looong intake runners for HP

worntorn said:

Although long bell mouths may add to the racy look of a machine, they do not, in my experience , add any performance where concentric carburetors are used.

Click to expand...

I don't wish to disparage Paul Dunstall's skills, but that's not what I found. I gained 4 horsepower simply by fitting longer bellmouths. I didn't even need to change the jetting.

Because of standing wave considerations, I suggest that the length of the inlet tract and the length of the exhaust only really suit one specific rev level. Everywhere else they are a compromise. If you find something that works well, you are lucky. The comment about the 4BHP boost to the BMW by fitting longer inlets is interesting. I use long bell-mouths on my Seeley because they came with the carbs and look good. Perhaps dyno testing does have some value ? I work on the basis that from experience, I know when my bike is going quick enough and when it is not. The big blast down the straights is not the only way to win races. A longer inlet tract might make the bike more controllable (smooth) when it is in and around tight corners. I doubt that comparing motors in dirt bikes and in road racers gives much information. When you are riding them the power characteristics you need are different.

acotrel said:

Because of standing wave considerations, I suggest that the length of the inlet tract and the length of the exhaust only really suit one specific rev level. .

Click to expand...

Sort of. It works best at one region of the rev range but also at various other harmonics.

I used this simple online calculator. You tell it the intake length and it tells you the RPM range that it will affect. Dead easy and it worked out just as they said.



The lower line is what I started with. Jetting moved it up to the middle line, and the longer bellmouths are the top line. As you can see the power has been boosted between aprox 5500 and 6700 and from 7200 to over 8200.

Good post, John. Nice to see real data.

Another bit of info supporting longer intake lengths is that the factory John Player racers, including conventional frame, monocoque, and bird cage bikes, used them, as did the factory flat track racers. Mick O'field's notes from the factory race shop show a total length from intake port face to end of bell mouth of 9.5" for 34 mm cabs and 10.0" for 36 mm carbs.

They did plenty of dyno and on-track testing, so I'm pretty sure they had good reasons to go long.

Ken

I was surprised to read the Dunstall comment about long bell mouths. Dunstall bikes won a lot of races and set some speed records as well. He definitely built some very fast Nortons.
His 1968 Dommie Roadster was the first sub 12 second 1/4 mile machine ever tested by Cycle World and was considered the world's fastest production bike at that time. He also set a world 1 hour speed record with the same type of bike in 67.

lcrken said:

Good post, John. Nice to see real data.

Another bit of info supporting longer intake lengths is that the factory John Player racers, including conventional frame, monocoque, and bird cage bikes, used them, as did the factory flat track racers. Mick O'field's notes from the factory race shop show a total length from intake port face to end of bell mouth of 9.5" for 34 mm cabs and 10.0" for 36 mm carbs.

They did plenty of dyno and on-track testing, so I'm pretty sure they had good reasons to go long.

Ken

Click to expand...

Interesting notes Ken. I seem to recall reading somewhere a few years ago about ports and intake tracts where intake tract volume for a given engine and desired performance characteristic was a bit of a unifying theme. Naturally, you want to get the velocity and flow efficiency without hitting too much of a sonic choke wall. Your two intake tract lengths with different size carbs prompted me to bring this up as it seems a bit contrary but makes me wonder if we can look at it in terms of energy stored within a port (ex. slightly longer runner (higher air velocity) to make up for the slightly slower air velocity in the larger carb cross section).

Probably too much coffee this morning.

Yes, there is energy stored in the mass of the moving air. It all comes down to the weight of the air that is delivered into the cylinders and that is helped by the inertia of that moving column of air; it aids in cylinder filling

One thing with my experience with the BMW: A friend ( well, friend on the internet ) tried the same thing and found no difference when he added the long bellmouths but he was using a completely different cam.
It could be that the cam makes a difference to the length required.

pommie john said:

One thing with my experience with the BMW: A friend ( well, friend on the internet ) tried the same thing and found no difference when he added the long bellmouths but he was using a completely different cam.
It could be that the cam makes a difference to the length required.

Click to expand...

I think it makes a HUGE difference John. Enough in fact to dictate the outcome!

Good point John. There are so many variables, change one thing and everything changes.
Also, Dunstall's comments are only in reference to the use of twin 32 mm Concentrics with his hot Street 750s or twin 30 mm Concentrics for hot Street 650s.
Not very large or exotic gear for the speeds and performance obtained!

Glen

Surely if the inlet ports operate on the basis of standing waves, longer inlet tracts should cause the motor to produce more torque at lower revs ? Same with exhaust pipes - the longer they are, the lower in the power-band the power comes on strong, if you don't change the cam ? I suggest the formula which applies is : wavelength equals the speed of the energy pulse in the medium divided by the frequency. So the higher the frequency, the shorter the wavelength, i.e. higher revs = shorter inlet.
I think the bit about bigger carbs working better with the longer inlets might be correct. Carburation is a very fickle jade. I use 34mm Mk2 Amals with long bell-mouths and standard size 30mm ports, but tapered from 34mm over the first 25 mm in the head of my 850. It all works extremely well. However I use methanol fuel which tends to hide-up tuning errors.

Throttle position matters too. That is, when you add a velocity stack, the throttle plate cuts the tube into two pieces, and each has a different resonance. So, simplified, there are two sets of waves - short ones and longer ones. Now, less simplified, they are muffled, chopped up, attenuated in signal because the opening and closing, of the throttle, by throttle position. changes the area of the reflector, as a percentage of the area of the tube, which is presented to that tube.

Also, the benefit (more volume/mass) of a long pre-carb velocity stack is attenuated as the throttle is lowered from wide open. When the throttle is from half way to wide open, that is when the additional volume/mass of the air added by that bell mouth really comes into play. At lower throttle positions, the more important length is between the valve and the throttle. At less than half throttle, more length before the throttle plate presents more inertia, more vacuum to pull fuel, to pre-load/richen the intake tract with fuel just before you re-open that throttle - bingo, better throttle response. So, when the throttle is partly closed, this may partly explain why "long intakes are better."

I never really believed in the inertia effect - always only about standing waves. However the extra lead on the Dynatek ignition system used on some Harleys makes sense. It is attached to a vacuum switch in the inlet port. If the motor loses vacuum at large throttle openings, the switch operates to change the ignition system onto a less aggressive advance curve. Also the shape of the taper on the needles in Amal Mk2 and Mikuni VM carbs is probably designed to compensate for loss of vacuum.
I know a lot of guys use flow benches when enlarging and reshaping inlet ports, however I suggest that if the inlet system is only about standing waves, using the flow bench might be erroneous. It might be the situation where the revs and the cam make a difference.

The flow bench is still very useful in improving port flow. In general, finding and improving on areas that restrict smooth static flow (valve guides, short side contours, valve and seat shapes, etc.) will also improve dynamic flow (pulsed, resonant, standing waves, sonic waves, whatever you want to call it). It was established decades ago by serious research and test programs at the major auto manufacturers, that measurements of static flow really are fairly good indicators of performance, as long as the port design also keeps the flow velocity high enough. The mistake a lot of tuners made back in the early days was to just keep hogging out the ports to increase the flow rate numbers, with no regard for flow velocities. I don't think any of the established modern tuners make that mistake. The trend for some years now has been smaller ports to keep velocity up, but with good port shapes to keep flow rate up.

I think we don't hear so much about flow bench use now because modern engines have much better port design to start with. On a lot of modern bike engines, the best you can do is just clean up any irregularities in the ports, and maybe fit larger valves. You don't often see the kind of radical reworking of ports that was common on engines like the early Harleys and American V-8s.

Ken

A while back father and son friends of mine had a highly developed pushrod Matchless engine in a featherbed frame, which they used for historic racing. I looked down the carb and the inlet port was huge. The rate of gas flow must have almost been stationary. I said to them 'why don't you fit a two valve Jawa engine to this, you would start where you have finished with the Matchless engine'. They actually did that and first time out with an almost unmodified engine, beat the top guy who had a new Molnar Manx. The problem is the father is never happy, so he is now working over the Jawa engine. I seen inside a few of those engines and there is very little opportunity for improvement. I suggest the mindset that 'big is always better' is pretty common amongst many tuners. Once you have enlarged the inlet ports in your Commando head, when the motor won't PULL, there is really no way back which is not extremely expensive.

lcrken said:

The flow bench is still very useful in improving port flow. In general, finding and improving on areas that restrict smooth static flow (valve guides, short side contours, valve and seat shapes, etc.) will also improve dynamic flow (pulsed, resonant, standing waves, sonic waves, whatever you want to call it). It was established decades ago by serious research and test programs at the major auto manufacturers, that measurements of static flow really are fairly good indicators of performance, as long as the port design also keeps the flow velocity high enough. The mistake a lot of tuners made back in the early days was to just keep hogging out the ports to increase the flow rate numbers, with no regard for flow velocities. I don't think any of the established modern tuners make that mistake. The trend for some years now has been smaller ports to keep velocity up, but with good port shapes to keep flow rate up.

I think we don't hear so much about flow bench use now because modern engines have much better port design to start with. On a lot of modern bike engines, the best you can do is just clean up any irregularities in the ports, and maybe fit larger valves. You don't often see the kind of radical reworking of ports that was common on engines like the early Harleys and American V-8s.

Ken

Click to expand...

If you want to examine state of the art, modern pushrod ports that are far ahead of things normally seen on motorcycles, have a look at the latest Chrysler 'Hemi' V8 road car engine efforts.

Momus said:

If you want to examine state of the art, modern pushrod ports that are far ahead of things normally seen on motorcycles, have a look at the latest Chrysler 'Hemi' V8 road car engine efforts.

Click to expand...

They are interesting, aren't they? More like "bathtub" chambers than hemi, and with dual plugs and D shaped exhaust ports too. I've seen a couple of Commando heads welded up and reshaped to that configuration, but never seen any flow or power comparisons. It does make it easier to get a higher compression ratio with flat top pistons.







Ken

Yeah, Mopar . . . kind of looks like a stretched Norton chamber.

In a large long port, with changing/different throttle positions, there are points where the jetting as you come off low to mid range corners can be laggy, lean, for a moment. That's where a pumper carb can get a jump on fueling that mass of air that's stalled in that long fat port. That is, for an instant, that air isn't moving fast enough to draw enough fuel . . .yet . . . so the pump gets a jump on the problem.

xbacksideslider said:

Maybe there is some confusion here. Most of the long runner set ups are a reaction, a fix for the loss of low speed power that comes with a new bigger carb. Then, the new longer runner, which actually helps the bottom of the power band, still gets blamed for the loss of low end that was caused by the big carb in the first place.

For illustration, put a longer runner (within reason) on a stock carb'd engine and the low speed power curve likely will improve but that long pipe will choke the engine at higher speeds.

Another way to say it - if you retain a short runner and change to a big carb, then between strokes the velocity/volume/mass in the port falls off too quickly, between strokes/valve events, as compared to a smaller carb. If, however, you add a longer runner to that big carb, then more volume/mass of air is transported over a longer time and that pushes the average velocity/volume/mass higher between strokes/valve events.

So . . . . within reason, with larger carbs, longer ports broaden the power band, help retain the low end.
Tell me if I'm wrong.

Another question though - where is it best, within that 14" port to place the carb? Close to or far from the head? According to inside diameter of the carb, is there a ratio that predicts the ideal length for a velocity stack?

Click to expand...

'Tell me if I am wrong'
I think you are correct.
I use 34mm Mk2 Amals and the ports in the head are tapered 2mm per side over the first 25mm. I've got Mick Hemmings rubber-mounted manifolds, and I use his rubbers which have a step in inlet tract to take out the step which is formed between the two aluminium parts, if you use radiator hose. The rubbers come in two types, one is able to be used with methanol, the other cracks. A while back, I was racing using a close ratio 4 speed gearbox with a very high first gear. The bike would lumber off without hesitation, but it was just that bit too slow for a clutch start to a race, unless I was very brave and revved the tits off it.
When I was a kid, the optimum length for a straight inlet port in inches was 28000 divided by the revs. Figure it out for yourself - the speed of sound is a mile per second, and the revs give you the frequency. It only works at one rev-range and like exhaust systems, longer pipes improve bottom end. I only use long bell mouths because the carbs are big. It is all about gas inertia and sound waves.
This stuff about standing waves and pulsed waves - it you can hear it, it is sonic. Longer pipes suit low frequency standing waves, however the optimum length is often multiples of the theoretical. . Have a look at a pipe organ. What interests me is the spray which often stands outside the entrance to the inlet tract when my motor is running, until the revs reach a certain level.

lcrken said:

They are interesting, aren't they? More like "bathtub" chambers than hemi, and with dual plugs and D shaped exhaust ports too. I've seen a couple of Commando heads welded up and reshaped to that configuration, but never seen any flow or power comparisons. It does make it easier to get a higher compression ratio with flat top pistons.



View attachment 3547

View attachment 3548

Ken

Click to expand...

The port design is a nice distillation: a real modern racing thing from professional V8 racing. The twin plug head hints at compromise for emissions and economy.