21 replies to this topic

[Moon Tricky]

Since the subject came up elsewhere, I took the liberty of starting a new thread on the subject, so now it can be "On Topic".

Here's my suggestion: 200k gif anim

Observe how in the fully open position, there is no restriction to flow at all as it simply forms a straight tube. I don't know what the flow characteristics would be like in partially open positions though...

[J. Edlund]

If you take a look at a few F1 engines you will notice that some of them use this type of barrel throttle while others prefer the butterfly.

Here is another method that has been common on touring car engines
http://e30m3performa...lides/index.htm

[Moon Tricky]

Quote

Originally posted by J. Edlund
If you take a look at a few F1 engines you will notice that some of them use this type of barrel throttle while others prefer the butterfly.

Here is another method that has been common on touring car engines
http://e30m3performa...lides/index.htm

365HP is quite impressive... is there any simple explanation as to why it gives poor control at part throttle?

[Ninja2b]

Moon Tricky: thats pretty much exactly the arrangement we came up with. At part throttle our CFD sims showed that it was poorer than a butterfly (that was at 45 degrees) but I don't have the figures here to tell me how much it was by. We didn't make a performance map (to my knowledge) so the part throttle performance from the dyno wasn't really something we looked into. As the test driver, I can say that I didn't notice any reduction in drivability compared with the butterfly setup.

In "V10 Formula One Engine Technology", Ian Bamsey states "When the Cosworth ED engine was converted from slide to barrel throttles, Cosworth engineer Mark Parish had some interesting observations on the effect they had. "This seems to be the reason for the enhanced drivability of the latest specification. However, the transient effects of a barrel throttle are something we can't see on the dyno, even though we have a transient dyno".

He also says "Compared to a barrel throttle, a butterfly is less prone to seizure and was a well proven option, that some felt offered enhanced control. The barrel throttle tended to be a more compact option, one that arguably worked just as well"

Both these quotes from p54.

[McGuire]

In around 1915, Harry Miller's carburetors employed a barrel-type throttle valve, and an array of fuel jets lined up in a row (like a pan flute) so that as the throttle was opened, the valve progressively uncovered the jets one at a time.

[McGuire]

Quote

Originally posted by J. Edlund
If you take a look at a few F1 engines you will notice that some of them use this type of barrel throttle while others prefer the butterfly.

Here is another method that has been common on touring car engines
http://e30m3performa...lides/index.htm

Along with non-linear effective opening rate, another drawback of the slide throttle is man do they like to stick. Even a dead bug can hang them up. Note in the photos how in this case, the edge of the slide has been sharpened like a knife in order to guillotine anything that finds its way in there to gum things up. Also note the big heavy return spring employed for the same reason. Put them both together and you do not want to stick your finger in there.

[Modulus]

I understood this type of barrel throttles to be universal in Formula One. The emphasis is on full throttle power means that barrel's are an easy choice. With the reduction to V8 engines, cars spend even more time at full throttle per lap, swinging the advantage further towards an unimpeded intake flow.

Part throttle performance is obviously not ideal, however, the gains must outweigh the problems. I'm speaking without seeing any emperical evidence. Please set me right if I have said something in error.


This is the first time I've seen slide throttles. I can imagine lots of stiction problems with large flat surfaces passing over each other. This wouldn't be so much of an issue with hydraulic throttle systems as used in F1 as they are very powerful, but I feel it would be difficult to package a sliding system in the Centre V of an engine. This space is usually entirely consumed with the airbox. Obviously with the BMW DTM photos (found with the link supplied in the 2nd post) the engine is an Inline-4 and the actuation seems to be longitudinally along the head, so such packaging issues won't be encountered.

[rms]

We really need linear power delivery v accelerator position.
Slide assy's can stick but given adequate maintenance and air filtering are no more prone to sticking than any other option.
Barrel assy's are a more compact option but along with butterfly's are more difficult to manufacture.
Slide assy's have been used on many formulae over many years.
Any progression problems are more a matter of engine mapping.

[McGuire]

Quote

Originally posted by Modulus
I understood this type of barrel throttles to be universal in Formula One. The emphasis is on full throttle power means that barrel's are an easy choice.

Not really. Actually it's more of a necktie-width issue.

Get an engine and a dyno. Place the engine on the dyno. Install two pressure gauges, one each tapped into either side of the throttle valve. If the two gauges read identical values with the engine running at max hp@rpm at WOT, you are good to go. A throttle doesn't get any better than that at WOT, and this can easily be achieved with a butterfly, a barrel, a slide, or a cast-iron steam valve if you please, if its effective flow rate is appropriate for the application.

So choose the type of throttle valve you like for whatever reasons you like, including part-throttle response, driveability, packaging, cost, contingency award eligibility, or your own personal sense of aesthetic satisfaction. But ironicallly, one reason that is not valid is max flow at WOT, because there it makes no difference what type of throttle you select. What matters there is getting the size correct.

[Moon Tricky]

McGuire, surely a butterfly valve creates turbulence which makes things different? If it's only a question of flow area surely it doesn't make any difference even at part throttle, other than potential non-linear response, as long as your valve is big enough for maximum flow at WOT.

And why have BMW even bothered to design an engine that throttles using variable valve lift, if the butterfly valve is just as good?

[shaun979]

Quote

Originally posted by McGuire


Not really. Actually it's more of a necktie-width issue.

Get an engine and a dyno. Place the engine on the dyno. Install two pressure gauges, one each tapped into either side of the throttle valve. If the two gauges read identical values with the engine running at max hp@rpm at WOT, you are good to go. A throttle doesn't get any better than that at WOT, and this can easily be achieved with a butterfly, a barrel, a slide, or a cast-iron steam valve if you please, if its effective flow rate is appropriate for the application.

So choose the type of throttle valve you like for whatever reasons you like, including part-throttle response, driveability, packaging, cost, contingency award eligibility, or your own personal sense of aesthetic satisfaction. But ironicallly, one reason that is not valid is max flow at WOT, because there it makes no difference what type of throttle you select. What matters there is getting the size correct.

Do you think it might not be a matter of just going bigger, but similar to port design where essentially you want maximum flow efficiency - required flow at minimum cross section to maximize velocity, and maximize pulse and ram effect? Same efficiency goal with bellmouths.

The way a lot of 4V race engines are designed, the portion of the tract all the way up to stack entry can be considered part of the port.

[Greg Locock]

Datapoint: the entire pressure drop on a /production/ intake system, up to a point about 4 inches short of thevalves, is less than 1 kPa, ie roughly 1%. That includes the butterfly, air filter, and a lot of duct.

So, unless we are arguing about how many angels can dance on the head of a pin, this is small beer.

[Ninja2b]

Quote

Originally posted by Greg Locock
Datapoint: the entire pressure drop on a /production/ intake system, up to a point about 4 inches short of thevalves, is less than 1 kPa, ie roughly 1%. That includes the butterfly, air filter, and a lot of duct.

So, unless we are arguing about how many angels can dance on the head of a pin, this is small beer.

I have reviewed a paper by Taylor published at the SAE international congress in 97 which would claim otherwise. Although the ports and beyond account for the vast majority of the pressure loss, the loss in the other regions is more than 1% - for high valve lift cases I believe the plenum losses were around 10% of the total loss in the system (although this was much lower for medium and low valve lift). The losses in the valve clearance region were by far the largest contributing factor.

Someone posted a link to a paper (which I still haven't gotten round to reading) which claimed that the flow disturbance from the butterfly had all but disappeared 10 diameters downstream of the throttle. Perhaps the prevalence of barrells (relative to road cars) in F1 is more to do with a combination of this fact combined with exceptionally small runners. 10 diameters downstream when we are talking a large diameter inlet with a very short length might = non-ideal flow into the ports, making barrels better than larger butterfly's (butteflies???) with a comparable flow area.

[McGuire]

There is a presumption that turbulence around the throttle blade and shaft is automatically a bad thing. Actually the edge of that throttle blade makes a very nice fuel shear point.

[Fat Boy]

Quote

Originally posted by McGuire
There is a presumption that turbulence around the throttle blade and shaft is automatically a bad thing. Actually the edge of that throttle blade makes a very nice fuel shear point.

The blade can also be used to turn the flow into the port. One engine I worked with had 'full-throttle' at 86-87 deg. This made a gnats dick more power than completely parallel to the flow because it turned the flow a small, but useful amount.

Turbulence is also a poor choice of wording. Damn near every bit of fluid flow that we talk about when having a discussion about racecars is turbulent (as opposed to laminar). Turbulent flow is not necessarily a bad thing. Unsteady, chaotic flow is a different matter, and that's what raises concern.

[Greg Locock]

Sorry that was poor wording on my part. The total drop in pressure across the intake system at full throttle is about 1 kPa, leading to a drop in VE of 1%.

What proportion of that is due to the butterfly valve, I don't know.

[Moon Tricky]

How about throttling with a variable speed supercharger? If you could reduce the speed for lower load conditions, and maybe open a bypass valve as well, you could go from 100% power down to 50% by going from 1 atm of boost down to naturally aspirated. It would help to have variable compression ratio as well though, otherwise the efficiency at low load wouldn't be so good. Once you'd got down to 50% power, normal throttling could commence, perhaps using variable valve actuation.

[J. Edlund]

Quote

Originally posted by Moon Tricky
McGuire, surely a butterfly valve creates turbulence which makes things different? If it's only a question of flow area surely it doesn't make any difference even at part throttle, other than potential non-linear response, as long as your valve is big enough for maximum flow at WOT.

The butterflies used in F1 applications usually have a nice aerodynamic shape. The small area in the flow path taken up by the butterfly can also, easily, be compensated for by making the diameter of the pipe around the butterfly a bit larger.

Quote

Originally posted by Moon Tricky
And why have BMW even bothered to design an engine that throttles using variable valve lift, if the butterfly valve is just as good?

Because they want to reduce part throttle pumping losses. Racing engines don't spend much time on part throttle so part throttle losses isn't really an issue for a racing engine.

[Moon Tricky]

Quote

Originally posted by J. Edlund
Because they want to reduce part throttle pumping losses. Racing engines don't spend much time on part throttle so part throttle losses isn't really an issue for a racing engine.

No, what I don't understand is, if the nature of any obstruction at WOT doesn't make any difference to performance as long as the flow area is big enough, why does it make any difference at any other throttle position?

[NRoshier]

Quote

Originally posted by Ninja2b


Someone posted a link to a paper (which I still haven't gotten round to reading) which claimed that the flow disturbance from the butterfly had all but disappeared 10 diameters downstream of the throttle. Perhaps the prevalence of barrells (relative to road cars) in F1 is more to do with a combination of this fact combined with exceptionally small runners. 10 diameters downstream when we are talking a large diameter inlet with a very short length might = non-ideal flow into the ports, making barrels better than larger butterfly's (butteflies???) with a comparable flow area.

The '10 diameters' rule is not in that article, but is instead a rule of thumb (given to me by an engineer working solely with turbulent flows) for establishing original flow conditions post disruption in turbulent flows. However there are so many variables in the intake tract of an engine that such rules of thumb are useful to consider but should be taken with a grain of salt.
Fat boy is absolutely correct too that we hope to have well developed and stable turbulent flows instead of chaotic...laminar is not possible as the intake port Re is far to high.

[J. Edlund]

Quote

Originally posted by Moon Tricky


No, what I don't understand is, if the nature of any obstruction at WOT doesn't make any difference to performance as long as the flow area is big enough, why does it make any difference at any other throttle position?

BMW's system isn't about perfomance, it's about pumping losses. If you employ low lift/early closing of the intake valves you can achieve the required throttling while reducing pumping losses at part load.

[Halfwitt]

There are advantages and disadvantages to both. Barrels have in the past been maligned as prone to hesitation, especially where fuel is sprayed from above. Fuel can tend to 'puddle' above the barrel at small openings, giving a few lean cycles, followed by too much fuel as the puddle disperses, then as the mapping tries to sort itself out, another few lean cycles. This clearly isn't such a great problem, as most F1 teams run barrels. Some other lower revving engines use them and, having longer inlets, are able in some cases to position the injectors below the barrel.

On F1 engines with big bores and small bore spacings (8.3mm between pistons as a minimum under the current rules), barrels tend to have long but narrow ports, in order to keep the weight of the barrel small. Some employ a double counter-rotating barrel to minimise weight and some perceive this to give some flow advantage too.

Others, possibly only one engine manufacturer in F1 today, continue with butterflies. These are less affected by weight considerations as they are inherently much lighter than barrels, and so can have a more optimal shape to the port, but nothing so simple as the Asiatech throttles shown above.

Slide throttles are felt to be difficult to make (they need to be very flat), set up (clearances and controlling friction) and maintain (wear). They fall somewhere between barrels and butterflies in terms of weight, but nobody has used them for a number of years in F1 (Judd GV perhaps) due to their shortcomings. A major advantage over barrels is the short length they occupy within the port.