Intake "Tee" Junction - Any Suggestions?

[sabre2fv]

I'm working with a couple of different older aircooled VW engines that use the stock intake manifold that has a vertical downtube from the carb that then transitions to a horizontal tube that feeds both sides of the engine. The Tee junction really seems to cut down on the flow, roughly 16% of what the carb can flow (~90 CFM down to 75 CFM). Because of the manifold configuration, porting opportunities are somewhat limited. It seems to me that this tee junction is very critical to improving performance/flow. Other than smoothing the short radius (which does help), does anyone have any suggestions that might help? Rules limit what I can do, but it seems that a basic understanding of what happens in the bend is a good start. I've tried squaring the sides of the upside of the short radius and even some dimples, all with either no effect or a decrease in flow.

I have a basic flow bench system, and basic knowledge (I think!) and am searching for ideas.

Thanks.

-Jim

[thomasvaught-1]

The idea behind the "tee" in a typical "air only" runner
is that you will have even flow to both sides. Trouble is when you add fuel and the mixture hits the bottom of the "tee" the fuel drops out of suspension.

Best way to reduce the mixture air/fuel distribution issues is to run a dual bore carb with a small balance slot to smooth out the idle modes. Carb has to be big enough to feed one runner by itself: (Typically from 2 to 4 times the normal carb flow).

Very difficult to evenly divide the fuel and air mix when the runners are close to each other as the first runner in the firing order will rob from the other runner.

Always interested in ideas on this subject.

Tom Vaught

[sabre2fv]

Tom,

I have long suspected that fuel drops out of suspension with this arrangement. I have received one PM suggesting I add a 'divider' directly below the downtube, and indeed one of manifolds I'm working on has this feature cast into it - but I want more! I also likewise expected that there would be 'charge robbing' between cylinders on opposite sides of the engine as the flow has to reverse itself every other firing stroke.

Unfortunately the rules only allow the use of a single carb so I have to optimize the tee junction as best as I can. As you state, there is difficulty in balancing the flow side to side, which I do see in the flow data. Adding to the fun, the horizontal runner lengths are different which changes the path length between the carb and each port - I don't know how important this is, but suspect it is a factor at some level; I'm stuck with it however...

Thank you for your comments.

-Jim

[thomasvaught-1]

If the rules allow: put a big "tall" plenum on it with shorter length runners that come in from the sides like a 4 bbl intake and you will be fine.

"Tall" keeps the fuel off the floor.

"Tall" lets the air/fuel turn into the runners.

Large plenum volume acts as a "storage tank" as well as a
"pulse damper" on the carb circuits for more even consistant
calibration control. Package is the only issue to work out.

JMO

Tom V.

[sabre2fv]

Tom,

HMMMMMM! I haven't thought of the plenum idea. I have to work within the confines of the existing manifold, but I certainly can open up the area below the carb and above the turns to emulate a plenum as much as possible. Do you know from experience if this will show a benefit on the flowbench or is it something that has to be tried on the dyno? I will be dyno testing, and will have different pieces to try. This might be another flavor...

Thanks!

-Jim

[106-1194218389]

I used to race VW's also. I had a machine shop and specialized in the Bug. Can you modify the stock intake at all, but just have to keep the stock carb? If so, put a larger down tube below the carb and flare the bottom where it hooks to the horzontal tube so you can get roundness.

John

[thomasvaught-1]

Agree! A larger "down tube" in diameter slows the velocity.
A Larger Diameter horizontal "Tee" basically increases the distance before a fuel droplet hits the floor.
A Flare at the roof of the horizontal cross tube (as was said) allows the fuel and air to "follow the curve" into the horizontal runners.

Would be really nice if we knew the exact rules then we could figure out how to get around them! LOL.

Tom V.

[68Corvette]

I have had luck with old harley davidson panhead engines by welding divider in T turn.
Users have reported about smaller jetting and less ignition advance needed and that bikes pull harder.

I as much material to outside curves that I get it ported to D-shape and as long turn as possible (also because of the divider is taking space)


Then lathe in thick disc



Nice long curves on both sides of divider, D-shaped inner bends wont show here, as tube tapers towards divider and then expands again when D-shape begins.



Finihed outside after lathing sealing surfaces.
inner curve is finished after lathing to give largest possible radius in "shortside"


My thinking was when making these to block pulses going between cylinders and try to bend A/F mixture better, rather than allowing it to turn 90-deg turn and separate badly plus hitting to wall.
it might have worked as jetting and timing went down.

Only "downside" was that idling was somewhat smoother
As these are made for street driven HDs I cant really say anything about top RPM performance.

[106-1194218389]

Good pictures 68Corvette! That is exactly what I was thinking and I feel Tom was also pointing out. Thanks for illustrating. On top of that you have "proved" the theroy which is very important. Did you shape the divider then after you put it in the middle? I could not quite tell. But your results prove it works.

John

[68Corvette]

Yes, it has also as large radius as possible in divider tip.

I had to grind it a little down to get a bit more CSA at the turn.
So its not as big as in welded separator2 picture.. if you compare it to finished separator2.

unfortunately I dont have CSA information, but if memory serves me it was somewhat larger than pipe CSA.. maby ~110-120%.

In one engine I also welded intake fully closed and regrind it totally differently.
That worked also really good, as HD runners are pretty goofy to start with.

[sabre2fv]

68Corvette, Tom and jfholm,

I have something quite similar except that the horizontal tubes are epoxied in and there is a bend in the down tube; the epoxy gets real soft when it sits on my gas BBQ for about 45 minutes...

I agree, very nice pics - look very similar to what I have.

I already have a nice radius where you filled yours in. I did find that increasing the short turn radius seems to increase the flow though the numbers aren't big, maybe a few CFM. The radius work also included changing the shape to more of a "D" shape, until I ground through the side of the casting. Ovalizing the passage at the bend might have helped a bit too; maybe .5 CFM. Per Thomas' and jfholm suggestions, opening up the area directly below the carb did seem to also help the flow, perhaps another .5 CFM. But I need to re-test, and possibly open it up some more. There is a divider in the bottom that is cast in - looks similar to what 68Corvette added. An interesting finding here, is that when I ground away the epoxy that I put on this divider (Long Turn Radius), the flow dropped about .25 CFM!

Tom,

The rule is that the manifold can be "modified". The term modified is defined as "To change a component from stock". Big blatant changes would be frowned upon, possibly leading to a rules 'clarification'. However, finding out what works and what doesn't, does lead to a better understanding of the flow and THAT is what I'm after; after that, I will try to gage how far to push the modifications. It does appear that the rule makers want a 'stock appearing' manifold.

Thanks for your suggestions - I'm learning

-Jim

[larrycavan]

[49-1183904562]

Larry;

3:30 Am Pennsylvania Time? Do we need to talk to Bruce about and intervention or do I even utter the words Flow bench Anonymous?

BTW That was Smoother idle for a Harley! It's a subtle thing on a ridged mount....

[68Corvette]

Smoother idle was only complain I got from other person.
Though he said that he could get some of it back by reducing idle RPMs.
I guess that could have something to do with HDs firing order.. maby 1st cylinder did not cause so much "robbery".