favor lower rpm torque production. Also causes peak power early. Reduced overlap to also favor lower rpm VE. Voila! Short duration, low or zero split, tight LSA truck cam.
Pick your EVO based on compression not head style or port shape. The pissed-on will push the air out.
#38
Moderator
iTrader: (20)
Intended Usage
The first thing in selecting cylinder heads is knowing the intended usage. How many cubic inches are you trying to feed and what RPM range will you want to make power? What you're really trying to do, whether you realize it or not, is to manipulate the torque curve for the application. You want to put the torque curve in the rpm range you need it, and then maximize the torque in that area.
Port Size
The port size, or more specifically the minimum cross sectional area (MCSA), is how you place the torque curve in the RPM range you want. Peak torque generally occurs around 240ft/sec so the MCSA can be sized accordingly using very simple math. Since horsepower is calculated based on torque and RPM, the higher up in the RPM range you make torque, the more horsepower you will make. It's common for people buy heads that will put the torque curve higher than what their valvetrain budget can afford.
Port Shape
When the ports are sized correctly, the shape of the port has a tremendous influence on the area under the curve. The better the shape, the higher airspeeds the port can maintain and the longer spread between peak torque and peak horsepower. The short side radius, which is basically the floor of the port as it turns to the valve job, is normally the hot spot. Too abrupt, and the air can separate off the turn and cause some bad turbulence. Tall ports with more gradual short turns tend to be better for this, which is why raised runner heads are so popular in racing. For lower RPM and lower airspeeds, sometimes a more abrupt turn is better to promote better emissions and idle characteristics.
Valve Size
I like to use as big of a valve as possible. The bigger valve achieves more curtain area at every lift point, similar to using a camshaft with more aggressive lobes that open and close faster. Unfortunately, most cylinder head porters will size the MCSA as a percentage of the valve, so a large valve often means a large MCSA.
There really isn't such a thing as too much airflow unless it is because the port is too big. At 28" H2O, which is the depression that most heads are tested on the flowbench, you won't flow more than about 142CFM/sq in of area. Flowbench numbers are almost completely useless IMO so I wouldn't worry too much about it
The first thing in selecting cylinder heads is knowing the intended usage. How many cubic inches are you trying to feed and what RPM range will you want to make power? What you're really trying to do, whether you realize it or not, is to manipulate the torque curve for the application. You want to put the torque curve in the rpm range you need it, and then maximize the torque in that area.
Port Size
The port size, or more specifically the minimum cross sectional area (MCSA), is how you place the torque curve in the RPM range you want. Peak torque generally occurs around 240ft/sec so the MCSA can be sized accordingly using very simple math. Since horsepower is calculated based on torque and RPM, the higher up in the RPM range you make torque, the more horsepower you will make. It's common for people buy heads that will put the torque curve higher than what their valvetrain budget can afford.
Port Shape
When the ports are sized correctly, the shape of the port has a tremendous influence on the area under the curve. The better the shape, the higher airspeeds the port can maintain and the longer spread between peak torque and peak horsepower. The short side radius, which is basically the floor of the port as it turns to the valve job, is normally the hot spot. Too abrupt, and the air can separate off the turn and cause some bad turbulence. Tall ports with more gradual short turns tend to be better for this, which is why raised runner heads are so popular in racing. For lower RPM and lower airspeeds, sometimes a more abrupt turn is better to promote better emissions and idle characteristics.
Valve Size
I like to use as big of a valve as possible. The bigger valve achieves more curtain area at every lift point, similar to using a camshaft with more aggressive lobes that open and close faster. Unfortunately, most cylinder head porters will size the MCSA as a percentage of the valve, so a large valve often means a large MCSA.
Don't mean to Hi-jack the thread and you all have made me feel better about my choice of TF's 220 fast as cast . I chose them because the flow at .500 lift is incredible. compared to 243's, BUT WHAT I NEVER SEE? Is how much AIR/FLOW does say a 4.00 bore and pic a stroke? HOW MUCH AIR FLOW DOES THAT PISTON/STROKE COMBO REALLY NEED? how much can it really suck in? I've never seen a flow chart like that?
#39
TECH Veteran
I do not thing the I/E biases and ratios matter as much as ppl think. Higher compression, work is extracted faster so ok to open valve earlier. Lower compression need to hold the valve closed longer to extract work due to slower burn.
Trucks run low compression due to towing and high load so ex needs to open later. Intake closes sooner to favor lower rpm torque production. Also causes peak power early. Reduced overlap to also favor lower rpm VE. Voila! Short duration, low or zero split, tight LSA truck cam.
Pick your EVO based on compression not head style or port shape. The pissed-on will push the air out.
Trucks run low compression due to towing and high load so ex needs to open later. Intake closes sooner to favor lower rpm torque production. Also causes peak power early. Reduced overlap to also favor lower rpm VE. Voila! Short duration, low or zero split, tight LSA truck cam.
Pick your EVO based on compression not head style or port shape. The pissed-on will push the air out.
Food for thought# Its a sbe ls1 car on board here running a 228R single pattern cam with some TFS 220s as cast out box. The car is makes
445 RWHP and 430 RWTQ on a mustang dyno with a out of date fast 90mm thats not ported, a SLP Y- pipe exhaust, SLP 1 3/4 headers. Aww yeah, the car is a 4L60 car with 3400 stall. Rearend is a Strange 12 bolt.
Not sure what the combo would make with a T56 but im sure its more not to mention ita room for improvement on intake,headers, and exhaust.
#40
Moderator
Thread Starter
iTrader: (4)
Join Date: Jul 2014
Location: My own internal universe
Posts: 10,446
Received 1,838 Likes
on
1,146 Posts
On the subject of *too much head* I have a theory, but nothing to back it up. Just thoughts.
Force = pressure x area. Pressure differential caused by absolute intake pressure and depression created by piston moving down and area being csa.
Bigger area means more force pushing air in
Bigger area means it also will flow more air with less velocity, so you lose momentum of the air column. In particular this will hurt filling the cylinder at lower rpm with a later IVC so there is more time for air to slow down and even go the wrong way as cylinder pressure exceeds manifold pressure.
This is where I think Smokey has a point. I think if you run less cam with a larger head, I think you can make it do quite nice and make a great torque curve. Possibly even broader.
Is there a such thing as too large a head? Yes, but I do not think within the confines of fitting two valves into the cylinder and having to make room for pushrods and coolant passages, and having to fit the whole thing into a car.
Here is an example of my thinking. Take a typical 400 cfm head on a ls7. Typical performance cam is going to be in the lines of 248/260.
Now put a 500 cfm head on there. You end up shrinking the ex valve further, etc, but theoretically we make it fit. It will revert worse so we need less overlap. Cannot start with air moving the wrong direction. So instead of a -16 IVO, start with a -4 IVC. By this point, piston is all the way up. And exhaust vacuum wave is good and strong. Now we are at 236/260.
My bet is the 500 cfm with that cam will make equal torque from 2000 rpm to peak torque and then carry further past peak.
Complete theory. Feel free to flame. Trying to learn.
Force = pressure x area. Pressure differential caused by absolute intake pressure and depression created by piston moving down and area being csa.
Bigger area means more force pushing air in
Bigger area means it also will flow more air with less velocity, so you lose momentum of the air column. In particular this will hurt filling the cylinder at lower rpm with a later IVC so there is more time for air to slow down and even go the wrong way as cylinder pressure exceeds manifold pressure.
This is where I think Smokey has a point. I think if you run less cam with a larger head, I think you can make it do quite nice and make a great torque curve. Possibly even broader.
Is there a such thing as too large a head? Yes, but I do not think within the confines of fitting two valves into the cylinder and having to make room for pushrods and coolant passages, and having to fit the whole thing into a car.
Here is an example of my thinking. Take a typical 400 cfm head on a ls7. Typical performance cam is going to be in the lines of 248/260.
Now put a 500 cfm head on there. You end up shrinking the ex valve further, etc, but theoretically we make it fit. It will revert worse so we need less overlap. Cannot start with air moving the wrong direction. So instead of a -16 IVO, start with a -4 IVC. By this point, piston is all the way up. And exhaust vacuum wave is good and strong. Now we are at 236/260.
My bet is the 500 cfm with that cam will make equal torque from 2000 rpm to peak torque and then carry further past peak.
Complete theory. Feel free to flame. Trying to learn.