Discussion: "Porting by Numbers" Darin Morgan

[Procision-Auto]

One word to discribe the contents of this audio CD: Humbling

If you've learned about cylinder heads from magazine lingo, and Internet
buzz words (like myself) for your entire life, you will probably be in shock
after listening to this summary of cylinder head basics.

You can pretty much toss out a majority of the information you picked up
from internet discussion boards, or at least re-think how you interpret the
data that everyone is throwing around.

I knew head flow wasn't the bottom line when picking a cylider head, but
according to Darin, flow numbers are likely the least significant information
that determine performance.

It makes sense that a flow bench cannot recreate dynamic pressure changes
throughout the engine cycle, or compensate for the lack of sonic and thermal
energy which is not present on the flow bench.

This was always obvious to me, and now moreso than ever, I wonder
how much weight a flow bench really has when making the perfect port.

Darin has me approaching cylinder head selection in a totally different manner
after last night: Cross sectional area, runner volume; valve sizes and
port shape.

Some questions for head porters and cylinder head techs:

1. Do you consider port shape for sonic energy, or reflected pulses when
designing the port? In other words, are there any tests to measure the
reversal of air flow, or how a tuned pulse might react when it reflects?

2. How many different test depressions are used to flow the port?
Do you also flow the head at each valve lift increment at each depression?
Surely, you all do. Which depression is most significant, and which is typically
used to shape the port; 28 in./H20?

3. Is there a major issue with flow/turbulence at the junction of the intake
port and head port if the surfaces do not make a smooth transition?
If not, what is a typical variance, or allowable step?

I have a slight lip going into the head port (approximately 1/16th of an
inch) as the intake runner is slightly ported. This was purchased used,
and I was basically too lazy to match the mating port.

That's all for now, but I'm sure the questions will keep coming.

[MadBill]

In reference to question 2., seems like anyone who uses engine pressure analysis equipment, e.g.: http://www.tfxengine.com/ , should have exactly the data needed to answer this. By charting the pressure drop across the valve at say 10-20° angular increments vs. valve opening/closing at each angle, one would know exactly what flow bench pressure delta to use at each lift point. (and could prioritize efforts on the most critical)
As intake delta is reputed to max out at ~140-160" H2O, this should be almost within reach of a good bench hot-rodded with an armful of extra motors, since the maximum will likely occur at ~ 75° -80° ATDC, well below peak lift. For the exhaust, not so simple! Since even a N/A engine can have 100 psi or more cylinder pressure at EVO, it would take a decent supersonic wind tunnel to accurately reproduce the pressure conditions, and that would still ignore gas composition and temperatures...

[Darin Morgan]

Although Bills responce is applicable and spot on, I find it complicated and hard to elaborate on. If one wishes to invest in that course of action, I say go for it.



(1) Do you consider port shape for sonic energy, or reflected pulses when
designing the port? In other words, are there any tests to measure the
reversal of air flow, or how a tuned pulse might react when it reflects?

YES we do. Absolutely! If you don't get it in the beginning you cant fill the cylinder later. Port has significant measurable effect on the pulses (amplitude of pulse) due to smooth transition of air flow as well as average and peak air speeds in the system. In other words, a slow, lazy, square port head with radical changes in both shape and area will not be able to instill a high energy level to reflect in the first place. I am sure many can elaborate and put this into a more descriptive format than I, but that about sums it up. Are there ways to accent the wave action and pulse amplitude? Absolutely. Are they simple enough to transcribe on a message board. No they are not because there are a multitude of ways of accomplishing the same task and it varies with engine design.

(2) How many different test depressions are used to flow the port?
Do you also flow the head at each valve lift increment at each depression?
Surely, you all do. Which depression is most significant, and which is typically
used to shape the port; 28 in./H20?

We use a standard test pressure of 28". Its about 1psi and every one uses the same test pressure for comparison. We do not vary the test pressure in attempt to simulate the dynamic pressure differential within the system. That would be impossible to recreate and the data would not be comparable. That scenario would be tantamount to over complicating an Anvil.

(3 Is there a major issue with flow/turbulence at the junction of the intake
port and head port if the surfaces do not make a smooth transition?
If not, what is a typical variance, or allowable step?

No there is not. The intake manifold should be an extension of the port. Minor changes in both direction and air speed are allowed. The engine combination your dealing with will be the ultimate factor in deciding what is an allowable and necessary in regards to change in direction and air speed. Of course the runner being slightly (ever so slightly) smaller than the port is the best case scenario.

I have a slight lip going into the head port (approximately 1/16th of an
inch) as the intake runner is slightly ported. This was purchased used,
and I was basically too lazy to match the mating port.

You can not have the port present itself to the air flow. Period. Don't let this happen. Take all measures to remedy this. There are epoxy resins capable of fixing this with very little time and expense. Never , ever, ever,ever bolt an intake up to a cylinder head that is larger than the port. Its horsepower suicide!

[headman]

At one time I subscribed to the notion of the intake manifold runner exit being the same or slightly smaller than the head's intake port entrance.
The purpose to avoid "hitting the wall".

Maybe this is still a suitable bandaid if the head's intake entrance is too large.



Yet, if the head port is exactly the size required, the intake should not have a smaller cross section at the exit either.

[Procision-Auto]

Thanks for the reply! How cool is this; having a reply from the man I
listened to driving in the car today.

We use a standard test pressure of 28". Its about 1psi and every one uses the same test pressure for comparison. We do not vary the test pressure in attempt to simulate the dynamic pressure differential within the system. That would be impossible to recreate and the data would not be comparable. That scenario would be tantamount to over complicating an Anvil.

I realize that 28" is the industry standard for comparison, but I was wondering
if that depression was also the most significant used to shape the port.

I heard you talk about low lift; mid lift; and high lift flow, mean velocities, etc.

Are most of your port designs figured mainly around 28", or are there several different depressions used to shape a port.

You can not have the port present itself to the air flow. Period. Don't let this happen. Take all measures to remedy this. There are epoxy resins capable of fixing this with very little time and expense. Never , ever, ever,ever bolt an intake up to a cylinder head that is larger than the port. Its horsepower suicide!

After hearing that turbulence can cause up to 15% HP loss in an engine,
I really started to question whether I should match port the manifold to
the head. I knew the step was a problem, but I had no idea it was,
"Horsepower suicide"! If I recall correctly, you had a PS motor drop 250
HP
due to a turbulent port!

That makes my decision easier to rip apart the top end and even invest in a
better manifold.

Since I've built this engine, I've gained even more data on tuning.
Now that I have the Reher-Morrison package in hand, it makes it very
difficult NOT to tear down the motor and apply the new methods and tips
learned.

[MadBill]

Yes, reading Darin's words about the effects of a step down at the manifold port junction made me wonder if the exceptional power (578 @ 6,400 RPM) we saw on a less than radical LS6 build up last spring had more to do with the perfect port match easily obtained due to the two-piece F.A.S.T. intake than much of the magic/luck of well-matched components!

Re test pressure, yes also. Even though 28" is the near-universal yardstick for comparing heads and has been shown to be a realistic and practical one, I was trying to say that if pressure analysis shows for example that the delta at low lift is only 12", we may be fooling ourselves developing seat contours at 28". Conversely, we are pleased to see our port doesn't die at 0.800" lift with 28", but knowing the pressure difference at high lift is 100"+, we might feel better if we could verify that it doesn't go turbulent at 40.

[Fatman]

(3 Is there a major issue with flow/turbulence at the junction of the intake
port and head port if the surfaces do not make a smooth transition?
If not, what is a typical variance, or allowable step?

No there is not. The intake manifold should be an extension of the port. Minor changes in both direction and air speed are allowed. The engine combination your dealing with will be the ultimate factor in deciding what is an allowable and necessary in regards to change in direction and air speed. Of course the runner being slightly (ever so slightly) smaller than the port is the best case scenario.

I have a slight lip going into the head port (approximately 1/16th of an
inch) as the intake runner is slightly ported. This was purchased used,
and I was basically too lazy to match the mating port.

You can not have the port present itself to the air flow. Period. Don't let this happen. Take all measures to remedy this. There are epoxy resins capable of fixing this with very little time and expense. Never , ever, ever,ever bolt an intake up to a cylinder head that is larger than the port. Its horsepower suicide!

Darin:
Is the runner being ever so slightly smaller than the port better than a perfect size matched transition? Assuming you could get the manifold on in the perfect location.

[cboggs]

Wow,..

nice thread starting here, ..

Port shape Plays a significant role in the shock wave tuning
and it's effects.

Darin makes reference to shape, .. consider the difference in port shape from
a 23* small block head to a SB2.2 port.
The 23* is very low, has sharp corner radius and some very severe changes in
shape and area. The SB2.2 on the other hand has a very gentle sloping port
to the valve with a large corner radius, no sharp bends and no
severe changes in shape. The SB2.2 has a much better shock wave pulse.

Darin, .. have you used the pressure transducers in the chamber to
measure chamber pressure?, . . the one that is built into a spark plug??
Can't the data give an indication of cylinder filling and shock wave effects?

As for test pressure, .. it's been hashed over and over here several times.
here's what I have found, .. I use the standard 28" for most tests, ..
and sometimes do a second test at 40" to check for turbulence.
I didn't find any use flowing above 40" for anything, .. I tried over 50" on
some DRCE stuff and SB2.2 and didn't really find anything, ..
in fact buy following the turbulence is bad theory I hurt some heads.

Curtis

[Procision-Auto]

Re test pressure, yes also. Even though 28" is the near-universal yardstick for comparing heads and has been shown to be a realistic and practical one, I was trying to say that if pressure analysis shows for example that the delta at low lift is only 12", we may be fooling ourselves developing seat contours at 28". Conversely, we are pleased to see our port doesn't die at 0.800" lift with 28", but knowing the pressure difference at high lift is 100"+, we might feel better if we could verify that it doesn't go turbulent at 40.

Yes, these are good points. I wonder how much the port changes once
the head is tested on dyno and track. IOW, is a majority of the port
complete after flow testing, or do major changes come into play after
track testing?

If I had to guess, I would think the flow bench is mainly used to optimize
air speed through the port and help eliminate trouble spots for turbulence.

[SWR]

Port shape Plays a significant role in the shock wave tuning and it's effects.

So in effect,if a port is "straight" i.e. no area change or taper,and the intake manifold tapers down to the gasket flange, vs. a fully-tapered-from-the-valve-up port/intake combo will tune differently (port volume/cfm being the same for both the tapered and straight port)?
What differences in tuning effects are you seeing?
Does the runner need to be more tapered,or longer/shorter than it would optimally be with a "straight" port/tapered intake manifold than a fully tapered combo?
How big an effect (in hp or %) would you think the difference to be with either combo,at the same cfm?

Sorry if I'm asking a bit too much today...have a headache and just don't want to think

[rmcomprandy]

Darin...or anybody else;
Why because of results with their testing MERCRUISER uses a Rectangular port B.B.Chevy intake manifold on an OVAL port head with a giant missmatch because in their words, "mid-range torque and top-end horsepower was increased without changing the low-end output".
Ford Marine also did the same thing with using a Cobra-Jet intake manifold on a regular D0VE port size head on their big-block to create the same giant missmatch for somewhat the same reasoning -

WHY did this occur...???

I was told that this mismatch was large enough to create TWO seperate sound wave pulses; can this happen...?

[Darin Morgan]

(3 Is there a major issue with flow/turbulence at the junction of the intake
port and head port if the surfaces do not make a smooth transition?
If not, what is a typical variance, or allowable step?

No there is not. The intake manifold should be an extension of the port. Minor changes in both direction and air speed are allowed. The engine combination your dealing with will be the ultimate factor in deciding what is an allowable and necessary in regards to change in direction and air speed. Of course the runner being slightly (ever so slightly) smaller than the port is the best case scenario.

I have a slight lip going into the head port (approximately 1/16th of an
inch) as the intake runner is slightly ported. This was purchased used,
and I was basically too lazy to match the mating port.

You can not have the port present itself to the air flow. Period. Don't let this happen. Take all measures to remedy this. There are epoxy resins capable of fixing this with very little time and expense. Never , ever, ever,ever bolt an intake up to a cylinder head that is larger than the port. Its horsepower suicide!

Darin:
Is the runner being ever so slightly smaller than the port better than a perfect size matched transition? Assuming you could get the manifold on in the perfect location.

The simple answer is No. Having the intake runner slightly smaller is not the correct approach, its the safe approach. Its the lesser of two evils but given both options I would prefer the port and manifold be sized properly and matched perfectly. If the port is to large and not sized perfectly for the engine combination, having a manifold that is smaller than the port can have a band-aid effect on the air speed and help the engine produce power. This was most often seen with Pro Stock cylinder heads in the late 1980s and early 90s due to using larger than necessary square port heads. Many engine builders (as is still the case today) had no concept of proper sizing and tended to over valve, over manifold and over port an engine. They found that leaving a step (most often on the floor) would increase both torque and power. They reasoned that this power increase was due to a decrease in reversion. There theory was only partially correct. Yes it did lower the amount of reversion but it did so by increasing the average air speed in the system not by the blocking effects of the step. I also surmise that there where some atomization benefits to the step as well but that is unproven.
Its all ancient history but I do see some people still using this technique today. With proper port sizing and induction system design there is absolutely no need what so ever for steps or miss matches of any kind.

[Darin Morgan]

Darin...or anybody else;
Why because of results with their testing MERCRUISER uses a Rectangular port B.B.Chevy intake manifold on an OVAL port head with a giant missmatch because in their words, "mid-range torque and top-end horsepower was increased without changing the low-end output".
Ford Marine also did the same thing with using a Cobra-Jet intake manifold on a regular D0VE port size head on their big-block to create the same giant missmatch for somewhat the same reasoning -

WHY did this occur...???

I was told that this mismatch was large enough to create TWO seperate sound wave pulses; can this happen...?

I think you have that backwards. Dont you mean to say that they use an oval intake on square port heads?

[Procision-Auto]

Darin...or anybody else;
Why because of results with their testing MERCRUISER uses a Rectangular port B.B.Chevy intake manifold on an OVAL port head with a giant missmatch because in their words, "mid-range torque and top-end horsepower was increased without changing the low-end output".

Let me take a stab at this one...

If I'm applying theory correctly: The step, or change in cross section in
the runner creates a tuned length at this point.

The air speed slows down as it enters a larger volume, and will also create
a density change in the gas/fluid.

When a pulse bounces back up the runner, it will reflect upon this change
in gas/fluid density creating a tuned length.

Similar techniques are used in exhaust primaries and secondary systems
in forms of steps and merge collectors.

[rmcomprandy]

That could be the case IF the port was larger than the manifold.
In these marine engines the intake manifold port is over an entire square inch LARGER than the head port. ???

Still don't understand why it works for them but, the testing shows - IT DOES what it's supposed to do.
Granted...they start their testing way down at 1000 RPM.