goldenchild---
Yes, but I have changed my mind about what seems to be the optimum configuration several times since my last post!
It's a strange story---
I was looking at the 2" pipe with the 3" reducer going into the 3" "Y" fitting, and trying more things, with it attached to the shop vac today. Mainly trying to figure out where the suction comes from, and how the larger size "y" fitting actually works to enhance it.
I came up with three things.
First, if I inserted a 2" PVC pipe into the suction port of the 3" "Y" connector, the suction up the 2" pipe increased when I pushed it up into the air stream within the 3" fitting. I thought it seemed like it would do that, but wasn't sure if having the 2" pipe up in there wouldn't bog down the air flow, and just prevent any increased suction. But it worked.
Secondly, I got to wondering how that would affect more suction ports in a multi-port system. Maybe having that suction pipe sticking into the air stream would mess up the downstream suction ports?
Third, I was wondering how the additional volume of air coming up through the suction port would affect the air density/pressure in the main blow-flow line. That, as well as turbulance from the first suction pipe being right in the air stream, might just counter any additional suction possible from additional suction ports.
So to get back to your question, I was then thinking, with the test setup I'm working with, the best configuration would be the 2" ID PVC pipe for air entry from the blower, going into a 2" to 3" reducer which goes into a 3" "Y" fitting. Then for the suction, a 2" PVC pipe going into a reducer to the suction port of the 3" "Y", but with the stop flange inside that reducer ground off to allow the 2" pipe to pass through and penetrate the air flow area.
Penetrating the air flow puts the air stream all around the suction pipe opening, making more "cavitation" area, and creating more suction. Cutting the suction pipe, on the end which goes inside the air stream, at a 45 degree angle, would result in, when that angle is combined with the 45 degree angle of the "Y" fitting's suction port, making the 2" suction pipe end-cut to be at a 90 degree angle to the blower air flow, creating even more opportunity for "cavitation" suction. But the bottom of the 2" suction pipe air stream opening is blocked from the blower air flow, by the 2" pipe, itself. So adding a collar to it, to extend straight down the center of the air stream far enough to go past the turbulance caused by the entrance of the 2" pipe, would allow the cavitation effect to exist completely around the 2" pipe tip in there. This would require a 22 1/2 degree cut on both the 2"suction pipe and it's collar.
At this point, I was thinking that a single-stage, rather than a multi-port, configuration would probably be more efficient, because of eliminating the efficiency losses at each port, and turbulence, and so forth.
But---the major problem with all this is that the thing
now resemples the drawing from Ammen's book, which jimdoc posted on the first page of this thread!
So, it's gone full circle! Oh no's!
Which figures, because why would he put it in his book, unless it was the optimum configuration. They must have already worked this all out.
So, why didn't Ammen's setup work for glondor? It seems to me it's probably because he used a 3" going into a 4". the cross sectional area of a 3" ID pipe is 7.07 sq. in., and the area of a 4" is 12.56 sq. in. That's less than double the area of the 3". Is it that it must be at least twice the cross area? Or did he have too many bend fittings, which would reduce the the flow as much as a smaller cross area? Or both?
Another interesting point about Ammen's configuration is that it is just the opposite of the one which I was setting up and testing, above. In Ammen's, the blower air comes in via the smaller pipe, and goes down the larger pipe. In the setup I was testing, the blower air feeds the larger diameter pipe fitting, and the suction is drawn out of the smaller pipe which is inserted into the larger diameter fitting. Does this make a difference? Does Ammen's design create cavitation around the
outside of the smaller blower air feed pipe, instead of creating it on the
inside of the smaller suction port pipe like the arrangement that I was testing? Apparently it does, because supposedly it works.
I don't have a cad program, and the only scanner is connected to my wife's computer, and it's late, so if you still want a diagram of the "Y" fitting configuration, I'll have to do it tomorrow.
As for me, I need to pick up another reducer, so I can test the suction with it stuck into the air stream, and with the 45 degree angle on the air stream end, and also with the extension collar.
It might turn out to be easier and less expensive in fittings costs, to just use a 3" 90 degree fitting, and drill a hole in the corner of it, and pass the 2" suction pipe into that and far enough up it and into the following ducting, to get past any turbulance from the 90. And the 2" air feed pipe from the blower would go through a reducer into the 90, on the other end from the one that the 2" suction port goes down. And that would make it look even more like Ammen's drawing in reverse.
Which brings up the question of, if the same 90 were used, but in reverse, with the blower fed through the 2" pipe going in through the corner of the 90, would the end of the 90 without the blower pipe going up it, then become the suction port, as in Ammen's diagram? I guess it would. And is one of these configurations more efficient than the other? I'm guessing they will be the same. Except that if you blow through the smaller pipe, into the 90, and suck through the other end of the 90,
you then eliminate two reducer fittings, and their cost.
Whichever way the venturi design goes, I think it's going to be very important to remember to compensate for whatever bends are in the exhaust ducting system, probably by going to a larger size than the venturi fittings section, which in my test case would be going to 4" exhaust ducting out of the 3" venturi casing.
If a system starts with 3", 4", or larger, out of the blower, the cross sectional area calculations are going to be very important, also ( 2 pi r squared).
I think that the maximum suction flow for a venturi system is going to turn out to be around half the CFM of the blower air out of it's feed pipe to the venturi unit. But I think the whole thing, with twice the size of blower, and the larger ducting required, would still be less expensive than an acid proof blower, unless you have a really long ducting run.