Airflow -port taper

Jim - thanks for taking the time and making the effort to share your hard earned knowledge.

Hobot - When I was a kid and tried to join in adult conversation, my mother used to say "Listen . . . . . children learn more when they are seen and not heard."

I found the comment interesting about the BSA Gold Star with the step after the GP carb. From memory there were two types of GP carburetor with the position of the low range jet reversed. On one it was ahead of the main jet, on the other it was behind. (GP1 and GP2) Perhaps the step caused turbulent flow which beneficially affects what happens within the choke of the carburetor. The question is obviously which type of GP, and why were the two types different ? I suggest that we are dealing with a resonating column of gas, not steady laminar flow.

'The GP design was the last in a series designed specifically for racing engines that started with the type 27 in the 1930s; was developed into the TT type with an enrichment (or emulsion) air slide located outside of the throttle slide bore; and then through to the RN (remote needle) type, where the throttle needle itself was located outside of the throttle bore.
This feature of remote needle was then carried through to the GP design. After the GP design was reintroduced into production by the AMAL Carburetter Company, it was a further development of the range, which became known as the GP2 type, where the pilot circuit air supply was arranged to come from filtered air. '

The GP and GP2 both had the slight reduction in bore just before the mounting flange on most of the models. The ones that didn't were the largest bore sizes in each series. On those ones they were bored out flush. The 1 1/2" GP on the DBD34 Clubmans Gold Star was the biggest in the series so it didn't have the lip in the carburettor.

I have forgotten where I read about the Gold Star intake ports being stepped down and why, but I have read it in at least two different places. I will start looking for it again so I can tell you where it is from. I know from observation that the late Gold Stars were made with this mis matching of carb and port sizing. I am pretty sure it was done on purpose.

I think just about all fuel injected engines have butterfly type throttles. This must cause some flow restriction and turbulence. I think if there was anything to be gained by having a perfectly smooth non restricted throttle body that some type of barrel type throttle would be in universal use on all really fast machines, but it isn't. Maybe a bit of turbulence is required.

The most powerfull Gold Stars made (measured on the Dyno) were the 500s with the 1 1/2" GP carb. Some people say the older Goldstars with the 1 3/8" GPs were actually faster for most racing though because of the wider power band.

This thread just shows me that I don't know anything about gas flowing or porting. I continue to fool around with it anyway. Right now I am trying to figure out how to add some features to the ports and valves on my Gold Star to prevent reversion.

xbacksider, I'm thick skinned and developed hobby to invoke sneers after the nasty reactions I got in public on Peel, especially afterwards and know any one who rides powerful motorcycles is not a rational mature behaving adult and you sir are one the best examples of it with video proof. You and me and most others want to feel like Peter Pan on our Commandos pull so just stating how Peel in her prime was configured that ain't hard to do or reverse. Pleased to learn of some examples lips/steps helped. Peel was geared tall enough that she could only pull 6000 rpm in 4th on 40 ft drop over 1/2 mile test section, if that makes any sense either. She was 3 mph slower going up hill direction. Peel was still a Combat pig power level with lips until muffler end shot out and became a pig again with other common combos on engine while leaving the exhaust the same. Don't make sense to me either but still getting off on the flash backs.

I've also heard that the Carb to intake step works well. The local Vincent engine builder /tuner does this on most of the rebuilds if they are built for power. Some just want stock but most want updates, same as we tend to do with Commandos.
So the bikes often get a mild porting, some kind of compression increase, higher lift cam, hotter ignition with a vacumn line to auto switch to a retarded curve when lugging, and a Carb size slightly bigger than the intake or port.
Whether this power difference coming from the step on his engines is real or imagined I don't know. I know that his own bike is extremely quick.

Glen

tricacent, I don't really know what 'reversion' is. I think it might simply be an theoretical explanation of something which works. My feeling is that predicting behavior in a resonating system from the results from a flow bench might not be realistic. Do we get 'reversion' when using a two into one exhaust system ?

worntorn said:

I've also heard that the Carb to intake step works well. The local Vincent engine builder /tuner does this on most of the rebuilds if they are built for power. Some just want stock but most want updates, same as we tend to do with Commandos.
So the bikes often get a mild porting, some kind of compression increase, higher lift cam, hotter ignition with a vacumn line to auto switch to a retarded curve when lugging, and a Carb size slightly bigger than the intake or port.
Whether this power difference coming from the step on his engines is real or imagined I don't know. I know that his own bike is extremely quick.

Glen

Click to expand...

They step them down because if the port was as large as the carburetor then the port velocity would be too low to be efficient.

And if the carb was as small as the port then it would not flow enough to make good power. Jim

worntorn said:

I've also heard that the Carb to intake step works well. The local Vincent engine builder /tuner does this on most of the rebuilds if they are built for power. Some just want stock but most want updates, same as we tend to do with Commandos.
So the bikes often get a mild porting, some kind of compression increase, higher lift cam, hotter ignition with a vacumn line to auto switch to a retarded curve when lugging, and a Carb size slightly bigger than the intake or port.
Whether this power difference coming from the step on his engines is real or imagined I don't know. I know that his own bike is extremely quick.

Glen

Click to expand...

I think we should consider why the Norton engine with the squish band is better than the Triumph engine with the fully hemisphered combustion chamber. If mixing is better due to turbulence, that might facilitate release of more energy from the fuel. The step after the carb. might help that happen ? Years ago there was a small fan device being sold which fitted into inlet ports with the intention of creating turbulence. I don't know whether it actually worked, or was just a gimmick. I do know this - if you use 12 to one comp. pistons in a Triumph engine, the side of the crown away from the spark plug is always coked up after a while. Centre plugs help a bit, however it seems to me that there is a lot of energy waste in the combustion process.

comnoz said:

worntorn said:

I've also heard that the Carb to intake step works well. The local Vincent engine builder /tuner does this on most of the rebuilds if they are built for power. Some just want stock but most want updates, same as we tend to do with Commandos.
So the bikes often get a mild porting, some kind of compression increase, higher lift cam, hotter ignition with a vacumn line to auto switch to a retarded curve when lugging, and a Carb size slightly bigger than the intake or port.
Whether this power difference coming from the step on his engines is real or imagined I don't know. I know that his own bike is extremely quick.

Glen

Click to expand...

They step them down because if the port was as large as the carburetor then the port velocity would be too low to be efficient.

And if the carb was as small as the port then it would not flow enough to make good power. Jim

Click to expand...

Why wouldn't the step be blended in ? You have a choice - top end or bottom end power. I can't see how fitting a big carb to a small port with a step, can achieve both.

tricatcent said:

I think just about all fuel injected engines have butterfly type throttles. This must cause some flow restriction and turbulence. I think if there was anything to be gained by having a perfectly smooth non restricted throttle body that some type of barrel type throttle would be in universal use on all really fast machines, but it isn't. Maybe a bit of turbulence is required.

Click to expand...

Fuel injected motors use butterflies because they are cheap and it is easier to implement a throttle position sensor. The restriction of the shaft and plate is not an issue because they are very large sized for the amount of airflow so the velocity is low and turbulence doesn't cause a problem.
Unlike a carb you do not need to maintain velocity through the FI throttle body to draw fuel. I use a 48mm butterfly on my bike. Jim

PS, reversion comes from the exhaust system -you are best off to deal with it in the exhaust.

acotrel said:

Why wouldn't the step be blended in ? You have a choice - top end or bottom end power. I can't see how fitting a big carb to a small port with a step, can achieve both.

Click to expand...

It takes a big carb to match the flow of a small port.
The step would be best if rounded a bit , but it doesn't take much. A step of up to 1mm doesn't even show up in an airflow test at carburetor velocities.
With a 4mm step a 3/16th inch radius pretty much makes it invisible. Jim

BSAs experiments showed that the step should be pretty sharp to provide the horsepower increase that they noticed on the dyno.

What I am planning to do to reduce reversion is shape the exhaust and inlet valves a bit differently. There is reversion at slow speeds on this engine. You can see it puffing out of the carburettor at idle. I am also switching to a cam with a bit less duration for the intake as well. The overlap will be about the same though so that may not make much difference. I just want the engine to run a bit better at the slower speeds. I don't think reversion is happening at higher engine speeds. Jim if you have any other hints about what I might do please tell me about it.

As far as what Worntorn said, about deep hemi engines having no squish bands, that is not necessarily always the case. I have a Triumph engine with pistons that have squish bands on both sides of the pistons that must provide more squish than the little squish band most Norton heads have. Part of the dome on both sides of the bore comes within .060 of the head at TDC. Both those bands are about 1/2" high and they account for about 1/2 the diameter of the piston. (there is no squish band where the valve cut outs are) The BSA has pretty good squish bands as well even though it only has 8.5 : 1 compression ratio. Even the original pistons from a Triumph Thunderbird which only had about a 7.5 to 1 compression ratio were designed with squish bands.

Thanks Jim.

What you have is basically a diffuser (diverging walls) or as I referred to earlier as an "evase". It would be interesting to see the third picture from the bottom (34mm Carb to 34mm to 30mm to 35mm) with an even longer diffuser that ends up at say 38 or 40mm; this would further reduce shock loss at the discharge end. I suspect the minor difference (118.2 cfm versus 120.8 cfm) has something to do with the lip radius (or lack there of) with the 34mm carb and 34mm manifold. This (34mm Carb to 34mm to 30mm to 35mm) basically is converting more of the kinetic energy (velocity) back to static energy (pressure). This diffuser also makes the arguement for bigger valves and larger ports for greater flow.

Yes, I know it goes without saying and yes there is the velocity trade off.

Your last picture is interesting in that you can achieve as good (actually slightly better) flow with less port volume and higher velocity - all good stuff. The down side is you may exasperate sonic choke. On the other hand, the third picture from the bottom would probably be just as susceptible to sonic choke since it converges to 30mm anyway so the botton picture would appear to be a winner.

Perhaps the air stalling at the 34mm to 30mm lip is forming some sort of desireable entrance bell naturally.

Last question is how would the abrut change in cross sectional area behave with the waves and harmonics - I don't have a clue.

All good stuff there.

tricatcent said:

BSAs experiments showed that the step should be pretty sharp to provide the horsepower increase that they noticed on the dyno.

What I am planning to do to reduce reversion is shape the exhaust and inlet valves a bit differently. There is reversion at slow speeds on this engine. You can see it puffing out of the carburettor at idle. I am also switching to a cam with a bit less duration for the intake as well. The overlap will be about the same though so that may not make much difference. I just want the engine to run a bit better at the slower speeds. I don't think reversion is happening at higher engine speeds. Jim if you have any other hints about what I might do please tell me about it.

Click to expand...

The sharp edge on the port would effectively make the port smaller which would increase the velocity in the active area of the port. They might find a bit more power by making the port smaller yet and matching the carb with a radius.

I would not think air "puffing out of the carb" at idle would be from reversion. Real reversion comes from the returning exhaust wave during overlap. The exhaust system would have to be really "off" to return the wave at idle. I guess you could change the exhaust length and see if the puff goes away.

It sounds more like very late intake valve closing is allowing the air to push back out the intake after BDC. I suspect any valve or seat damming would hurt the intake flow but I can't say for sure.

I would make sure the intake valve is closing at the right time and make sure the valve is centered on the seat. Beyond that you may have to live with it or install a softer cam. Jim

Dances with Shrapnel said:

Thanks Jim.


Your last picture is interesting in that you can achieve as good (actually slightly better) flow with less port volume and higher velocity - all good stuff. The down side is you may exasperate sonic choke. On the other hand, the third picture from the bottom would probably be just as susceptible to sonic choke since it converges to 30mm anyway so the botton picture would appear to be a winner.

All good stuff there.

Click to expand...

Making the diffuser longer and larger does increase the flow -although not very much.

I have kept most of my testing similar to what you see here as it is something that can physically be done with a Norton head and end up with the correct tuned length.

And Yes, the setup in the last picture was the winner on the dyno on a 750 cc race engine. Jim

forgive me some more to clarify factoids. I did study rocket science but dropped out for less math in slide rule days good vision and steady hand was as important as the current length of digital displays for enough significant digits. Also in VietNam era rebellion not to end up in govt or major industry complex.

Steps are used to prevent both Exhaust Reversion flow suck backs and the reversion by sound pressure fronts bounce backs. To apply same anti-reversion logic to intakes, the step up would have to be into a bigger port than the manifold-carb, exactly opposite of all those that liked the step down effects in intakes of their engines. Norton exht has a bit of reversion lip, more so if D port made and Harley vendors sell anti-reversion cones that are supposed to deflect the reversions to dead pocket not valve pocket, which is the reverse of comnoz finding a slight funnel out tube flow increases.

The original racer reason FI went to butterflies was the great suction developed at hi rpm low throttle causing the flat slides to hang an instant then over release a bit which caused pilots to mess leathers just like the Drouin Lake injectors though the racers certainly had pull closed cables. Butterflies main value is ease to operate smoothly, all the rest comnos lists followed afterward to refine to current state. If ya look into FI butterfly hop up you find ways to trim them to flow a bit better yet... instead of size changes which screws with air and temp and pressure sensors so back to start on re-mapping digital brains.

We all love the roaring sound blasts out of well tuned power but some of us like the imploding howling of intakes even better.
http://www.youtube.com/watch?v=Mf8YFg4IAyY

Oh yeah might turn those 'velocity' stacks upside down.
http://www.youtube.com/watch?v=3N5Rxn-AHYA

acotrel said:

tricacent, I don't really know what 'reversion' is. - - - - - - -
- - - - - - - - - Do we get 'reversion' when using a two into one exhaust system ?

Click to expand...

Megaphonitis is a perfect example of reversion.During valve overlap,the exhaust can push some of the fresh intake back out the intake port.It later gets drawn into the cylinder,but it's been past the main jet 3 times and the mixture becomes rich.You'd also have some exhaust products remaining in the cylinder,instead of all fresh charge.

When it comes on the mega,all this reversion stops and it runs well.At less rpm,it can run lousy.

Here's a link to the GP carb on the BSA,and Roland Pike's experiments http://beezagent.blogspot.com.au/2009/02/roland-pike-autobiography-chapter-26.html

hobot said:

forgive me some more to clarify factoids.

Click to expand...

Your forgiven if -you post your pig fetish elsewhere. Jim :?

I

tricatcent said:

As far as what Worntorn said, about deep hemi engines having no squish bands, that is not necessarily always the case. I have a Triumph engine with pistons that have squish bands on both sides of the pistons that must provide more squish than the little squish band most Norton heads have. Part of the dome on both sides of the bore comes within .060 of the head at TDC. .

Click to expand...

I'm not sure when I said that, but it sounds very intelligent, even though it might be wrong Perhaps in some other thread?
If your squish clearance is 60 thou then it is likely ineffective. It seems 30 to 50 thou is a good range, with more power and quench happening at 30 than 50. Much closer than 30 Thou and there is the possibility of piston contact with the head at high rpm.

Back to flow

Glen

worntorn said:

If your squish clearance is 60 thou then it is likely ineffective. It seems 30 to 50 thou is a good range, with more power and quench happening at 30 than 50. Much closer than 30 Thou and there is the possibility of piston contact with the head at high rpm.

Click to expand...

At less than 0.045",the pistons are likely to touch the head at times if you have aluminium conrods.
0.045" is safe,and you're likely to get that if you leave out the cylinder base gasket on a T140 (just use Loctite 518 instead).Some people manage to get away with a little under 0.030",but they have to be careful during warm-up and also avoid over-revving.