This article was first published in 2002.
The flow of air into the end of a pipe is affected a helluva lot by the shape of the edge. Air that's being drawn into a pipe that's been cut off square suffers from turbulence around the sharp edge of the pipe, reducing substantially the amount of gas that's actually able to get into the pipe. That's the theory, and a few years ago I got a company to put it to the test.
(It's important, cos the amount of air that can get into the engine is responsible for how much power it makes, and in turn that depends on how much air can jam itself into the intake runners.)
This test was performed on a 3-inch pipe, but just the same idea applies to smaller diameter tubes.
Three intake designs were tested on the flow-bench. First up was a straight piece of 3-inch tube, cut off square at the end. The actual ID of the tube was 73mm, giving a cross-sectional area of 41.8 square cm. At a pressure differential of 3 inches of water, the 215mm-long tube flowed 255 cfm.
Next up was a 3-inch pipe running a straight flare. Again the tube was 215mm long, with the length of the 11-degree flare being 75mm. The mouth opening was 115mm in diameter, giving an intake area of about 104 square cm. The airflow with this design was noticeably improved, with a measured flow of 318 cfm recorded.
This represents a 25 per cent increase over the straight pipe!
Finally, a custom-made bellmouth was trialed. Again 215mm long, the bellmouthed intake used the same flare angle as the previous pipe. This time, though, a 22mm tube was rolled around the top, being brazed to the edge of the flare. Plastic body filler was used to fill any imperfections, giving a very smooth radius entry into the tube.
And the result? 322 cfm was the measured, up over the straight flare by 1.2 per cent.
While this test showed that a rolled-edge bellmouth gave only a very small improvement over a straight flare, the use by all manufacturers of bellmouths on the intake runners of their production cars (the bellmouth is usually cast into the shape of the plenum/runner junction), and in all forms of fabricated intake systems used in professional level motorsport, shows that perhaps better results can be gained in real life than shown in the flowbench test.
Given the differences in the type of flow between a cylinder head and a flowbench, that's quite likely to be the case...
But the problem in using bellmouths on the end of intake runners has been in forming them - how do you make them, when they aren't an item that you can usually buy off the shelf?
Metal spinners can make very nice custom bellmouths - but you'll usually need to pay for them to make the former, a cost normally considerably greater than the charge for making even twelve bellmouths! I've also seen bellmouths (well, more correctly, slightly curved flares) made in exhaust shops using a hydraulic press and a mandrel, but again if you need to pay for the tooling, the cost becomes very high.
But, now we have a solution. Well, it's not a total one, but for the cost there is nothing - and we mean nothing - that comes close.
The bellmouths are appropriate if you want intake runners starting at about 47mm internal diameter - although if you want slightly smaller runners of - say - 1¾ inches (44.5mm), or slightly larger runners (eg 55mm) that still wouldn't be a problem.
Your New Bellmouth, Sir
Step One of your new supa dupa intake system fabrication is to visit a Woolworths supermarket.
Now, while there are Woolworths stores all around the world, the research for this story involved the visit to just one Australian Woolworths. So in Australia you'd be pretty confident of finding what we're looking for, but maybe in other countries' Woolworths they might not know what the hell you're talking about.
Especially when you ask for a stainless steel eggcup.
Yep, that's right - you want the stainless steel eggcup (they only sell one type). And look, if you're a bit reluctant in embracing the idea, buy just one. At AUS$1.84 it isn't going to break the bank, is it now?
When you get your eggcup home, put it on the photographic backdrop and take a pic of it - then, it should look just like this.
Tilting it to get a better view while your photographic assistant holds a reflector should make it look like it does here. As you can see, a not very attractive eggcup - unless you have a house full of stainless-steel-and-glass furniture, anyway. Or toddlers, I guess. Throw this eggcup onto the floor in a tantrum and all it will do is bounce. Not the egg though...
Now turn it over and what do you see? Well, I see what looks to be an abstract of the intake of a jet engine, but maybe you'll see something quite different. Anyway, perhaps we will both agree that the unpolished side isn't quite as attractive as the other surface. You agree? Well, we're now reduced to discussing metal finishes on eggcups...
But if you turn it back over again, and apply a file through the upper edge, something will happen. Because stainless steel is quite tough, using a hand file will take you a while - better to apply a power sanding belt or similar tool. We did it with a file.
When you grind through (just
through, don't mangle it), you'll find that it looks like this. Bend the dislodged piece back and forth until you can take it out, file the sharp inner edge smooth with a half-round or round file, and you'll find yourself left with....
... this superb bellmouth! And remember, it cost you a total of AUS$1.84 plus some sweat.
Look at what you now have. A hand-formed stainless steel bellmouth with an opening diameter of 60mm and a lip radius of about 6mm, complete with a gentle taper heading down into the runner.
And how do you join the bellmouth to the intake runners? That could be done in a number of ways - sweetest would be to TIG it to a length of thin-gauge tube. But that will also be the most expensive approach. A high-temp epoxy adhesive could also be used, or you could flare the intake runner tube to take the bellmouth as a push-in element (the 'tube' section of the bellmouth has a near-constant taper). Alternatively, and especially if you just want to test variations in performance on say a multi-throttle intake system, you could use rubber hose and clamps to trial them into place.
Whatever attachment approach you take, you'll need to smooth the inner flow join transition with an appropriate tool - eg a round file.
Well, we're sure that Woolworths tested the product in great depth - as an eggcup. How well it flows as a bellmouth on an engine we don't exactly know. But we did do some flow testing, using a manometer and a constant airflow source. Compared with a bare tube end, there was a distinct and measurable drop in flow restriction with the bellmouth in place. .
No surprise there - with a shape like this, you'd' certainly expect it to be a damn sight better than a lot of the sharply-cut-off-tube fabricated intake manifolds that we often see.
In fact, even if you're only vaguely thinking that in the future you might need to fabricate an intake manifold, it might be worth buying 2, or 3, or 4, or 5, or 6, or 8, or 12, or even 16 of them...
The tuning of intake systems is a black art, and one that is far better carried out by doing back-to-back engine testing than sitting around theorising. At least three variations in the design of the bellmouth section of an intake runner will affect performance:
- the diameter of the opening
- the radius of curvature of the lip
- the taper of the runner immediately after the bellmouth
In addition, the length of the runner (measured from the lip of the bellmouth to the intake valves) is critical, as is the shape and size of a box (ie plenum chamber) placed over the intakes, and even the shape and length of the tube that feeds air to that box.
One of the most interesting back-to-back tests we've seen done was conducted on a Ford Escort 2-litre that was running a quad throttle body system. The full story is at "Escorting Power". On that very highly tuned engine, changing runner length gave power variations of up to 12 per cent! And, while we didn't specifically cover it in that story, changing the bellmouth shape also altered power.
Another story that we have done that touches upon the critical aspects of bellmouths is our interview with race engine builder, John Sidney. In it he describes trialing on the engine dyno bellmouth trumpets with different tapers, lengths and lip radii. In answer to a question about whether injectors placed across the mouth of the bellmouths obstruct airflow, he said:
"When you get a big oil leak on the [engine] dyno - and there's smoke blowing all around - you see the air drift up over the engine and then it gets to a point near to the edge of the bellmouth and it gets sucked around the rolled edge. The air comes from here [gesturing to the rolled edges of the bellmouths]."
The shape of that edge is critical! See the full interview at "AutoSpeed Interview: John Sidney of John Sidney Racing".