Old but interesting article on sqaure vs cathedral heads

Old but interesting article on sqaure vs cathedral heads

Author
Discussion

ringram

Original Poster:

14,700 posts

248 months

Wednesday 23rd April 2014
quotequote all
The LS Evolution
Ingenuity or Idiocracy?

I wasn't put on this planet to be the ¡§Glenn Beck¡¨ of cylinder heads but at times I am told my thoughts are as controversial if not more then Beck's. The LS platform itself is controversial in nature. It has upset many men that have tried to improve upon its design, a humbling creature for sure. I find myself daily dealing with the same phone calls...over and over again. After a few years of answering the same questions ad nauseum I figured it was time to put the pen to the paper (or the keyboard to the screen).


A little about me:
.
I run a small white label induction shop. Most of my time is spent doing port mastering for engine builders and cylinder head shops. The rest of it is spent dealing with other induction issues. I am fairly proficient in all modern domestic and import engines. Never heard of me? Well...then I am doing my job correctly! 95% of my time is spent supplying product and design information to other cylinder head shops, engine builders and racers. I generally don't deal directly with the public. Why would I take the time to write this? It certainly isn't because I have a bunch of free time! I feel that educating the end user only benefits everyone, which makes this industry stronger as people make better decisions on how they spend their hard earned dollars.

The questions myths/misconceptions I deal with.

1. Square port LS heads are slow and sluggish.
2. Big velocity / Low lift flow builds big TQ. (Cathedral ports are for velocity)
3. Factory LS heads are too thin for boost
4. The lower the valve angle, the more power
5. Re-engineering the General


1. Square Port LS heads are slow and sluggish.

Most people hate change. When something changes in their life most people react by identifying what has changed. When the GEN 3 engine was released everyone looked at the intake port and commented how SMALL it was. At that time most people instead of looking how tall the port was immediately commented that the small port would never run and the LS engine was a bad design.

If we took the time to measure the port we would find that the average width is 1¡¨. That is skinny compared to its SBC predecessor, but what it lacks in width it more than makes up in height. An average cathedral port head measures 2.750¡¨ tall dwarfing even its big uncle (Mr. BBC) in height and giving as an average cross sectional area at the intake flange of 2.625¡¨sq. In comparison your average out of box aftermarket 200 cc SBC head will average 2.3¡¨sq. at the flange with the OEM port being even smaller. Simple math tells us that when used on a 346 cubic inch engine the General learned that big ports are a good thing a long time ago. From appearance though the cathedral port head still ¡§looks¡¨ small so many never think to measure it and call the mental picture of small good enough.


When you look at an LS-3 or LS-7 head it doesn't take an optical comparator to tell it is one big fat port! To the untrained eye many comment that the head is TWICE as big as its cathedral port parent. Before we judge another book by its cover maybe we better measure it before we open our mouth. Average port opening on a LS3 (L92) head is 1.250 ¡§ wide X 2.475¡¨ high. It also carries a fatter corner radius (about .150¡¨sq). That gives us 2.944¡¨sq. at the intake flange. What looks like twice as big turns out to only be about 12% larger? That is one fact we can't argue: the square port heads are only marginally bigger at the intake flange. What about that BIG HONKIN¡¦ INTAKE VALVE! 2.165¡¨ in Diameter. The discharge coefficient and velocity through the port must be so slow raw gas drips out the exhaust pipe if you let this monster idle!

Au contraire mon frere

Let¡¦s come back to our friend MATH again. We will be working with him through the rest of the document so be nice to him and he will be nice to you. I promise we will keep it at a grade school level though. Before we get there we need to agree on a few things.

Everyone's flow bench is different (or so the aftermarket wants you to believe). So we need to agree on some flow numbers for this experiment.

1. Average ported LS6 heads (or copies) w/ a 2.055 valve (Most I see carry a 91% throat and flow around 320 cfm peak).
2. Average ported LS3 heads (or copies) w/ a 2.165 valve (Most carry the same 91% throat and flow around 365 cfm peak).

It stands to reason if we can figure out the area of the throat and compare it to the amount of airflow we can get a quick down and dirty idea of which cylinder head is ¡§faster¡¨ in airspeed.

LS6 head to start.
2.055 * .91 = 1.870 throat diameter

A simple formula for finding the area of a circle is (diameter x diameter x .7854)
So, 1.870 x 1.870 x .7854 = 2.746 ¡§sq. of throat area.

Then 320 cfm/2.746 ¡§sq. = 116.53 cfm/sq¡¨ of throat area.

Now I am not going to bore you with the math the whole way through this, so get your calculator and see if you find the same results as I do!

LS3 (big fat and slow...right?)

365 cfm for every 3.046¡¨sq of area, or more simply, there is 119.83 cfm/sq¡¨ of throat area. Much slower....WAIT! HOLD ON. More cfm / sq¡¨ = higher velocity. If we move more air through the same size hole the only way we can do it is by moving it FASTER!

So much for that big ugly port being slow. Now what does all this mean? And why are people lying to me? Read on:





2. Big velocity / Low lift flow builds big TQ. (Cathedral ports are for velocity)

(or...Brawndo has what plants crave...electrolytes)

Mass marketing's purpose is simple: Scare the consumer into spending money. Turn on your TV and think about what someone is trying to sell you. ¡§You need to get rid of your earth killing incandescent light bulbs and purchase compact fluorescent bulbs (that contain mercury) as they are safer and conserve electricity¡¨. Think about that for a moment.

This is where you are going to think that I am going to start sounding like Mr. Beck. Honestly... I don't howl at the moon... I don't attend tea parties nor believe in conspiracy theories. But facts are facts. Marketing has won over truth for many years. Who invented the radio? Marconi? Think again! Mass marketing even made its way into your history book in school! I don't honestly believe all marketing is bad. But let¡¦s remember what the purpose of marketing is. The people doing the marketing want to get the pictures of the dead Presidents out of your wallet. A white lie here or there won't hurt anyone right ?

At the current point in time most of the aftermarket cylinder head companies have large investments in the castings that are on there shelves. Core box cost, foundry time, etc. etc. Big bad GM comes by and throws a bucket of water on the fire with the square port head. What to do? Remember... a little white lie can't hurt anyone right?

It looks BIG.. Let¡¦s tell them it is TOO BIG to make TQ. It will be some big fat lethargic heap. The real small cathedral port head will make BIG TQ. Small ports and smaller headers MAKE BIG TQ! Yeah, RIGHT! If we keep this up Bernoulli is going to climb out of his grave.

Let¡¦s take a look back and try to think like a GM engineer.. The cathedral port head was designed with many goals in mind. Again I am going to keep this stupid simple (it is a bit more complex in real life...but not much more). Let¡¦s look at only 3 of the goals. 1. It needs to go like Hell. 2. It needs to be in a nice small package. 3. The valve train must be stable.

Again...the simple approach.

ƒ{ It must go like hell (Read... big ports almost 2.7¡¨sq to be exact!)
ƒ{ It needs to be a nice small packages (Read... The motor must be tinny!)
ƒ{ Stable valve train (Read.... everything must be as tight as possible)

Well I don't know about you, but if I needed to build a small motor but I needed big ports I would try to use every square inch of room available the best I could. But we still need a big port and we don't want push rods and rockers all over the place. So we need to stand the rockers up and keep the manifold as tight to the center-line of the engine as possible. The best solution is to go tall! Nice tight narrow intake ports allow us to keep the push rods perpendicular to the camshaft but we get to keep our nice big port!

But what happens when the port isn't big enough? Why wouldn't it be big enough? 3 words... Corvettes and Pickup Trucks. GM's flagship car the Corvette must run really well. GM's number one seller, the trucks must also run well. New casting technologies allowed GM to produce a new style of cylinder head at a price point like no other. Less than $150 cost per head! This allowed more cost to be applied in other places (read offset rockers and additional vehicle engineering) The L92/L76/LS3 cylinder head was born.

If we compare the original LQ series 6.0l truck to the new L92 / L94 (square port) engine we see TQ output is up roughly 12% yet the new engine is only 3% larger? Obviously the square port head does a pretty good job of making power in short order! Remember our friend MATH?


You will also get sold on ¡§hour glass¡¨ ports. These are basically cylinder heads with a point of restriction somewhere in the middle of the port to increase velocity. Many want you to believe that high port velocity is where torque production comes from. Instead of using math and making this one complicated... lets use some stupid simple theory.

If we want to make more torque we need to find a way to apply more pressure to the piston so it can apply more pressure to the connecting rod and finally the crankshaft. What applies the pressure to the piston anyway? The answer is simple...HEAT! Let¡¦s turn our air/fuel mixture into something we can see a little bit better... say... Sand.

The more sand you put into a jug... the faster.... the more heat you will make.

Would you rather use a funnel or an hour glass?

3. Factory LS heads are too thin for boost.

I think I have heard it all now. Next they will be selling fire insurance for underwear.
The picture above is of a very early L92 prototype. It is one of the thinnest of the L92 castings I have measured at .470¡¨ thick. Is .470¡¨ thick enough? Well in short, Yes! Assuming you aren't going outlaw 10.5¡¨ racing the .470¡¨ deck is more than adequate. A good way to think about it is comparing the LS engine to a small block ford (A small block Ford??? He is crazy!). If you have yet to figure out the LSx engine and SBF have the same bolt pattern I am here to confirm that for you. The point of this experiment is to stop the BS of the L92 heads having a deck that is less than .300¡¨ thick. The aftermarket guys must use a different measuring device for sure to come up with that kind of gibberish!

So what is the secret to getting a 4 bolt cylinder head to seal correctly? For the average boosted engine the solution is simple. But it is a secret so I guess we keep it that way.

What fun would it be keeping it a secret? You really are dying to know anyway. The secret is stupid simple. Surface finish. correct gaskets and a good tune! THAT¡¦S ALL THERE IS TO IT!

A 20 finish on the block and cylinder head, a good set of MLS gaskets, a good tune and you are set! I generally don't recommend tuners by name. One thing I will say is that if the tuner will fly to your house and tune your car for $50.00 he is not your guy! A bunch of great tuners exist and you will more than likely have to wait in line and pay for their service. At the end of the day though they are worth their weight in gold!


4. The lower the valve angle...the MORE power (really..it should read the MORE problems)

It wasn't long after people started modifying cylinder heads that they started moving valves around. Luckily today most people understand why and when you want to alter valve placement or angle. I guess a few exist that still don't get it though. We need to apply a little theory and math together to get a good idea of why we may or may not want to adjust the valve angle. A few thoughts first:

ƒ{ Some believe magic happens when you reach a certain valve angle (normally 18 degrees)
ƒ{ In reality the engine does not know what angle the valve is on
ƒ{ the relationship of the intake port entrance to the valve angle is what produces more power (What we refer to as approach angle)
ƒ{ Engines with valve angles steeper than 20 degrees gain bore clearance by standing the valve up. Obviously this does not affect the LSx engine with its factory 15 degree valve angle.

If we want to move air effectively into an engine common sense would tell us that wewant to make the path as straight as possible. Due to valve train and engine packaging we only can move so much in either direction. Since the LS head carries a shallow 15 degree valve angle the intake and exhaust valves stay far away from the engine bore when opening even at valve lifts over 1¡¨. The only benefit of moving the valve angle any shallower would be to make the combustion chamber smaller. In reality the 15 degree chamber can support volumes as small as 45 cc's. So what other benefit do we gain by making the valve angle shallower? The answer may be shocking to some. From here on out, unless we can raise the intake port great amounts to offset moving the valve angle, we only make things worse with shallower angles. As we stand the valve up without raising the intake port entrance we are making the turn over the short side steeper and steeper increasing the energy it takes to turn the charge into the combustion chamber.

Now the bad part. We have moved our intake valve to something less than 15 degrees without raising our intake port entrance. It has given us no benefit as the intake valve is already far away from the bore wall when opening and we don't really have a use for a combustion chamber smaller then 45cc's. But by moving that valve we have blessed ourselves with a slew of other problems. What problems you ask?

ƒ{ None of the current altered valve location castings alter the width or height of the casting. This is understandable as the casting needs to be a direct bolt on. Since the valve package has been relocated the compound angle of the rocker packages has been severely compromised. No chance of running a larger push rod unless you want to grind into the valve cover rail or intake port. More lateral scrub as we can't move the rocker high enough...etcc.
ƒ{ The additional lateral load on the valve directly affects valve guide life
ƒ{ The solution for some is to run a stud mounted rocker system to help the guide life problem. This does help the guide life a small percentage but at a huge cost to valve train stability.
ƒ{ In a race situation a custom piston will be required.

OK... enough with the negatives. All this for nothing! We gain nothing with the altered valve location! Let¡¦s keep our rigid factory rocker and valve location!

5. Re-engineering the General

Obviously anything good can be made better.... Right? Well some obstacles certainly exist with the factory LS square ports (and are common among all factory LS heads). Most can be solved cheaply enough that it is not an issue. The one big issue that only affects the hardest core racers is the lack of valve spring pocket diameter. The factory configuration does not allow a spring diameter greater than 1.400¡¨ without major modification. How in the world could GM go through all of this engineering and miss something so simple?
The answer is simple....THEY DIDN'T! Plans have been in motion for many years on a performance version of the factory square port heads. Finally today demand is high enough for GM to manufacture its ¡§LSX¡¨ line of street heads with more material in all the right places plus support for valve spring diameters up to 1.650¡¨! All this with a cost that rivals most of the aftermarket companies with no worn valve guides or lost promises to boot!

At the end of the day it is obvious that the good people of GM have done their homework!


Comments or Thoughts? Shoot me an email!

Dennis Wheet Jr.
AFD racing
dennis@afdracing.com





S800VXR

5,876 posts

200 months

Wednesday 23rd April 2014
quotequote all
square or sqaure ....? Typical NA type mistake. wink

KMud

2,924 posts

156 months

Wednesday 23rd April 2014
quotequote all
yes Trust the OE bits, be suspicious of those who think they can do better without telling you what has been compromised to 'better' the original bits.

ArnieVXR

2,449 posts

183 months

Wednesday 23rd April 2014
quotequote all
Bet that felt good to get that off your chest. Luckily, when doing my own research I always buy cheap stuff that looks cool. Thousands of FI horses can't be wrong wink

stevieturbo

17,260 posts

247 months

Wednesday 23rd April 2014
quotequote all
Wayyy too much to read.

Bottom line. They both work great, and work even better with boost wink

stigmundfreud

22,454 posts

210 months

Wednesday 23rd April 2014
quotequote all
Dont start him off!

Also interesting to note that the 2.5in and stock paper filter has more than enough cfm unless you start to turn to boost

ringram

Original Poster:

14,700 posts

248 months

Thursday 24th April 2014
quotequote all
Its a head comparo gents.

Also airflow NA vs Airflow FI is essentially the same, the pressure delta is larger that's all..

Same principles apply.

stigmundfreud

22,454 posts

210 months

Thursday 24th April 2014
quotequote all
its down to the perssure? Does that mean more volume of air per CC multiplied by the bar/psi figure?

MyM8V8

9,457 posts

195 months

Thursday 24th April 2014
quotequote all
stigmundfreud said:
its down to the perssure? Does that mean more volume of air per CC multiplied by the bar/psi figure?
Through the same ports (pipework) air will flow at a higher velocity at higher pressure. That means that more air mass should be delivered to the cylinder. Problem being that when you apply pressure the temperature increases.

(Volume is not as important as the mass of fuel mix and there is a difference.).

Not trying to force Granny to suck eggs BTW.

Its also important to note that small restrictions in air pipework can be overcome simply by compressing the air. Therefore smaller heads can flow more air when induction is forced.


Edited by MyM8V8 on Thursday 24th April 13:17

ringram

Original Poster:

14,700 posts

248 months

Thursday 24th April 2014
quotequote all
Yep heads are flowed at a certain depression, aka pressure delta.
FI just means the delta is larger so flow is likely higher..
Guess where the extra power comes from... more mass.

cnc heads give you more mass at any set delta whether atmospheric or forced.

MyM8V8

9,457 posts

195 months

Saturday 26th April 2014
quotequote all
For a technical paper, his analogy about sand or air flowing through an hour glass is not really on. I could suggest he uses a snowstorm device for the sand, to get more mass through, but its as pointless as the comparison he makes. Air is a compressible fluid, sand is not.