tuning hilborn EFI stack length

Hi All,

running hilborn EFI 2 7/16 IR injection set-up on my 434 SBC in my canam replica. I have used the pipemax program ( PipeMax36xp2 ) in the construction of my headers and am now planning to use it to trim stack length so I tune to either the second or third harmonic. My engine is a low revver and was dyno'd at 595hp at 6300 and 565 ft/lbs at 4800 with a world products single plane intake and a 1050 dominator on it. I have the dyno data loaded into the software and am getting the following advice from the software (lengths are inches):

---------------------

- Induction System Tuned Lengths - ( Cylinder Head Port + Manifold Runner )
1st Harmonic= 35.097 (usually this Length is never used)
2nd Harmonic= 19.920 (some Sprint Engines and Factory OEM's w/Injectors)
3rd Harmonic= 13.907 (ProStock or Comp SheetMetal Intake)
4th Harmonic= 10.946 (Single-plane Intakes , less Torque)
Note> 2nd and 3rd Harmonics typically create the most Peak Torque
4th Harmonic is used to package Induction System underneath Hood

---------------------

I followed up with an email to Larry Meaux who advised that the third harmonic usually creates the most average torque across the range, and that the 2nd usually has the highest peak but is "peakier".

It is around 12 inches from the back of the intake valve to the top of the hilborn intake, meaning stack length will be 8 inches for second harmonic and only 2 inches for the supposedly broader curve delivered by 3rd harmonic tuning.

Clearly the original guys with the big stacks were tuning at or around the second harmonic ( http://rcd.typepad.com/rcd/2007_06_26_IMG_3412_20_28Small_29.jpg ), and did seem to have peakiness issues with the BBC until they went to staggered length to compensate for the BBCs two different intake port designs ( http://www.airportjournals.com/Photos/0609/X/0609022_7.jpg ).

My thinking is that an 8 inch stacks would look better and be more in keeping with cars of the day, but 2 inches might work better and would be easier to build a low profile plexi cold air box around.

Any experiences or thoughts?

Cheers, Andrew Robertson
 

Terry Oxandale

Skinny Man
Thanks for sharing that with us Andrew. Have you already posted the results of your program on header tube length. If not, I'm sure they would also be of interest.

Terry
 
Thanks for sharing that with us Andrew. Have you already posted the results of your program on header tube length. If not, I'm sure they would also be of interest.

Terry

Hi Terry,

here is the header recommendation for engine from the Pipemax program:

----------------------------------------------


--- Single Primary Pipe Specs --- for 433.893 CID from 4300 to 6800 RPM
Diameter= 1.807 to 1.932 Length= 31.8 to 34.6 inches long
--- 2-Step Primary Pipe Specs ---
1st Dia. inches= 1.807 Length= 15.9 to 17.3
2nd Dia. inches= 1.932 Length= 15.9 to 17.3
--- 3-Step Primary Pipe Specs ---
1st Dia. inches= 1.807 Length= 15.9 to 17.3
2nd Dia. inches= 1.932 Length= 7.9 to 8.6
3rd Dia. inches= 2.057 Length= 7.9 to 8.6
--- Header Collector Specs (Conventional Straight Tube) ---
Diameter= 3.309 to 3.559 Tuned Lengths= 17.9 best and also 9.0 or 35.8
--- Header Collector Specs (Megaphone or Diffuser Cone Shape) ---
Diameter= 2.809 taper to 3.809 Megaphone/Diffuser Length= 17.9 inches
Best HP/TQ Tuned Collector Lengths= 17.9 , 35.8 , 71.6 , 143.3 inches long
Worst HP/TQ Loss Collector Lengths= 26.9 , 53.7 , 107.5 , 214.9 inches long
Note-> all Pipe Diameters are OD and based-off .0625 inch Pipe thickness
---- Primary Pipe's Harmonics ----
1st Harmonic = 138.8 inches long ... typically never used
2nd Harmonic = 52.9 inches long ... longest recommended
3rd Harmonic = 31.8 inches long ... highly recommended , best Torque Curve
4th Harmonic = 22.2 inches long ... shortest recommended
5th Harmonic = 16.8 inches long ... typically never used
6th Harmonic = 13.3 inches long ... typically never used
7th Harmonic = 10.9 inches long ... typically never used
8th Harmonic = 9.0 inches long ... typically never used
---- Collector's Harmonics (includes Intermediate, Muffler , TailPipe) ----
1st Harmonic = 143.3 inches long ... longest with Mufflers and TailPipes
2nd Harmonic = 71.6 inches long ... longest recommended with Mufflers
3rd Harmonic = 35.8 inches long ... more bottom-end Torque
4th Harmonic = 17.9 inches long ... highly recommended , best Torque Curve
5th Harmonic = 9.0 inches long ... reduced Torque , more top-end HP sometimes
6th Harmonic = 4.5 inches long ... reduced Torque , not recommended

-----------------------------------------

based on this I built a non-stepped design with 32 inch primaries in 1 7/8 inch, and an 18 inch collector into a chambered muffler which I treated as the end point in the system because of it's volume.

Cheers, Andrew
 

Terry Oxandale

Skinny Man
Great info. I will be using smaller primaries (1 3/4") and a smaller displacement, but I'm thinking the two cancel each other out and that the lengths will be similar.
 

Kevin Box

Supporter
Andrew

For what its worth we used the following lengths in Jet boat engine for 358 cu.in.
Intake valve 2.125 (cut down 4V)
Exhaust valve 1.7 (4V)

Exhaust
Primary 34
Collector (open) 20


Intake (measured from the back of the valve) 15.5
Intake diameter 2.5 inch


Exhaust was calculated and the intake was initially from calculation suppopsed to be around 13 inches. I used 8 bits of hose to lift the rams up and down to get a peak and it ended up at 15.5

We tried sliding an extension on the exhaust but above 22 peak dropped off.

Our aim was for peak torque at 5500 and becuase it was a jet boat was not too worried about bottom end

Cam was was 238 @ .050 and 630 lift
Note: To get the torque in the right place it ended up advanced approx 4 Deg further than Kelford suggested. (pre Kevin Ban days)
Never had it on the dyno in that config but Jet unit suggested power was around 460 HP @ 6000
All I remember is lots of playing to make the engine work for a jet unit as your a re really trying for torque and HP peak almost within 1000 RPM

Hope the lengths give you a comparison to start with.

KB
 
Andrew

For what its worth we used the following lengths in Jet boat engine for 358 cu.in.
Intake (measured from the back of the valve) 15.5....the intake was initially from calculation suppopsed to be around 13 inches. I used 8 bits of hose to lift the rams up and down to get a peak and it ended up at 15.5

Hope the lengths give you a comparison to start with.

KB

Hey Boxy, how's it going? Thanks for this. I made up a spreadsheet and it looks like around 15 inches (back of valve to end of stack) is about right, and your experience sounds like that's in the ballpark. This length puts the third harmonic at approx 5600, and the fourth harmonic at around 4400, which is either side of peak tq at 4800 so should produce something quite broad - at least in theory. I need it too - shifting at 6300 the porsche 930 box drops RPM back to the 4400 on 2-3 change.

Cheers, Andrew
 
Hi Andrew
I think you guys are pretty close with your stacks. With our sprintcar engines at 410 Cu In 10" is close to peak power but torque picks up right through to 18". so 15" is close. Of course every engine is different and as soon as you say this is the right recipe the bloody cake wont rise!
However there is a formula that I use for headers, and this is a starting point, a rule of thumb. I have never come across two engines that run identical on the dyno. If you dont have acess to a flow bench or computer programs and the like, use this method.
The primary length of your headers are usualy from 1 1/2" to 2 1/8" and from 26" to 36" long. So the primary length should be:
1.8 x the volume of your engines cylinder for mild street use.
2 x for hot street
2.4 x for Drag race.
Use small diameter and length for torque and large diameter and short for upper HP
Example
434 Cu In 434 / 8 = 54.25 Cu In per cylinder
Hot Street x 2 54.25 x 2 = 108.5
Primary header length 1 7/8" x 32". 1 7/8" is 1.875"
1.875 x 1.875 x .7854 x 32" = 88.35 Cu In
1.875 x 1.875 x .7854 x 36" = 99.4 Cu In closer to your 108.5
After 36" its best to go up in diameter
2" x 2" x .7854 x 34 = 106.8 Cu In
So 2" dia x 34" long is twice the volume of one cylinder good starting point for a hot street engine.
Iv'e used this method many a time and have had good results.
I hope there is something in there that may help
Woody.
 
Not specifically trying to be a wet blanket here, but...

Practically all of the currently available engine simulation programs -- like Larry Meaux's PipeMax -- on the market today that are priced within the range of the average enthusiast have some shortcomings that become readily apparent when you look at the physics involved in the systems they attempt to model.

With regard to exhaust system design, you need the header to accomplish two tasks; first, route the spent exhaust gasses from cylinder head to some predetermined point during the time allotted (fancy way of describing RPM) with a minimum amount of energy supplied by the piston to affect this transfer of mass. In plain English, you want an exhaust system with as little back pressure as possible. That's the easy part. From here on though, things become complicated quickly. As everyone reading this forum knows, the flow of exhaust gasses through the header is not steady state. The pressure and resonance waves that are the backbone of "exhaust scavenging" are affected by so many variables that are not even considered by these programs (other than as "default value" assumptions buried in the calculation equations) that the resulting stack-up of these fudge factors often knee-caps the accuracy of entire simulation. Ideally, you want the reflected negative pressure wave that is created at the end of the primary tube to arrive back at the cylinder during overlap during the RPM range you want the system tuned for. Collector lengths and rates of taper play their own part in this, but let's not go there just yet...there is plenty to think about here for the moment.

Let me give you an example: The resonance waves I mentioned above move at the speed of sound...but what is that? The speed of sound in dry air at standard pressure and temperature is usually given as:

[331.4 + (Temp C x .06)] meter/sec.

(For those interested, a utility to convert to imperial units of measure is available here: Speed of Sound

For clarification, it's easier to consider the question in terms of the "local" speed of sound at a given point and given direction of moment within the primary pipe or collector. As the wave moves back up the primary pipe toward the cylinder it encounters changes in both the velocity and direction of the medium (exhaust gas) in which is it moving. This effect alone can be drastic...there are areas within the system where the velocity of the exhaust gas itself can exceed the velocity of the wave moving through it in the opposite direction, so that the wave actually "loses ground". At other points in the system, the opposite is also true...the wave will be helped along by a "tail wind". These "head wind" and "tail wind" values rarely if ever cancel each other out over the length of the primary pipe, so using a single "default" value table for wave velocity based on primary pipe length/cross-sectional area and pressure differential...even one with a EGT modifier...is practically useless.

Now, let's toss some other variables like pressure, density, and temperature...all of which vary over the length of the system and with other design parameters (such as thermal transfer characteristics of the construction material itself) into the mix. It becomes easy to see why simple programs (usually based on some derivative of Hermann von Helmholtz' resonator theory) that do not consider these factors can only get you so close...and I use the term "close" here in the hand grenade sense. At RPM levels over 7000, a couple of inches of primary tube length either way can be the difference between the reflected wave getting back to the cylinder at the proper time, getting back to the cylinder after the exhaust valve has already closed, or before the intake valve has opened. Nature is often a real mother when it comes to details...especially with the physical sciences. The same factors affect intake tract tuning but the dynamic values on that side of the engine are far lower, so the magnitude of potential errors is less. This is why acceptable estimates for intake tract length are often obtainable from simpler formulae.

Is this all much adieu about nothing? Here’s why I personally think not: Most members here are at least familiar enough with camshaft technology to recognize that it’s quite common to see cams ground (particularly for naturally-aspirated small block Ford applications) with considerably more exhaust duration than intake duration. While certain applications will always call for at least some of this, we’ve found that it isn’t the absolute necessity many engine builders think it is – if the exhaust system performs at optimum efficiency. Like every other part of the engine’s gas exchange process, cylinder scavenging during the valve event overlap period is a function of pressure differential. This magnitude and duration of the pressure differential as created by the exhaust system can be manipulated to a great extent by its design, and when done properly it allows for effective cylinder scavenging requiring fewer degrees of overlap. The camshafts we have ground for our complete head/manifold/carb/header road course packages nearly all feature less exhaust duration as compared to intake, and are ground on wide lobe separation angles (116-118), often resulting in less than 40 degrees of valve overlap and getting the exhaust valve back on the seat less than 10 degrees ATDC. These engines are extremely efficient over a wide RPM range, and many of them would be very drivable on the street. While proper exhaust system design will never be a substitute for multi-valve cylinder heads and VVT, in my experience it does open the door to aspects of engine performance that have often been left on the table.

In my experience, there are three commercially available programs that allow/require the necessary data input and are sufficiently robust to predict exhaust system performance to a professional degree of accuracy:

ANSYS V.12 (a computational fluid dynamics analysis package)

Ricardo Wave (a comprehensive induction and exhaust system simulation package)

Optimum Power Technology Virtual 4 Stroke with ANSYS/Fluent add in

The least expensive of these software packages will set you back a bit over $40K for a single 12 month seat, and all three utilize an arrayed processor network made up of the service subscribers (and requiring some sort of X-Server connection) in order to access the processing capacity necessary to perform such simulations in a manageable amount of time.

I know and like Larry Meaux and I respect his work. I hope he doesn't feel "hated on" because I can't buy-in here.
 
In my experience, there are three commercially available programs that allow/require the necessary data input and are sufficiently robust to predict exhaust system performance to a professional degree of accuracy:

ANSYS V.12 (a computational fluid dynamics analysis package)

Ricardo Wave (a comprehensive induction and exhaust system simulation package)

Optimum Power Technology Virtual 4 Stroke with ANSYS/Fluent add in

The least expensive of these software packages will set you back a bit over $40K for a single 12 month seat, and all three utilize an arrayed processor network made up of the service subscribers (and requiring some sort of X-Server connection) in order to access the processing capacity necessary to perform such simulations in a manageable amount of time.

Hi Stan,

Thanks for this - great insight - and didn't think you could spend $40k in annual license fees on engine simulation software!

I'm using a triangulation approach - i.e. using $40 software, using chryslers old and very basic maths on ram tuning, asking the list, scanning the net, reading guys like Vizard Yunick Jenkins etc etc - I think thats a fairly solid way of getting into the ballpark, especially given the wealth of empirical data available given how common the engine combination is. When all this stuff points to a general sweet spot, that's when I cut to length and move on to the next challenge! I know I'm leaving a wee bit on the table here and there, but I'm building a car in my garage and that's the compromise I have to make.

Cheers, Andrew
 

Kevin Box

Supporter
Andrew

Yes you are correct.
Assemble the best bit of info that is available at the time
Sort through and chuck out the BS
Make use of the rest to formulate your plan and DO IT !!
If in "Doubt" try it and if it doesnt work put it down to experience or find out why not.

Your lengths seem to be in the right ball park so go with them.

good luck

KB
 
Back
Top