Nissan L-Series Cylinder Head Flow Bench Testing and Development

[MONZTER]

I've decided to build a detailed thread to document L-Series cylinder head flow testing and the results. Like many aspects of these engines, it's sometimes hard to know what to believe or how to interpret what you see and read. Test setup, equipment, testing methods, and just straight-up BS can all contribute to the confusion.

I don't sell a product or service and have nothing to gain from the results I post, so my goal is to keep the testing unbiased and as consistent and repeatable as possible.

Feel free to comment or criticize as I move through this. I'm happy for any input.

 

First up is the equipment.

 

I have a SuperFlow 600 converted to run the digital FlowCom directly to my laptop where I use PortFlow analyzer from Performance trends.

 

This software sets up the test and aids in the data logging and reporting.  It also compensates for Altitude, Barometric pressure, ect.  So no math or reading of the old school manometers.  Its super accurate and very repeatable.

 

 

I built a L-Series specific test stand and replaceable cylinder bores to test the common sizes, 84mm - 240z , 86.5mm 280z, and 89mm - Big Bore.  This has a considerable effect to the max flow due to bore shrouding, so its important to test on the correct size.  This is the first place numbers can get "offset"  If some guy is flow testing a Z head on his 4" Chevy bore, you bet the numbers will look good.  More on this later.

 

These bores are made of nylon and seal to the test stand with an o-ring.  However any leakage can still occur, so the software takes a zero measurement first, and this leakage is then subtracted out automatically

 

 

Next up is calibration - the bench should be calibrated on a regular basis to check for issues.  You do this by bolting the supplied test plate that is matched to the machine and simply flowing it.

 

Here you can see the plate bolted on, it has calibration flow percentages written on it to compare

 

 

 

Ans this is the screen on the laptop, you can see the bench is within .1 percent calibration for the intake and exact for the exhaust

 

  

 

 

Ill flow and document the heads I have as well as some I will borrow from friends.  Some will be stock, some will be modified to various states.

 

Ill also do some test to help point out important details.  So stay tuned

 

Thanks - monZter

Edited September 6 by MONZTER

[MONZTER]

The first head to document will be an old E88 from my 72.  I used this head a long time ago worth 40DCOE triple Webbers.  In typical fashion long long ago I did my best to do what I thought was right.  I blended the seats to match the bowls with a grinder, and port matched the head to the intake, opening up the port about 1" in from the entrance to blend the new opening (not good). other than that, stock valves and chambers

 

So lets flow it.  Hoping to borrow a stock E88 to compare later.

 

Some pics for referance

 

   

 

   

 

Here are the head Specs

 

 

Cylinder 2 Intake with 84mm Bore

 

 

And Cylinder #2 Exhaust with 84mm Bore

 

 

I then switched out the bore to 86.5mm and tested again, here is a comparison chart of the effect of only a larger bore

 

 

 

 

 

Sorry the image is hard to read I need to figure out how to upload better resolution pics, but you can download the PowerPoint below

E88_Flow_Bench_Comparison_Cyl2_84mm_vs_86.5mm_FIXED (1).pptx

Edited September 8 by MONZTER

[DuffyMahoney]

Super cool, thank you for posting.

[calZ]

This has the potential to be one of the best threads in this site's history. It would be awesome to end up with complete database of head flow for the different combos. 

[DuffyMahoney]

My E31

[jeffer949]

Very cool. Are you getting a pamz head?  I forget if that was you or not.  Be interesting to see how it does. One thing ive always been curious on is the difference between the runner size on the exhaist on a n42 p90. The middle two are much wider than the outer 4. Picture is of an n42 and a p90 im rebuilding.   

 

Im starting a cylinder head rebuilding buisness and while im not going after the high performance market its amazing the difference a valve job can make. We have come along way since the 30/45/60 valve job. I have seat cutters with 6 or more angles and radius exhaust and so o. Im not flowing heads but thankfully there are guys online doing so. Its amazing the flow they get by a little tweaking hear and there

[DuffyMahoney]

No PAMS head for me.  I am pretty proud to own one of Andy Kazanis's heads.  I believe he and a few others think the E31 flows the best of any of the heads, with enough effort.  

 

 

[MONZTER]

Sorry, been busy as usual, I have tested my Welded N42 I ran on my 240 for years, I have also tested my PAMS head.  Ill post the reports when I get a moment.  Hopefully soon.  My concern is that as some know and some do not.  Flow and cfm do not tell the whole story.  Port volume and thus port velocity along with a proper MCA for the actual displacement and peak HP RPM all combined with needed flow is the best answer, not just who flows more.  So, Ill discuss this in detail as well so we can look at each head and its proper position on top of what engine.  Its a lot to digest, so still thinking how to present it all.  I also need to figure out how to get some decent resolution images to post with all the number legible.  Thanks for the people who posted some numbers for referance.

 

[DuffyMahoney]

Yeah the PAMS have huge ports, the Japanese tuners seem to prefer huge ports.

 

I think my port sizes being roughly stock and flowing as high as they do, so work very well. 

[MONZTER]

Duffy, If you want I will run your head in my calculators and Ill tell you what it shows.  I need to know

 

Target / peak HP RPM

Measured Intake flow at 28"

Approximate minimum port diameter, not necessarily the flange port opening

 

 This will tell me what engine capacity it will like

 

If you tell me these specs

Bore

Stroke

Rod Length

Target / peak HP RPM

 

I can tell you what flow it wants as well as the recommended minimum port diameter based on velocity

 

Port velocity is just as important or more than flow, too much or too little is not always better

 

[DuffyMahoney]

My 3.1L

 

10.2:1 compression ratio 

88.5mm bore. N42 block. Sonic tested. 

 

LD crank by marine crankshaft. Crank Off-set grind to 84.6 stroke
Finish rod journals @ 1.8889" 

Thrust Finish 1.27”
R & R plugs with threaded plugs
Dynamic Balance
Ion-Plasma nitride, straighten & micro-Micro polish all journals

 

E31 head. Port work by Andrew Kazanis

Dsi valves Diameter:   45/36.5mm Length:  118mm, titanium retainers. Isky springs. 

Bonk cam. Roughly stock exhaust and intake port sizes. Vapor honed inside and out. 

 

221 CFM flow at .45-.5 lift. 

 

Rebello JE Eagle pistons and toyota rods 

 

Mine is 88.5x88.5x.25x3.14159x 84.6x6 /1000 = 3122cc 

 

My intake is port matched to my head, throttle bodies are tapered bore to my intake, they are 50-48mm (at the butterfly) then 45mm at the DCOE flange. 

[MONZTER]

 

Duffy, thanks for the numbers.

 

Let's see how this works out.

 

My plan is to approach this two ways. First, "I have this cylinder head, what bottom end does it want?" Second, "I want to build this engine, what cylinder head do I need to support it?"

 

Let's start with the first one.

 

I've made a bunch of Excel calculators to help automate this process. Since I don't have all the details of your head, I'll need to make a few educated guesses.

 

We have a 45mm intake valve and 221 CFM, which I'm treating as corrected to 28" H₂O. You mentioned roughly stock-sized ports, and this is where things get interesting. Flow numbers alone only tell part of the story.

 

I'm going to assume a 7000 RPM peak horsepower target and a 35mm MCA (minimum cross-sectional area), expressed as an equivalent round diameter. The actual minimum might be at the flange or somewhere farther into the port. We won't know without measuring it.

 

MCA is important because it determines the calculated average velocity through the smallest section of the port for a given airflow.

 

For this example, I'm using 320 ft/s as a conservative flowbench-based design target. The L-series has a relatively low intake port with a sharp turn over the short-side radius. Picture a race car going around a track. You can carry plenty of speed down the straight, but enter a tight corner too fast and you're going to have problems making the turn.

 

A modern high-port motorcycle head with a straighter approach to the valve might tolerate 350–360 ft/s. These are design guidelines, not hard limits. Port geometry and flow stability ultimately determine what's acceptable.

 

Also, these are steady-flowbench-based calculations, not instantaneous velocities in a running engine. Dynamic sizing methods use piston demand and RPM and can produce considerably higher numbers. We've used approximately 613 ft/s as a dynamic design reference, but that's not a direct conversion from 320 ft/s.

 

First example: 7000 RPM and 320 ft/s design target.

 

 

The calculator gives us a 2.854-liter engine.

 

Is that correct? Maybe, maybe not. It depends on what you're building.

 

With our assumed 35mm MCA and a conservative 320 ft/s design target, the calculator suggests approximately 2.85L as a starting point for a 7000-RPM power peak.

 

Notice the calculated velocity at MCA is 355.7 ft/s. That's based on the head's reported 221 CFM through our assumed 35mm minimum area. It's not a measured local velocity.

 

The calculator is actually sizing this engine around approximately 199 CFM of required flow, corresponding to our 320 ft/s target. The head's full reported flow capacity remains 221 CFM.

 

Now let's increase our design velocity target to 350 ft/s.

 

 

 

Now we get a 3.122-liter engine.

 

Same head, same assumed port size, same 221 CFM and the same 7000-RPM target. The difference is that we're allowing a higher calculated velocity at the engine's required airflow, approximately 217.5 CFM.

 

But can this particular port support that demand efficiently? That's the question.

 

If the airflow becomes unstable, especially around the short-side turn, cylinder filling may stop improving as RPM increases. This is where an engine can start running out of steam earlier than expected.

 

A 3.1 or 3.2-liter engine needs more than a good peak CFM number if you expect it to keep making power at high RPM. The port needs the correct area, shape and flow characteristics to support that demand.

 

OK, one more. Back to 320 ft/s, but lower peak horsepower RPM to 6250.

 

 

 

Now we're at 3.197 liters.

 

Nothing about the head changed. We're simply asking it to support peak horsepower at a lower RPM, where a larger engine can operate with less airflow demand per unit of time than it would at 7000 RPM.

 

Here's how the three examples compare:

 

 

The required-flow values come from the workbook's sizing model, not measurements of the running engine.

Notice that 355.7 ft/s never changes. It's calculated from the same 221 CFM and 35mm MCA in every example. What changes is the engine's calculated airflow requirement at its target RPM.

 

So what's my point?

 

Flow numbers alone don't tell you how much power an engine will make or where it will make it. You need the right combination of airflow capacity, port area, displacement and RPM.

 

None of these combinations is necessarily bad. They're just different engines that need different supporting components.

 

A smaller displacement with a longer-duration cam might be happier running higher RPM. A larger displacement might make a better street engine, producing strong torque and reaching peak power sooner. Each would need its cam timing, intake runner length and exhaust system matched to the intended powerband.

 

The mistake is assuming you can have everything simply by adding displacement, camshaft and RPM without considering airflow and velocity.

This is why I think people need to be clear with their engine builder about what they actually want the engine to do, then select the components to support that goal.

 

Getting back to your head, there's another important part of this.

 

MCA is the smallest cross-sectional area anywhere along the port. Average CSA is calculated from port volume divided by centerline length. They're different measurements and tell us different things.

 

Let's say your E31 has a 35mm MCA at the flange. At 221 CFM, that calculates to 356 ft/s. But if the port opens up to 37mm before the short-side turn, the calculated average velocity at that larger section drops to 318 ft/s. That's a significant difference on a low-port L-series head where the air has to make a sharp turn.

 

This is why we need to know not only the MCA, but where it's located and how the port area changes toward the valve. Even with lower average velocity, poor short-side geometry can still cause flow separation.

 

Port volume is useful too. For reference, a cleaned-up E88 measured approximately 104cc, a welded E88 measured 111cc, and the PAMS JP geometry calculates to approximately 129cc. These represent quite different port sizes, although volume alone doesn't tell us where the restrictions are.

 

Ideally, I'd like to know your actual MCA, its location, and the intake port volume. With those measurements, we can make a much better assessment of what your head can support instead of guessing from peak CFM alone.

 

And that gets us away from the bigger-is-better mentality and toward building an engine where all the components actually work together.

 

Next, we'll look at the second approach: "I want to build this engine, what cylinder head do I need to support it?"

[DuffyMahoney]

Looks like I am 37.08

 

This is the porting plate used to port match my intake to my head.  Each aspect is port matched. Picture is prior to port matching, but shows the bends

 

My porter (Andy) secret sauce is how he does his short bend, but I have no idea exactly how he does it.  He seems to think the E31 can flow the best and is the best for "race engines"

 

My car is a street car, so overall my goal is fun drive around and doing canyon driving.  

Edited Tuesday at 07:46 PM by DuffyMahoney

[MONZTER]

Duffy, that helps quite a bit.

With the 37.08mm measurement being the equivalent diameter of your port opening, and that the flange is at or near the MCA, this changes my original calculations considerably.

 

At 221 CFM, the calculated velocity through that opening drops from 356 ft/s with my original 35mm assumption to about 317 ft/s. That's almost exactly the 320 ft/s design target I was using for the L-series port.

 

Using that new dimension, 221 CFM and a 7000 RPM peak horsepower target, my calculator comes up with a 3.172-liter engine. That's remarkably close to your actual 3.122L.

 

So based on these assumptions, your displacement and intake port size appear to be pretty well matched for a 7000 RPM power peak. That's encouraging, especially considering your goal is a responsive street car rather than an 8000 RPM race engine.  (At least what I think, for what its worth)

 

I also did a rough port-volume calculation using your 45mm intake valve, an assumed 90% throat (40.5mm), an 8mm valve stem and the port centerline length from my E88 measurements. Assuming a gradual expansion from the flange toward the throat, I get approximately 124cc. That's only a geometric estimate, since we don't know the actual CSA progression or how Andy shaped the bowl and short side.  We will use this number later to look at average port dimeter.

 

What Andy is doing around that short-side turn is particularly interesting. If he's increasing the area and controlling the flow as it approaches the valve, that could help explain how he's getting those flow numbers without an excessively large flange opening / MCA.

 

The one thing we still don't know is whether 37.08mm is actually the smallest section or whether the port necks down somewhere else.  On L-series it is typically pretty close to the flange or about an inch in before the turn.

 

But this new information makes your 221 CFM result much easier to reconcile with the port geometry and displacement, and it suggests your existing combination is reasonably close to the type of driving you described.

 

The next step would be looking at your actual cam timing and intake runner length to see how well they're matched to that 7000 RPM target.

 

We can do that in the rest of the tabs in my calculator Im working on

 

 

But look when we put that head on a 2.8, A real ripper at 8k

 

[DuffyMahoney]

How would you like intake runner length? Total from the stacks to the head?  

 

Cam is .525" lift and duration of 240 @ .05 valve lift

Edited Wednesday at 10:51 PM by DuffyMahoney

[MONZTER]

Hi Duffy, I calculate it  from seat to bell mouth end but I already know port length, so Intake runner length from Head flange to Throttle body.  (I was guessing 110mm), TB Length of 88mm long, another guess, and 65mm velocity stacks a third guess, so it will be good to know exactly, thanks

[DuffyMahoney]

I am getting roughly 294.5mm +-.25mm  total length from head flange to velocity stack bell.  

 

Throttles with the stacks are =153mm

Phenolic spacers =10mm

Intake = (harder to measure since it isn't straight at all, center of bore to center of bore is 131.5mm

 

Stacks have a max opening of 80mm, they taper down to 50mm at the throttle opening, then 48mm throttle blades, then 45mm at the intake to throttle flange.  Then slowly taper to the port size of 37.08

 

 

Edited Thursday at 05:00 PM by DuffyMahoney

[MONZTER]

Duffy thanks, this lets me complete most of the analysis.  Lets have a look at the basics first, I've adjusted details too better match your set up.  I've also set it to track given your cam choice as it is better suited for higher RPM power

 

 

 

and the resultant port sizing based on what you told me (Intake)

 

Exhaust

 

Intake Valve

 

Exhaust valve

 

Bowls

 

 

so far all your specs align quite nicely with the workbook prediction for an optimal engine for trackish use with a 7000rpm peak power band.

 

Ill get you runner length specs tomorrow and simulated dyno run numbers

 

 

[DuffyMahoney]

11 hours ago, MONZTER said:

Duffy thanks, this lets me complete most of the analysis.  Lets have a look at the basics first, I've adjusted details too better match your set up.  I've also set it to track given your cam choice as it is better suited for higher RPM power

 

 

 

and the resultant port sizing based on what you told me (Intake)

 

Exhaust

FindEngine Parts

 

 

Intake Valve

 

Exhaust valve

FindEngine Parts

 

 

Bowls

 

 

so far all your specs align quite nicely with the workbook prediction for an optimal engine for trackish use with a 7000rpm peak power band.

FindEngine Parts

 

 

Ill get you runner length specs tomorrow and simulated dyno run numbers

 

 

Super cool stuff

[ihavearustedz]

Super interesting read! Thank you. Today I learned about MCA (minimum cross-sectional area) in the port and how it relates to choosing a bottom end. Love it. What's the easiest way to measure the MCA in the port?