108 replies to this topic

[The_R1_Kid]

Hello all, I have been wondering about cylindrical throttle bodies. Besides the obvious advantage at full throttle, are there any draw backs at part throttle. Compared to a normal butterfly throttle body. Does anyone have any more info or pictures? these are the only pictures I can find.
are cylindrical throttles still in use? I think koenigsegg uses them.

[OfficeLinebacker]

what closes them? The cylinder rotates so it's like a piece of the cylinder is rotating instead of a plate?

I know turbulence in the intake charge helps distribute the AF mix so not sure why any one approach would be better than another.

[Canuck]

I have to be honest - I'm not seeing the barrel in the 2nd and 3rd pictures. I've asked about these type before, on here, somewhere. Seems to me the consensus was they were suited for WOT and rather unsuitable for street-level applications.

Hollow camshafts...trick.

[gruntguru]

I have to be honest - I'm not seeing the barrel in the 2nd and 3rd pictures.

Its the metal bit - half way up the CF bit.

[Canuck]

Are you certain? If you look at the first shot, the barrel appears to ride inside a metal sleeve, not inside the CF trumpets.

If they are, I need a gizmo-spotting tune-up...must be getting old.

[Tony Matthews]

Are you certain?

gg is right, but it is slightly confusing as the cylindrical shaft that is the throttle not only has ports in it that at full-throttle line up exactly with the carbon intake, but has a flat machined on the side facing the centre of the vee. This flat, I think, faces upwards towards the injector when the throttle is closed. The sliding plate system also gives an un-interupted bore at full throttle, perhaps it's easier to achieve speedy, precise movement with a cylinder. Butterlies still seem very popular.

Edited by Tony Matthews, 24 June 2009 - 12:57.

[Catalina Park]

Butterlies still seem very popular.

If you want more airflow at full throttle you just make a bigger hole.

[McGuire]

Barrel valve throttles are not new. Harry Miller was using them by 1915 and they were not new then.

Against: weight, cost, complexity, packaging.

For: Less air restriction at WOT. This is mitigated by the fact that, just as Catalina Park says, it is a simple matter to size a butterfly valve for true open throttle, with some tiny pumping loss due to air friction. The barrel valve's power gain will be incremental enough to make it difficult to quantify.

A new-ish angle on all this... wet flow studies have established that the edge of a throttle butterfly, closed or open, is an excellent fuel shear point. As blobs and globules of fuel (blobules?) strike the blade, they are mashed into smaller blobs and globules, a good thing. (When fuel droplets in the airflow strike each other they tend to merge; when they impact other things they tend to separate. This is fascinating to watch live: it's almost like the fuel droplets are looking for each other.) If the fuel is injected downstream of the throttle valve the issue is moot, but with a carburetor or trumpet nozzles (as shown here) it will be significant. Here, barrel valves are combined witth trumpet nozzles -- at apparent cross purposes, from this perspective at any rate.

[OfficeLinebacker]

Barrel valve throttles are not new. Harry Miller was using them by 1915 and they were not new then.

Against: weight, cost, complexity, packaging.

For: Less air restriction at WOT. This is mitigated by the fact that, just as Catalina Park says, it is a simple matter to size a butterfly valve for true open throttle, with some tiny pumping loss due to air friction. The barrel valve's power gain will be incremental enough to make it difficult to quantify.

A new-ish angle on all this... wet flow studies have established that the edge of a throttle butterfly, closed or open, is an excellent fuel shear point. As blobs and globules of fuel (blobules?) strike the blade, they are mashed into smaller blobs and globules, a good thing. (When fuel droplets in the airflow strike each other they tend to merge; when they impact other things they tend to separate. This is fascinating to watch live: it's almost like the fuel droplets are looking for each other.) If the fuel is injected downstream of the throttle valve the issue is moot, but with a carburetor or trumpet nozzles (as shown here) it will be significant. Here, barrel valves are combined witth trumpet nozzles -- at apparent cross purposes, from this perspective at any rate.

This concept of a fuel shear point is, I suppose somewhat compatible with the idea of turbulence aiding mixture?

[Tony Matthews]

This concept of a fuel shear point is, I suppose somewhat compatible with the idea of turbulence aiding mixture?

I find that driving with my head out of the widow, at an angle of about 45 degrees to the direction of travel, mouth slightly open, the gin and tonic mixes very well.

[OfficeLinebacker]

I find that driving with my head out of the widow, at an angle of about 45 degrees to the direction of travel, mouth slightly open, the gin and tonic mixes very well.

..while listening to music with heavy bass and bobbing your head in time?

[gruntguru]

..while listening to music with heavy bass and bobbing your head in time?

Perhaps listening to those 16 Cossie trumpets at 20k+ would be better?

Edited by gruntguru, 26 June 2009 - 04:58.

[cheapracer]

Hollow camshafts...trick.

You mean like millions of Tpyota's for the last 30 years not to mention their crankshafts...

[phantom II]

100 millionth Toyota was built today.

You mean like millions of Tpyota's for the last 30 years not to mention their crankshafts...

[Greg Locock]

You mean like millions of Tpyota's for the last 30 years not to mention their crankshafts...

Good catch -who has the patent on expanding the tubes to lock the individual cams in place?

[Tony Matthews]

..while listening to music with heavy bass and bobbing your head in time?

What, and risk spillage? Have you seen the price of gin?

[Powersteer]

This concept of a fuel shear point is, I suppose somewhat compatible with the idea of turbulence aiding mixture?

At the kind of air speed and high pressure fuel injection system these engine run, I am not too sure how critical that equation is.


The construction is amazing, having the cam caps horizontal to be able to carbon fiber as much of the cylinder head as possible. Its like having a cylinder head chassis straight down the middle and everything else carbon fiber, what a beauty. On mulsannecorner.com, the Judd JV5 has the barrels intergrated with the cylinderhead and it was one of read it was done to cut weight more than anything else fromt he JV4. The siamese liner has what looks like a structural link, wow, must have been a headache cast around this.

Edited by Powersteer, 25 June 2009 - 06:17.

[gruntguru]

The siamese liner has what looks like a structural link, wow, must have been a headache cast around this.

Are you sure that's not a gasket?

[Powersteer]

Are you sure that's not a gasket?

Yeah, I guess you are right, trick stuff.

Edited by Powersteer, 25 June 2009 - 07:29.

[phantom II]

GM

Good catch -who has the patent on expanding the tubes to lock the individual cams in place?

[OfficeLinebacker]

At the kind of air speed and high pressure fuel injection system these engine run, I am not too sure how critical that equation is.

Good point. I suppose it's easier to model atomization from a nozzle and flow through a smooth bore than to try to imagine the air/fuel flowing around all kinds of strange bends and around pointy bits.

One thing I've seen in racing catalogues is valves that have had the stem necked down between the back of the valve and the valve guide, presumably to increase flow. I don't see it here...so I suppose it's usefulness is suspect at best.

[Canuck]

Perhaps listening to those 16 Cossie trumpets at 20k would be better?

Guess I need to open up a Toyota one of these days...

[gruntguru]

Good point. I suppose it's easier to model atomization from a nozzle and flow through a smooth bore than to try to imagine the air/fuel flowing around all kinds of strange bends and around pointy bits.

One thing I've seen in racing catalogues is valves that have had the stem necked down between the back of the valve and the valve guide, presumably to increase flow. I don't see it here...so I suppose it's usefulness is suspect at best.

Looking at those stems, I would say they are at the limit in terms of thin. No point in stepping up to thicker diameter in the guide area unless there is a specific reason eg side loads from rockers would require more bearing surface area in the guide.

[Tony Matthews]

Looking at those stems, I would say they are at the limit in terms of thin. No point in stepping up to thicker diameter in the guide area unless there is a specific reason eg side loads from rockers would require more bearing surface area in the guide.

Also looking at those stems, in original pic #2, you can see that they are stepped, or necked, but the change in diameter is small, and the neck is very long, so it isn't that obvious...

[McGuire]

Good point. I suppose it's easier to model atomization from a nozzle and flow through a smooth bore than to try to imagine the air/fuel flowing around all kinds of strange bends and around pointy bits.

Atomization is essentially a myth.

[OfficeLinebacker]

Atomization is essentially a myth.

then what's it called? lol surely it's a relatively homogeneous mixture?

[McGuire]

then what's it called? lol surely it's a relatively homogeneous mixture?

Noo, we wish. This is a gross idealization/simplification but the bulk of the fuel portion of the intake charge -- the best-moving part of it anyway -- travels down the center of the port like a string of snot (or several strings) then breaks or shears over the edge of the intake valve. Along the way, every time it strikes a shear point (throttle blade, flow vane, valve guide boss, etc) -- it breaks apart, then begins to reform. Turns do the same thing to a lesser extent. Some fuel sticks to the wall; some skips and tumbles along the wall; some shatters and is thrown back out into the flow. You can see why this is: For one thing fuel and air have different mass.

And so the air-fuel charge as it lands in the cylinder is not terribly homogeneous. It can vary locally from 100 percent fuel to 100 percent air. So once we ignite this mess, the flame front does not advance across the chamber at a uniform rate. The burn rate speeds up as the flame front moves through more combustible zones and slows down through the less combustible zones. Mapped, it looks like a shape-shifting one-cell organism. In the voids on the tail end of this process (the "end gases") is where we get detonation. There are little pockets or bubbles of intake charge that refuse to be ignited by the flame front, yet they are being crushed and heated by the expanding combustion all around them. At some point they self-ignite, blam. Thus it is that early ignition timing and lean A/F mixtures both tend to produce knock, by promoting the conditions in which it occurs.

[OfficeLinebacker]

Noo, we wish. This is a gross idealization/simplification but the bulk of the fuel portion of the intake charge -- the best-moving part of it anyway -- travels down the center of the port like a string of snot (or several strings) then breaks or shears over the edge of the intake valve. Along the way, every time it strikes a shear point (throttle blade, flow vane, valve guide boss, etc) -- it breaks apart, then begins to reform. Turns do the same thing to a lesser extent. Some fuel sticks to the wall; some skips and tumbles along the wall; some shatters and is thrown back out into the flow. You can see why this is: For one thing fuel and air have different mass.

And so the air-fuel charge as it lands in the cylinder is not terribly homogeneous. It can vary locally from 100 percent fuel to 100 percent air. So once we ignite this mess, the flame front does not advance across the chamber at a uniform rate. The burn rate speeds up as the flame front moves through more combustible zones and slows down through the less combustible zones. Mapped, it looks like a shape-shifting one-cell organism. In the voids on the tail end of this process (the "end gases") is where we get detonation. There are little pockets or bubbles of intake charge that refuse to be ignited by the flame front, yet they are being crushed and heated by the expanding combustion all around them. At some point they self-ignite, blam. Thus it is that early ignition timing and lean A/F mixtures both tend to produce knock, by promoting the conditions in which it occurs.

So it's really the passing over the edge of the valve that does the mixing eh?

Also, we're mostly talking about road cars here, right? In the pics it doesn't look like there's much space to re-form from a mist into a string of snot.

Finally, we all know cooler air and fuel mean denser intake charge means more power. What about experimenting with heating the stuff to at least encourage evaporation?

[Tony Matthews]

In the pics it doesn't look like there's much space to re-form from a mist into a string of snot.

Finally, we all know cooler air and fuel mean denser intake charge means more power. What about experimenting with heating the stuff to at least encourage evaporation?

I think you will find, OLB, that a string of snot, to use McGuire's charming simile, doesn't need much space to form - but I digress...

The heating thing is not uncommon, at least in the days of carburettors, as water-heated - or conduction-heated from the exhaust - inlet manifolds are well known.

Now I am going to be jumped on and told that it was to prevent carb icing in the winter...

[The_R1_Kid]

I think you will find, OLB, that a string of snot, to use McGuire's charming simile, doesn't need much space to form - but I digress...

The heating thing is not uncommon, at least in the days of carburettors, as water-heated - or conduction-heated from the exhaust - inlet manifolds are well known.

Now I am going to be jumped on and told that it was to prevent carb icing in the winter...

Didn't Smokey Yunick mess with heating of fuel prior to the carb or injector?

[gruntguru]

So it's really the passing over the edge of the valve that does the mixing eh?

Also, we're mostly talking about road cars here, right? In the pics it doesn't look like there's much space to re-form from a mist into a string of snot.

Finally, we all know cooler air and fuel mean denser intake charge means more power. What about experimenting with heating the stuff to at least encourage evaporation?

Don't get too concerned with McGuire's extreme description. F1 achieves very high levels of "atomisation". There isn't too much re-forming (agglomeration) either otherwise they couldn't afford to put the injectors so far from the valve head. They would certainly like smaller droplets though - if it wasn't for the 100bar (1500psi) limit on fuel pressure, they would be running a much higher pressure. (Common-rail diesel systems run at 1000 - 2000 bar (15,000 - 30,000 psi).)

Regions of 100% fuel and 100% air in the cylinder are individually extremely small and collectively constitute a small percentage of the total volume.

[gruntguru]

Now I am going to be jumped on and told that it was to prevent carb icing in the winter...

Also to prevent the "string of snot" from pooling in the manifold so you are quite correct. Unfortunately heating and even excessive evaporation is detrimental to VE and therefore power.

Edited by gruntguru, 26 June 2009 - 21:12.

[Tony Matthews]

Unfortunately heating and even excessive evaporation is detrimental to VE and therefore power.

I realise that , and realistically the mixture needs to be cooled again before reaching the combustion chamber, but that means a long, probably serpentine, duct. Also, I suppose there is a chance, with the cooling, for precipitation, or whatever the engineering term is for a return to a snot-like fuel/air conglomeration. Is this where direct injection wins? I am not familiar with the pro's and con's of this system, although it sounds logical.

Do you mean that F1 limits injection pressure to 100 bar, or that this is a natural limit for gasoline? I know that the lubricating properties of diesel fuel is beneficial.

[OfficeLinebacker]

I realise that , and realistically the mixture needs to be cooled again before reaching the combustion chamber, but that means a long, probably serpentine, duct. Also, I suppose there is a chance, with the cooling, for precipitation, or whatever the engineering term is for a return to a snot-like fuel/air conglomeration. Is this where direct injection wins? I am not familiar with the pro's and con's of this system, although it sounds logical.

Do you mean that F1 limits injection pressure to 100 bar, or that this is a natural limit for gasoline? I know that the lubricating properties of diesel fuel is beneficial.

F1 regulations limit the pressure.

I've always been interested in fuel efficiency in my road cars so this whole fuel atomization thing is like one of my hobbies. In "tuner" circles people talk about CAIs and in "ecomodder" circles we talk about WAIs. (Cold/Warm Air Intakes)

Remember in the old time carbureted V8s there was a heat riser thing attached to a heat-sensitive valve? When the engine was cold air was drawn from the area of the headers! Presumably there was a point at which the intake manifold was plenty hot so the valve closed and air was drawn from next to the radiator or wherever.

[cheapracer]

Remember in the old time carbureted V8s there was a heat riser thing attached to a heat-sensitive valve? When the engine was cold air was drawn from the area of the headers! Presumably there was a point at which the intake manifold was plenty hot so the valve closed and air was drawn from next to the radiator or wherever.

Just about every carby engine ever made had some basic form of inlet heating either air, exhaust or water.

I think the first 'hot up' trick I ever heard of as a kid was to cut the 'exhaust pipes' off a VW inlet manifold to gain 3 hp

[gruntguru]

I realise that , and realistically the mixture needs to be cooled again before reaching the combustion chamber, but that means a long, probably serpentine, duct. Also, I suppose there is a chance, with the cooling, for precipitation, or whatever the engineering term is for a return to a snot-like fuel/air conglomeration. Is this where direct injection wins? I am not familiar with the pro's and con's of this system, although it sounds logical.

Certainly can't have longer ducts - the runner length is a critical part of the power equation. (No doubt you knew that)

Direct injection has the potential advantage of charge stratification, putting the fuel where it is most useful - away from crevices and chamber walls. It also permits high levels of scavenge without losing fuel in the process. Lots of benefits for road cars in the areas of fuel efficiency and exhaust emissions. On the downside it is more difficult to utilise all the air in the chamber without fairly high levels of turbulence, tumble and swirl. Tumble and swirl are bad for VE because the energy required has to come from the intake flow. The other way to produce turbulence is squish which also can't be increased beyond a certain point without compromising flow through the valves.

[McGuire]

So it's really the passing over the edge of the valve that does the mixing eh?

Also, we're mostly talking about road cars here, right? In the pics it doesn't look like there's much space to re-form from a mist into a string of snot.

The application really doesn't matter. Where you have gasoline traveling in air you will have the same essential behavior, differing only in degree. I can't show you F1 or anything else current, but I can show you stuff from a previous thing that is no longer sensitive. Here the "fuel" is fuel sytest fluid with UV dye added, functionally identical to gasoline except non-flammable. This is typical air-fuel flow:





This is the chamber end. See the strings?


Here is an interesting experiment in which eight extreme high-pressure nozzles were installed in an intake bell. See the eight discrete streams headed off into the distance? The interesting part: at the intake valve there are eight discrete streams...

Edited by McGuire, 27 June 2009 - 03:58.

[cheapracer]

So your saying the fuel likes to bond together and doesn't like to atomise?

Edited by cheapracer, 27 June 2009 - 04:05.

[OfficeLinebacker]

So you're saying the fuel likes to bond together and doesn't like to atomise?

I think that's exactly what he's saying.

[OfficeLinebacker]

I saw this video several days ago and after McGuire's images I figure some video warrants a response in kind.



I think it's interesting how it looks like a cloud forms over the entire bank of trumpets.

[gruntguru]

I saw this video several days ago and after McGuire's images I figure some video warrants a response in kind.



I think it's interesting how it looks like a cloud forms over the entire bank of trumpets.

That is the "standoff" referred to in the "balance tubes multiple carbs" thread. It is caused by extreme reversals of airflow in each runner as negative and positive pressure waves race up and down in the runner - bouncing off the ends. F1 engines make maximum use of these pressure pulses to "supercharge" the cylinder at certain rev bands (usually just before the power peak and also at one or two lower rpm's) with the arrival of a positive pressure wave at the cylinder just before the intake valve closes. The rpm where this occurs is determined by the runner length. The standoff is fuel spray ejected from the trumpet and is sucked back in very soon after.

Here's another interesting video many of you will have seen - a camera in the combustion chamber. Definitely not a F1 - a 2 valve something or other (McGuire?) probably carbureted. Note the areas where excess fuel resides and the flame is very lazy and late - esp' around the intake valve seat.

Combustion chamber video

Edited by gruntguru, 27 June 2009 - 05:53.

[gruntguru]

So your saying the fuel likes to bond together and doesn't like to atomise?

If two or more droplets collide they will merge and unless the energy of the collision is very high, they will remain merged - held together by surface tension.

[NRoshier]

Duckworth did his masters thesis on port barrel throttles (PBT) and some years ago I contacted him about this. His view was that there was no performance advantage over a 'well designed' (this was not stated) throttle plate system in most circumstances, but there was the potential for minor part throttle economy gains due to the ability of the PBT to bias flow in the port. Later research for some information that seemed to indicate that at WOT the turbulence in the intake airstream can last around 10 diameters of the throttle plate. In this respect the PBT may well have an advantage of tradition throttles if packaging forces the throttle to be close to the valves.

[Tony Matthews]

Fascinating! Thanks NR

[gruntguru]

Later research for some information that seemed to indicate that at WOT the turbulence in the intake airstream can last around 10 diameters of the throttle plate. In this respect the PBT may well have an advantage of tradition throttles if packaging forces the throttle to be close to the valves.

The Cossie in post 1 has short runners - the whole thing is only about 5 diameters long!

[McGuire]

I saw this video several days ago and after McGuire's images I figure some video warrants a response in kind.



I think it's interesting how it looks like a cloud forms over the entire bank of trumpets.

...except that if you look closely, you will see that the cloud is not all that cloudy. It's especially visible in the first (longitudinal) view: in the reversion aka standoff "clouds" over each cylinder you can see the afore-mentioned snot-strings of fuel building brom the bottom up, forming and growing with the throttle open and then collapsing as the throttle is closed. First 20 seconds of video, cylinders in left of frame. Spike-shaped, in the center of the inlets -- can you see them? With side lighting and slow-mo they would be even more evident. And these are forming in the standoff columns. Imagine what the main fuel charge is doing.

I tried a screen capture and didn't quite catch it at peak but this will show you where to look:





The other noteworthy thing about this video is running the engine without the airbox, obviously done for visual effect but I hope someone was standing just out of frame with a fire extinguisher. If a few drops of fuel had splashed the wrong way, then we would see a half-dozen Frenchmen running around in sheer terror... which would make great cinema, come to think of it. Better than anything their film industry has done in years.

[OfficeLinebacker]

Thanks McGuire. I am going to have to say for others though that part of what makes that image is the top edge of the throttle plate. You can seen the very tops of them swing into view on a couple of pulls when they actually open the throttles gradually. It seems that the injectors end up kind of pointing at the underside of the throttle plate at WOT, though that might be an illusion.

Also, it's interesting, it seems Renault are the only ones that test their engines with airbox off; of all the "Formula One Engine" clips I found on YouTube only the above and this one, which shows much briefer clips of the injector nozzle area, are sans airbox.

http://www.youtube.c...feature=related

I apologize if I've added confusion since the OT of this thread is cylindrical throttle bodies and I've posted two video clips of engines with traditional butterflies.

[Tony Matthews]

I apologize if I've added confusion since the OT of this thread is cylindrical throttle bodies and I've posted two video clips of engines with traditional butterflies.

Where would these threads be without a bit of added confusion, I'd like to know!

[The_R1_Kid]

Duckworth did his masters thesis on port barrel throttles (PBT) and some years ago I contacted him about this. His view was that there was no performance advantage over a 'well designed' (this was not stated) throttle plate system in most circumstances, but there was the potential for minor part throttle economy gains due to the ability of the PBT to bias flow in the port. Later research for some information that seemed to indicate that at WOT the turbulence in the intake airstream can last around 10 diameters of the throttle plate. In this respect the PBT may well have an advantage of tradition throttles if packaging forces the throttle to be close to the valves.

thanks for the information

[Tony Matthews]

thanks for the information

Er, does that sound familiar to anyone?...