Airflow -port taper

Status
Not open for further replies.
Dances with Shrapnel said:
Interesting demonstration comnoz. I am pretty sure the difference in the flows is due to the difference in shock losses which is a function of velocity. The 35mm has 36% greater cross sectional area so velocity will be 73% of that of the 30mm side. A higher velocity discharge will incur a greater shock loss. The greater the air velocity at the discharge the greater it wants to contract (vena contracta).

If interested, do a Google search on the words <mine fan evase>. Basically evases reduce the shock loss of discharged mine ventilation air.

Same thing as comnoz demonstration and how it applies to porting.
Click to expand...

Shock losses occur when there are shock waves...shock waves occur when the flow is supersonic. Flow in Jim's tube and carbureted manifolds is subsonic.

A vena contractor occurs when fluid leaves a sharp edged orifice.
 
jaydee75 said:
If I remember from college (40 yrs ago), a diverging tube with a seven degree included angle is the most efficient for flow thru a tube. That's why exhaust megaphones are made the way they are. I know Jim is on the right track, I'm just trying to remember the optimum angle.
Jaydee
Click to expand...

My memory says seven degrees is about right....too much divergence, and the flow separates from the walls, creating turbulence, which sucks energy off the flow.
 
texasSlick said:
Shock losses occur when there are shock waves...shock waves occur when the flow is supersonic. Flow in Jim's tube and carbureted manifolds is subsonic.
Click to expand...

Never mentioned anything about "shock waves" only "shock loss", all relatively low fluid velocity phenomena; it happens all around us and is fundamental fluid dynamics. Closest I came to a discussion on supersonic is when referring to sonic choke as part of the trade off between smaller ports and higher velocity. There are actually a few tricks to help skin that cat also including a Helmholtz resonator. It all has to do with making the best of the time available to get air into the combustion chamber ......but I diverge here.

texasSlick said:
A vena contractor occurs when fluid leaves a sharp edged orifice.
Click to expand...

Yep, good description and pretty much spot on to what Jim "comnoz" has been showing here. Even with the clay rounded edges it is still relatively sharp and abrupt discharge with very little chance of energy recovery.

The whole point is the observed difference in flow rate is due to the relative velocities and resultant shock loss at each end when flipped around.
 
Shock loss happens anytime flow is disrupted -whether it is a turn , a surface irregularity or a valve guide sticking into the flow.

The higher the velocity of the air -the greater the shock loss. The air needs to be slowed down to make a turn or go around the guide with less loss.

I have used valve guide fairings. A fairing on the cylinder side of a guide helps a little. I could never see much difference with the fairing on the carb side of the guide.

I made up some mini-fairings one time by machining the part of a large 850 guide that is in the port, into a teardrop shape. They didn't make any difference that I could see. The fairing was probably too short to be of any value. Jim
 
comnoz said:
Mark said:
X-file said:
I think the problem here is that the flow test was done with a radius on BOTH ends of the tube.
Click to expand...
I noticed that too.
Jim, why did you do it like that?
Click to expand...

A cylinder head has something of a radius at both ends. One at the carb mouth and the other at the valve entry into the cylinder with a good mulit-angle valve job. Of course the valve end radius is smaller.

The tube was just made to show the effect of the wall taper.

I have a cylinder head that has been bored large and strait to the valve guide to hold a port insert. I have made a lot of different inserts with different tapers and shapes and tested them both on the flowbench and the dyno. This has been where I have done the most research.

Here is the head and a couple different inserts that I have tried. The head has been on my 880 for years up to the big blow-up. Jim

Airflow -port taper


Airflow -port taper
Click to expand...

Does "wall taper" only apply to the cylinder head port, or does the theory include the entire intake ie inlet manifold, and carb?

Meaning: I know you have said in the past that an unbutchered 30mm RH10 head is optimum for an 850. Are you also saying/ suggesting / alluding to the fact that the ideal carb and manifold should also be 30mm (assuming twin carbs)?

And does it extend to velocity stacks... should they be parallel?

Jim, I realise that intake charge 'art' is a constant trade off between volume and speed, and generally, more volume seems intuitively good, but I struggle to relate to the CFM figures; How much does a Commando need?
 
Hard to beat the sensation when various flow features are felt in the seat of the cycle. John/Dances set us straight on "shock loss" vs "shock wave" choke terms, thanx.

Sorry I meant Al-ment and glad to hear your basic data found it didn't hurt and might of helped some with interrupters down stream of valve. Goldie Locks compromise to make surface texture smooth enough to avoid drag yet rough enough to energize thickening boundary layers, so sliding scale of size depending on the application, intakes or semi trucks, which Dreer has investigated and invested in.

The reason for divergent megaphone is to spread the rpms of effective 'softened' reflected low pressure sound waves aka: negative polarity front to arrive at valve about right time not push spent gases back into chamber and even better if can help suck intake into chamber sooner just before exht closes. Step ups in dia. reflect neg wave back while step down reflect positive wave back. Peel may hit 1600'F EGT in 1st header bend so may have to use Inconnele[sp] for first 7-8 inches in 1.5" dia. then step up to 1 3/4" ID for the next ~18" to the Y-junction of the headers then 2" for another foot or so into the long dong megaphone with exit tapper down to 1 3/4" over last 4". If Peashooters really were reversed cone megaphones instead of just straight though tube with pretty cover they would work better but way louder. Its the collector section beyond a merger before muffler that makes most help in exhaust flow design.

Thanks for taking the time comnoz to test entry-exit readius details to fill in our blank open minds, or at least my mind. Radius bell the intake and sharpen the exits.
 
Fast Eddie said:
Does "wall taper" only apply to the cylinder head port, or does the theory include the entire intake ie inlet manifold, and carb?

>It would apply equally through the whole intake tract but there are a lot of other considerations when applying it.
A carb flows considerably less than a straight tube -there are a lot of things to disrupt the airflow. IE-a 34mm carb = a 30 mm strait tube.<

Meaning: I know you have said in the past that an unbutchered 30mm RH10 head is optimum for an 850. Are you also saying/ suggesting / alluding to the fact that the ideal carb and manifold should also be 30mm (assuming twin carbs)?

> The best way is to get rid of the curved manifold. When the manifold needs to be there, I generally like to use the larger manifold to slow the flow through the curve and then step down using a radius at the port entrance.
A little taper in the manifold seems to work OK. There is lots of turbulence in the manifold anyway. The carb should be 4 to 6 mm larger than the smallest part of the port with a conventional carb. A little smaller with a smoothbore.<

And does it extend to velocity stacks... should they be parallel?

>If the velocity stack is the same ID as the carb then they need to have parallel walls and a radius entrance.
Many stacks are much larger than the carb bore and then the taper does not matter much. They are more just a way to keep the standoff contained and don't contribute to the tuned length. <

Jim, I realise that intake charge 'art' is a constant trade off between volume and speed, and generally, more volume seems intuitively good, but I struggle to relate to the CFM figures; How much does a Commando need?
Click to expand...

>A long stroke motor does not really need a high flow potential. I don't have any minimum numbers - I like to get the highest numbers possible with a minimum port diameter of about 28mm for a mild street 750 to 32mm for a full race 850. I like to see at least 450 fps velocity at 20 inches pressure. Any time I have increased CFM numbers at the expense of velocity I have netted a lose of power "under the curve"
It is very easy to make the port flow more than the valve can when using stock sized valves which kills the velocity in the port. < Jim
 
How does port flow cfm relate to pulsed intake flow though. ?

Don't Manx Nortons for example get more power from pulse tuned intake/ehaust ports than mere flow numbers ?
Equally, intake length for a Commando may be as/more important than port size.

Rawlins/Baker sure got some good speed/power out of that 850 with big ports. ?
143 mph average 2 way over the flying kilometre (mile ?) isn't hanging about...
And was apparently still quite strong for average street riding - being registered and on the road.
 
Rohan said:
How does port flow cfm relate to pulsed intake flow though. ?

>It doesn't -it is only a way to compare ports and modifications. You still have to do the dyno tests to see what makes power. Then you do the flowbench tests to see what worked and how to make another head work.<


Don't Manx Nortons for example get more power from pulse tuned intake/ehaust ports than mere flow numbers ?
>yes<
Equally, intake length for a Commando may be as/more important than port size.
>.that needs to be right also<

Rawlins/Baker sure got some good speed/power out of that 850 with big ports. ?
143 mph average 2 way over the flying kilometre (mile ?) isn't hanging about...
And was apparently still quite strong for average street riding - being registered and on the road.
Click to expand...
 
[quote="Rohan"
Rawlins/Baker sure got some good speed/power out of that 850 with big ports. ?
143 mph average 2 way over the flying kilometre (mile ?) isn't hanging about...
And was apparently still quite strong for average street riding - being registered and on the road.[/quote]

I would not say that you can not get good peak power with large ports. Peak power is what counts when you are looking for max MPH.

What I look at when building a motor for street or roadracing is "power under the curve" That is the area under the horsepower line over a rpm range that is going to be used when accelerating. [determined by the rpm drop when shifting]
If you have a close ratio trans then you can put a narrower powerband and possibly a higher peak power to use. But if you have a normally spaced gearbox then a short peaky powerband isn't much good when you have to wait for the rpms to build back to the powerband on each shift. Jim
 
I think too many of you are equating separated flow turbulence, and eddy flow vortices as "shock" losses. Such loss of "smooth" flow reduces the kinetic energy of the flow, but "shock" loss is very inappropriate. Shock waves occur in supersonic flow and do cause very large gains in entropy and that requires energy to sustain it. As an aerodynamicist, you say shock loss, I think shock wave, and think big entropy gain, which translates to big stagnation pressure loss (stagnation pressure is a measure of flow energy)

Skin friction is a function of flow velocity. If the tube diverges, the velocity slows, reducing skin friction, the reverse occurring with a converging tube. Turbulent boundary layer skin friction is greater than laminar. With either type boundary layer, more flow will pass thru the tube with lessor skin friction.

Hobot-speak is difficult to interpret, but his explanation of choked sonic flow is fundamentally correct. I am too lazy to make the mathematical calculations, but I suspect near sonic choked conditions may occur in a Norton manifold at 7000 -8000 plus RPM . Sonic velocity, or near sonic will occur eventually as the RPM increases. The only way to stave it off is to increase tube diameter. Once sonic speed is hit, there is no way to get more mass into an engine by increased velocity (the flow is "choked" and simply will not go faster. If an engine is supercharged, more mass can be put into the cylinder by pressure, not by velocity.

As the velocity in a tube increases, so does the boundary layer, and this has the effect of "reducing" the effective diameter of the tube, leading to a further increase in fluid velocity, and concomitant increase in boundary layer. Thus the choked flow condition happens suddenly.

Students often ask " what happens if the driving pressure is increased when the flow becomes choked? Answer...a shock wave forms, causing "shock" loss of stagnation ( driving) pressure.
 
Thanx I think Tex. Pressure is another way of saying denser at same temperature. Once particles are at speed of sound that's as fast as they go so only packing more through at speed of sound makes more power in chamber.
Surpercharges can both speed up the flow in a port faster than just sucked in and further pressurize the flow thru the sonic choke area, ie: more molecules moving close to sonic speed. Un-boosted Nortons are not capable of approaching sonic choke flows as would blow up at the rpms required. Peel @ 8000 with 10 PSI boost on standard size 6mm intake valve with 32 mm ports will only see maybe Mach 7 so not much a restriction if any sonic choke wise. Peel's Mach index value comes out around .16 to .17 so not an issue as .55-.60 is troublesome range. If ya study up on the cam profile to help ease sonic choke a drag cam is the way to go there.

Plug in your data for fun here
http://www.rbracing-rsr.com/machcalc.html
http://www.wallaceracing.com/machcalc.php
 
Whoa!!! Slow down. As I mentioned earlier about the trade off of smaller ports and higher velocities and the concept of sonic choking, there is more than one way to skin this cat. The valve open time available can be better used to move airflow and spend less time in sonic choke - use a Helmholtz resonator to achieve higher average flow rates while reducing peak flow rates. Agreed that a Norton at high rpm may likely encounter sonic choke with smaller ports but................ that is not what this thread is about nor is it about supersonic shock waves.

In this thread a simple experiment was presented and the difference in air flow results is due to the shock loss, yes shock loss, well below the speed of sound. This may not be a term many people are familiar with outside of fluid dynamics and ventilation systems. Judging by comnoz data (30-35mm dia ports and 100-140cfm), it is well below supersonic so lets not confuse things here.

Simply stated, in the experiment presented by comnoz, when a fluid exits a pipe, tube or conductor, there is a pressure loss due to not being able to convert the remaining kinetic energy into potential energy. Simple energy balance. The taper of the tube has nothing to do with the phenomena, the air velocity at exit has everything to do with the loss.

Normally the experiment would be run both ways at constant flow rate and a measure of inches water gauge would indicate the relative difference in pressure (pressure loss) to achieve the flows.
 
I have not been able to approach sonic flow speeds in a Norton intake port with a little less than 1 bar of available air pressure. About 540 fps is the highest speed I have seen with a "bent" port like a Norton. There are some formula 1 engines with very straight downdraft ports that were approaching 650 fps a few years ago. I don't know what they are up to now.

A port can plug due to turbulence at average speeds much lower than sonic speeds if the shape is wrong. You can hear the turbulence on the flowbench and beyond that point added pressure will not increase the flow CFM -in fact added pressure will often cause the CFM to drop.
Is it because flow in certain areas of the port have reached sonic velocitys? I don't know. Jim
 
comnoz said:
I would not say that you can not get good peak power with large ports. Peak power is what counts when you are looking for max MPH.

What I look at when building a motor for street or roadracing is "power under the curve" That is the area under the horsepower line over a rpm range that is going to be used when accelerating. [determined by the rpm drop when shifting]
Click to expand...

Be interesting to see that dyno chart for the Rawlins/Baker factory Norton Villiers 850 bike then. ?
It was pretty strong in drag racing at the time too - wasn't it doing less than 12 sec runs.
 
Dances with Shrapnel said:
In this thread a simple experiment was presented and the difference in air flow results is due to the shock loss, yes shock loss, well below the speed of sound. This may not be a term many people are familiar with outside of fluid dynamics and ventilation systems. Judging by comnoz data (30-35mm dia ports and 100-140cfm), it is well below supersonic so lets not confuse things here.

Simply stated, in the experiment presented by comnoz, when a fluid exits a pipe, tube or conductor, there is a pressure loss due to not being able to convert the remaining kinetic energy into potential energy. Simple energy balance. The taper of the tube has nothing to do with the phenomena, the air velocity at exit has everything to do with the loss.

Normally the experiment would be run both ways at constant flow rate and a measure of inches water gauge would indicate the relative difference in pressure (pressure loss) to achieve the flows.[/quote

" It is well below supersonic.....etc". Exactly....so keep shock waves and shock losses out of it. Call it something else it you insist. Shock waves do not exist in subsonic flow. Shock loss is a term this aerodynamicist is not familiar with in subsonic flow.

"Normally the experiment....etc" Comnoz kept the pressure differential constant and measured flow variance both ways....one can interpret the effects either way.

Entance and exit effects are not part of the experiment if comnoz's pressure taps were BETWEEN the tube ends.....So just where were they? Until this can be clarified, I'll stand that skin friction is the dominant mechanism governing the pressure drop at constant flow, or flow drop at constant pressure.
Click to expand...
 
There are no pressure taps in the flow tube. The flow tester is simply set to maintain 20 inches of differential between one end of the tube and the other. Jim
 
Comnoz thanks for the flow magic show. To max power with boost one lowers CR for High Boost, ie" over 15 PSI one Bar. That requires special fuels. Part the reason I decided to up CR with a bit of boost on top took into consideration the self limiting slower flow "shock loss"? nature of the poor ole early Combat head aka: 'CHO' to me. Ken told me don't even bother to test as would just depress and by golly that's how it feIt on Peel. So so Sorry for the fella 'hogging' his out. I think I read somewhere the later non Wovernheaven Combats had better reduced size porting. Calc's imply Peel could hit 17:1 nearing 8grand but that don't consider the self choking CHO nature which should flatten out the torque curve even if the blower boost curve is steepening. I think I'll get the power band I seek pretty tame 920 down low, wakes up hard mid range for digital traction then levels off pull toward top off by the clogging that makes throttle control more innate one to one sense w/o the harmonics coming and going on singing tire tunes, even in lower gears. Its a pretty crude impeller so likely its mostly cavitating and pooping out by then too.

Airflow -port taper


Ugh, What phenomena is cavitation demonstrating - shock loss, shock wave front, sonic choke or zero point vortex energy?
Airflow -port taper


Dances, your exit size point struck a cord with me though exhaust is on my mind most. I'm big fan of Helmholtz resonance and have seen applications to both intake manifolds and exhaust systems. Mostly 2smokes though. Can ya think of an example where the head ports or valves too small so sonic choke limited power before it blew up? Only boosted engines that I know of have sonic choke issues if sticking with Intermittent Combustion engines. I've read that on exhaust headers its usually worth while to go up 1/8" over factory to allow for the boundary layer and ease exiting. I am in love with my long dong Dunstall spread of helpful harmonics as Peel couldn't get out the way of ordinary C'do's with one 2-1 header. Strange as the P!! 30" dual open pipes kicked my ass start to finish better Peel ever could in a bee line. Had a mystical Ozark's experience during an earthquake and a stadium size cave near by with a very narrow opening. Ever heard of a "booster bottle"?

Airflow -port taper


Pretty easy to cut a jagged crude intruding gasket Jim, if still in the mood to see what that does. Both ends in case Dances is onto something.

Airflow -port taper
 
comnoz said:
I have not been able to approach sonic flow speeds in a Norton intake port with a little less than 1 bar of available air pressure. About 540 fps is the highest speed I have seen with a "bent" port like a Norton. There are some formula 1 engines with very straight downdraft ports that were approaching 650 fps a few years ago. I don't know what they are up to now.

A port can plug due to turbulence at average speeds much lower than sonic speeds if the shape is wrong. You can hear the turbulence on the flowbench and beyond that point added pressure will not increase the flow CFM -in fact added pressure will often cause the CFM to drop.
Is it because flow in certain areas of the port have reached sonic velocitys? I don't know. Jim
Click to expand...

Agree with all stated. Regarding last sentence:

The numbers you cite are AVERAGE fps....sonic choke can occur if velocity on a particular streamline approaches sonic. For example, the velocity profile of laminar flow in a round tube is a parabola of revolution. In this case, the velocity of the stream tube on the axial centerline is twice the average mass velocity. Thus with your numbers, the centerline velocity would be 1080 fps....definitely near sonic (depends on temperature). As I stated earlier, once the flow approaches a choked condition, things happen to accelerate the process....which you also allude to above. Maintaining laminar flow in such a flow condition is doubtful. The laminar profile will disrupt to a turbulent, dropping the centerline velocity to near average mass flow velocity, which removes the sonic condition, but at the expense of greater skin friction. It is useful to keep in mind that nature ALWAYS takes the path of least resistance....in this example, nature eliminates the sonic shock wave which is a big energy hog.


comnoz said:
There are no pressure taps in the flow tube. The flow tester is simply set to maintain 20 inches of differential between one end of the tube and the other. Jim
Click to expand...

In that case, the entrance and exit effects are part of the experiment. A radius at each end minimizes the effects. I do not think these effects are the principal reason why flow is different both ways....that is my professional opinion...I'm done.

Slick
 
Status
Not open for further replies.
Back
Top