Manifold port size and peak torque rpm - air velocity limit: NA vs. FI

[edjza80]

Hi Folks,

my own thoughts on this topic, and some others discussion have stalled... and thus, i open the floor to you guys.

basically I have no argument with the conventional numbers of manifold port and valve throat peak airflow velocities (in NA engines), and their relationship to rpm and VE. However, I am having all manner of problems understanding what happens in a forced induction application.

how does FI change your expected peak airflow velocity? does FI still follow the general rule of decreasing returns over a manifold port velocity of 300 (mach 0.4ish), or a port/valve throat vel of mach 0.6? as it appears that FI engines still makes power at rpms where predicted velocity values say it shouldnt.

The only way my head is resoving it is by assuming the manifold velocities are still hitting a ceiling around the mach 0.4-0.5 point, and from there, any further increases in pressure result in increased VE only through increased charge density rather than escalating velocity?

Related to this is a basic question regarding predicted changes in the reynolds number and boundary layer thickness of an FI intake charge vs an NA charge? Is there a difference due to the abs pressure and predicted velocities of the charges? and given this can you address the theoretical required manifold port sizes for high performance FI engines, and qualitatively, how would this differ to an NA manifold port size?

original thread below...

http://www.v-eight.com/tech_forum/viewtopic.php?t=268

Regards
Ed

 

[Lou Scannon]

I'm not sure I'm following your question properly... perhaps you could explain the mechanisms you think are affecting velocity and Mach number, when going from N/A to FI with no other changes.

 

[edjza80]

ok, problem is we've done some calcs, as well as studied a few built engines with known paramteres (both NA andFI)

the results have suggested that the manifold velocities are getting 'very' high on the FI engines, and the torque and peak power positions on the RPM axis do not seem to relate to manifold and port velocities as expected with NA engines.

for eg, on an NA engine, to increase the RPM where peak power occurs, we need to open up the port and manifold dias allowing peak velocity to occur later. NA engines on stock manifolds seem to hit a ceiling once a certain intake velocity s reached, and neither RPM nor position of peak torque/power can be raised fruther. Its posible to predict where peak power will occur simply by caculating what rpm peak intake velocity will occur.

on the FI engines however, this seems nowhere as predictable. stock manifolds have been shown to be highly variable as to where peak power occurs, and appears almost unrelated to port velocities. second to this, simulations utilising wildly different port dias seems to have very little impact on the RPM position of peak power.

is this because:

a) VE is increased due to intake charge density rather than velocity?
b) FI engines still obey the common velocity limits wrt diminishing returs, but the increase in charge density contiues to increase VE despite decreasing intake flow efficiency?
c) does the increase in charge pressure and density greatly alter the reynolds number and effective boundary layer, further complicating thecalcs

d) qualitatively, can someone describe the optimised design of an FI manifold port compared to an NA manifold port

Cheers
Ed

 

[patprimmer]

a) Yes
b) Yes
c) I don't know.

Changes of temperature of the inlet charge changes the calcs considerably.

Ratio of inlet to exhaust flow become upset as the supercharged engine will produce a lot more exhaust gas than a similar NA motor.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[Warpspeed]

Not sure that the intake velocities will be that much higher. If you supercharge to twice atmospheric density, you can get twice the mass flow at the same flow velocity.

If this were not true, flow velocity would just rise quickly into sonic choke, and the huge horsepower that forced induction is known to produce just would not be possible.

Another thing to consider is the type of supercharger. A centrifugal will produce a fairly well defined boost pressure at a given impeller Rpm. But flow at that boost pressure will be highly dependent upon any up stream flow resistance.

A positive displacement screw supercharger virtually pumps a constant volume per revolution almost regardless of up stream back pressure. Any resistance to flow will just cause the boost pressure to rise high enough to overcome the restriction.

As Pat has already pointed out, the resistance to mass flow through the whole engine is not all in the induction system. Exhaust back pressure is as great, if not a greater evil.

Everything after the supercharger is just a flow restriction, but the type of supercharger will have a very great influence on both boost pressure and mass flow at any particular Rpm.

 

[edjza80]

yeah, i know that arguing the finites of intake design is a small peble in the whole engine, but its been an academic argument among friends so far, and its got us all chewing our nails.

i do agree that I expected the mach limitations on the two types of inatkes to be the same.

so the question stands - in an NA engine, peak power occurs just after valve throat port velocity hits around mach 0.6 and one can reasonably model an engines performance and RPM band by looking at the manifold and head port dimensions, and considering these numbers.

on an FI engines however, these max port velocity numbers would be expected to occur much earlier in the RPM band, with charge density / mass flow, and ultimately increased VE's occuring well past this velocity point.

so, is port velocity thus even worth considering in a model for FI engines in terms of its predictive valve (as it is in NA engines)?

and secondly, my orginal question still remains: does the boundary layer (and effective port size) change due to the increased density of the intake charge? and do we need to consider redesigning the ports in the 'mind-model' to account for this boundary layer?

cheers
ed

 

[patprimmer]

The increased pressure difference between the manifold and the cylinder will obviously increase flow speed, but it will be increasing the speed of a higher density column of air. The density of this column will obviously drop as it flows into the cylinder. This process is very dynamic with the volume of the cylinder increasing as the piston moves down, the effective CSA of the choke point changing as the valve closes and opens, the time available for pressure equalisation reducing with rpm, and the localised temperature changes as the pressures change at various points in the manifold, port, valve throat/seat area and chamber.

I guess your presumptions re mach numbers and port choking are based on average speed, not peak speeds which will be considerably higher.

re your question about boundary layer and Reynolds number, try the site in the following link.

http://naca.larc.nasa.gov/

It is the precursor to NASA and has a lot of research data done between the Wright Bros and till shortly after WW11

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[gruder]

edjza80

An interesting question:
Also hard to answer due to all the misconceptions about intake velocities regarding N/A engines. We can calculate & manipulate all the figures we want regarding airflow velocity but in "reality" it is what the engine "actually" sees in the way of flow that determines the ACTUAL intake velocity.
This is the reason we can have X flow on the flow bench which should relate to X HP etc but in reality things can & usually are quite different.

If we work the figures regarding port or runner size vs velocity we may end up with the calculated mach figures but these are surely not what the engine will see in reality.
If it was just a matter of calculations there would be no engine builder better than the rest, there is a lot more than meets the eye or the calculator or the flowbench etc.

Forced induction is more predictable due to the known induction pressure but there are still many variables as to what the engine will actualy see in regards to realtime flow.

We have configured winning intake systems with incredibly small choke points that defy all know beliefs regarding this subject.
What may seem to be a limiting point to some may be a starting point to others, we all have different views but when the flag drops the bullshit stops.

 

[headman]

My take on this is that lower mach index numbers are based on early desighn ports that take a sharp bend and fuel and air sheer @ higher speeds .an example would be 2 ports identical in volume and cross section also having the same valve size and i dentical flow numbers ,one of the ports being straight @ the valve and one being turned nearly 90 degrees .the straight port is going to fill the cylinder better @ high rpms and continue to make power .working with ports that make a major bend always have higher measured velocities @ the apex of the floor you may see 420 fps @ 28"h20 every time that i work the floor and lower this to say 380 to 390 fps the hp tends to continue to rize .this leads me to believe that with a straight port .6 mach may still make hp because the air does not have to deal with the bend and not sheering @ high velocities and leaving a larger window.