maf and throttle body diameter

[blue2000s]

It lowers losses to have larger tubes up to the throttle body. The pressure losses due to friction in a tube are directly related to the speed of the flow running through the tube.

Set the throttle body at something, 76mm, then run a 76mm tube 3 feet up to the throttle body and pull a vaccum. Measure the speed of the air at the inlet of the tube. Now run exactly the same experiment but start with a 100mm tube that gradually drops to 76mm at the throttle body. The speed of the air at the inlet is lower. If you look at the flow rate through each of these for a given vaccum level, the decreasing tube will have a higher flow rate. It's analogous to running a set amount of electrical current through a fat wire or a thin wire at a specific voltage. The thin wire heats up more due to resistance.

I think I showed the attached pictures before of flow entering a "T". When the center stream of the flow leaves the entering section and hits the back wall of the T, it stops or stagnates. The flow along the sides of the entering tube is pressed out by the stagnated flow from the center of the tube and makes it's way out towards the outlet of the T. In other words, the flow makes it's own wedge. Adding a wedge, without many, many experiments or fluid modeling, is more likely to hurt than to help with the airflow.

The more interesting thing about the picture is the big "dead" zones as the flow near the walls leaves the entering tube. This is known as separation. The tube in the T section may as well be half as large because the air isn't doing anything in the corners. If the entering tube where rounded as it merges with the T, the air could more easily follow the contours of the transition and use more of the volume of the tube. There is a potential for a HUGE benefit to airflow.

[jaykay]

Originally Posted by blue2000s

It lowers losses to have larger tubes up to the throttle body. The pressure losses due to friction in a tube are directly related to the speed of the flow running through the tube.

Set the throttle body at something, 76mm, then run a 76mm tube 3 feet up to the throttle body and pull a vaccum. Measure the speed of the air at the inlet of the tube. Now run exactly the same experiment but start with a 100mm tube that gradually drops to 76mm at the throttle body. The speed of the air at the inlet is lower. If you look at the flow rate through each of these for a given vaccum level, the decreasing tube will have a higher flow rate. It's analogous to running a set amount of electrical current through a fat wire or a thin wire at a specific voltage. The thin wire heats up more due to resistance.

I think I showed the attached pictures before of flow entering a "T". When the center stream of the flow leaves the entering section and hits the back wall of the T, it stops or stagnates. The flow along the sides of the entering tube is pressed out by the stagnated flow from the center of the tube and makes it's way out towards the outlet of the T. In other words, the flow makes it's own wedge. Adding a wedge, without many, many experiments or fluid modeling, is more likely to hurt than to help with the airflow.

The more interesting thing about the picture is the big "dead" zones as the flow near the walls leaves the entering tube. This is known as separation. The tube in the T section may as well be half as large because the air isn't doing anything in the corners. If the entering tube where rounded as it merges with the T, the air could more easily follow the contours of the transition and use more of the volume of the tube. There is a potential for a HUGE benefit to airflow.

Are you using the CFX package in Ansys to model the flow?

Looks to me like the optimum flow regions are scribing a 'Y' pattern! The diverter being it the centre bottom stagnant zone. It looks to me that this shape would reduces losses but.... Don't forget that we are drawing air in and there are varying pulses. I would be concerned with any tendancy for the banks of cylinders drawing from side to side over the diverter. This may not happen to any large degree but in this case I would think a tee would be better.

I went and look at the ipd and low a behold they have relief hole put through the diverter. They look just big enough for pressure balancing side to side.

Is the Y pipe model next???

[2003S]

Originally Posted by blue2000s

I think I showed the attached pictures before of flow entering a "T".

I'd be very interested in seeing a simulation like this for Pedro's techno insert in the T junction... Any chance you can whip one of those up?

[The Radium King]

blue - i get that it lowers losses to have larger pipes up to the constriction; my question is that, if the throttle body is the constriction, why did porsche put it in the system at all?

regardless, the ipd dynos show that a larger throttle body with a better-designed plenum is a winner on the 986 (the 987/cayman have a resonance flap on the oem plenum that is removed in the ipd product, hense i think the poor low rpm performance on the ipd kit for these cars).

the idea is to make the ipd product without the ipd $. as has been pointed out, the pedro product is flawed in that it doesn't address the turbulent flow that occurs on the inside radius of the 'T'.

the best solution i see is to fab a 'Y' pipe out of silicone that mates to the flange/aos end of a cayman plenum and throttle body.

[jacabean]

what kind of realistic gains can be made by this mod ? if it is just a few h.p. like 5 or less i do not think it is worth the cost. if you can get about 10 to 15 h.p. then i think it would a good value. does anyone have any solid numbers for gains on a 3.2 engine ?

[blue2000s]

I'm using Fluent in these simulations.

A Y will be more efficient than a T for airflow in a steady state, that's correct.

Keep in mind that with three cylinders to a plenum, there is a pretty much constant flow of air into the plenum when the engine is kept at a constant speed. There is always a cylinder pulling air. The pulses that influence secondary pressure waves are important, but they are secondary in nature so the influence on flow is on a much lower order than the primary vacuum pulled by the cylinders themselves.

Here are three more simulations varying on the one I posted earlier. The first one just adds a sloped wedge to the original T. The second adds a large radius to the T junction and the last one combines the radius with the wedge.





The volume flow rate results show that the T and T with wedge have almost identical flow rates. The radius-ed T with and without the wedge are both almost the same flow rates but are about 30% greater than the standard T.

[jaykay]

Originally Posted by blue2000s

I'm using Fluent in these simulations.

A Y will be more efficient than a T for airflow in a steady state, that's correct.

Keep in mind that with three cylinders to a plenum, there is a pretty much constant flow of air into the plenum when the engine is kept at a constant speed. There is always a cylinder pulling air. The pulses that influence secondary pressure waves are important, but they are secondary in nature so the influence on flow is on a much lower order than the primary vacuum pulled by the cylinders themselves.

Here are three more simulations varying on the one I posted earlier. The first one just adds a sloped wedge to the original T. The second adds a large radius to the T junction and the last one combines the radius with the wedge.





The volume flow rate results show that the T and T with wedge have almost identical flow rates. The radius-ed T with and without the wedge are both almost the same flow rates but are about 30% greater than the standard T.

Well done; well put. I was just considering any tendency to pull from bank to bank when the firing order is such that there is a strong intake pull from one side versus the other.

Looks like in theory the 'y' has some definite benefits...now comes the hard part part of testing varying shapes and geometries.

[Johnny Danger]

Am I correct in saying, rather than spending a $1000 + on the IPD set-up, it sounds like using a 74mm 996 throttle body, along with a 996 "T" that has been modified by Pedro, is the best and most cost effective way to improve the oem design ?