Well, I stumbled onto a thread recently and was intrigue with the velocity stack. In particular the application on a tubular intake and accomodates a filter. After dwelling on the idea for a few days, I got more and more excited about the idea. Background info: I am a mechanical engineer in the oil and gas industry. During school I "specialized" in fluid dynamics, meaning I simply took alot of advanced courses in the subject. Anyway, my currently employer has Cosmos FloWorks software for fluid flow analysis. This is not extremely highend analysis software, but still powerful and the liscence is enough to buy a new car every year.

My analysis is not complete at this point, but I started by modeling a typical cone filter on a 3in pipe. Boundary conditions include a pressure differential between the "filter element" and the outlet of the tube of 1-bar. I basically was looking for pipe outlet fluid velocities and mass air flow rates. I assumed ~100F air temps. I next modeled a "stack" similar to the Prototype Stack that is floating around here. The main geometry that I alter for my analysis was the radius (Rl) of the inlet lip and the radius (Rv) of the "funnel" which tapers into the 3in pipe ID. So effectively Rv changes the effective length on the funnel/taper. My first go round was to look at the following combinations:

Small Rv with Small Rl
Big Rv with Small Rl
Small Rv with Big Rl
Big Rv with Big Rl

I will post the results in a minute. Final analysis is nearly complete.

 

I am a bit confused what Rv is as compared to Rl...

If Rv is the radius at which the funnel curves inward to meet up with the 3" intake tube, what then is Rl? You say it is the radius of the inlet tip... Could you explain this?

But to your research and analysis

 

I'm interested to see your results. What software are you using to do this analysis?

Note:

I was doing some research on Velocity Stacks recently and came across this. This will be very benefical for your test.



You will notice that there isn't a constant radius on the inlet lip. Its actually an ellipse. Therefore, there is no cone present.

 

Note that Rl and Rv are tangent at their intersection point.


Modified by BryanPendleton at 3:44 PM 12/1/2005

 

Well, at first glance the results would lead you to believe it is pretty insensitive to the geometry of the "velocity stack", with the exception that LONG is not good, though only by ~5% in mass flow rate. Compared to a typical cone filter on a pipe though, the velocity stack represents a significant improvement with over a 40% improvement in mass flow rate. I will be doing some further analysis into geometry optimization. I'll post results as they become available.

The nice thing is once I have a design I am happy with, I will "plot" it on our rapid prototype machine, so I'll have a functional ABS model to test. Stay tuned, and if anyone has any information, feedback, tips, etc. by all means lets hear it. This research if for myself and others.

 

I will definately try that .667D to D ratio for the eliptical entrance. Not sure where ratio came from, but "looks good".

 

Yesterday on the flow bench we took the clay ring around the port that works like a velocity stack off and flowed the port "bare". It went from 235 to 220 cfm @28" with no other changes.

 

Interesting, I would like to see the lengths a bit more similar. I know that is hard as you are working with pieces you have.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by drdisco69 &raquo;</TD></TR><TR><TD CLASS="quote">Yesterday on the flow bench we took the clay ring around the port that works like a velocity stack off and flowed the port "bare". It went from 235 to 220 cfm @28" with no other changes.</TD></TR></TABLE>

damn!

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by DonF &raquo;</TD></TR><TR><TD CLASS="quote">Interesting, I would like to see the lengths a bit more similar. I know that is hard as you are working with pieces you have.</TD></TR></TABLE>

What would this show you since the two that seem close to similar produce somewhat similar results (the two on the top)? &lt;--Just trying to learn here not questioning your methods.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by drdisco69 &raquo;</TD></TR><TR><TD CLASS="quote">Yesterday on the flow bench we took the clay ring around the port that works like a velocity stack off and flowed the port "bare". It went from 235 to 220 cfm @28" with no other changes.</TD></TR></TABLE>

i know what your mean. i flowed a head without the radiused inlet guide and dropped like 30 cfm the other day.

and thanks for the great information on the velocity stacks

 

PROPS TO YOU for doing some very interesting research!
(waiting for the results)

 

I dynoed a car the other day with two style velocity stacks-back to back. One was the Prototype Racing intake that I have in my posession. The other was the customer's generic Ebay stack; the inlet horn was wider and flatter. The Prototype intake killed the generic intake everywhere, not just in one part of the powerband-a bit of a surprise to me. It would be interesting to test different air horn shapes and their effect on power and air flow.

 

Interesting. Maybe thats why they charge so much for their velocity stack LOL

 

how about a velocity stack design like this one http://www.turbohoses.com/velocity_stack.htm

 

glad i went with the prototype

 

anybody know the specs on that prototype racing velocity stack. AKA what radiuses were used for that and over what lengths?

 

any opinions on the hks mega flow? i think it was one of the first aftermarket intakes to incorporate a velocity stack

 

I have further analyzed the pressure differential graphs and velocity contour maps and I retract my early statement about length being BAD. I am incoclusive with regard to length at this point. The reason config 4 does not perform as well as the other config is the "effective entrance angle". The other configs radius from the ID to the front plan gradually and smoothly. This allows the pressure gradient to "mushroom" out further, thus drawing in more air. IMO, The Prototype Stack is going is a better design that the TurboHose Stack. The TurboHose Stack is similar to my Config 4 for the "funnel" does not transition tangent to the front plane smoothly.

Now, if allowed (as in Config 2 & 3, and the Prototype Stack) the pressure gradient will actually curve around the backside of the lip some. IMO, the full radius though is not necessary, as the gradient only effectively reaches around maybe another 30deg, not the full 90 as in config 2, 3 and the Prototype Stack.

I will post some photo of pressure gradient maps to clarify the "mushrooming" effect.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by b19coupe &raquo;</TD></TR><TR><TD CLASS="quote">I dynoed a car the other day with two style velocity stacks-back to back. One was the Prototype Racing intake that I have in my posession. The other was the customer's generic Ebay stack; the inlet horn was wider and flatter. The Prototype intake killed the generic intake everywhere, not just in one part of the powerband-a bit of a surprise to me. It would be interesting to test different air horn shapes and their effect on power and air flow.
</TD></TR></TABLE>

Please post a photo of the Ebay Stack.

 

Similar design to the Prototype Stack where radius from the OD all the way to the ID. Notice how the pressure gradient mushrooms out more, effectively "feeding" from a larger volume of air as compared to the Config 4 pressure map shown second.

Constant radius inlet stack (similar to Prototype Stack)


Config 4 Stack

 

 

B19coupe, when you tested the two velocity stacks did you simply add that "ebay stack" to the block coupling shown on the intake with the Prototype Stack. Reason I ask is the additional length on the Prototype Stack would negate your results with regard to comparing the velocity stacks only.

 

Intake length was the same. The intake pipe shown is actually longer than what was used on the car for the test. Customer's intake pipe was used with his stack, short elbow was used with the Prototype intake.

 

Since there is someone in here that actually has a brain about flow, could you give me a clue as to the positives and negatives of the j's racing style intake tub? From what it seems, it really doesn't have too much of an effect over a straight filter on a stick, but i'd like to at least see the science of it.

 

I suspect that your question was aimed at BP, but will throw out my 2 cents. The J's is great below 4,000 in my experience but suffers at high rpm. From my rudimentary understanding of flow, I surmise that there is too much turbulence in the J's at higher rpm for optimal flow.