51 replies to this topic

[Canuck]

This is likely another one of those simple questions I'm just not seeing - perhaps I'm not drinking enough water or eating enough fish today...

The statement was made that "Linear valve speed is proportional to piston speed x bore stroke ratio divided by square root of the number of intake valves/cylinder.

I understand the statement, I don't understand why. Certainly valve size is proportional to bore, and lift is proportional to valve diameter, but...if I take an existing engine, bore it .5" (it's not that outlandish, 883 to 1200 H-D Sportster conversion), substitute proportionaly larger valves...oh hey.... fog lifting slightly and adjust my lift to .25 new valve OD then my valve train linear speed does change. And of course with more valves, they're smaller OD so their respective lift is reduced, reducing their linear speed compared to 1 valve.

Right?

While I'm on valvetrain topics, I recall reading a thread that discussed both length and diameter of the intake port. Initially I thought it was the HRM thread but the only formula I could glean was 86,000/Desired RPM = length. I've not been able to find the other thread, can someone refresh my memory?

And how about that .25 x valve diameter? I typically run anywhere from .29 to .32 because, that's how the cams are ground. Typically the larger the lift, the longer the duration, but my favourite cam is a short duration, big lift that results in a .31 ratio. However, the cam works VERY well if a bit hard on valve train.

[Joe Bosworth]

Canuck

The formula for the relationship of linear gas velocity, cylinder volume and RPM that I have been using since the 1950's is, (for metric calculations):

RPM = 23.275 * gas velocity * (valve diameter mm *.9)^2 / cc per cylinder

I find that gas velovity of 80 m/sec +/- a little bit keeps coming up as optimum gas velocity from the early days of Coventry Climax race engines through Cosworth and on to the latest Honda CBR1000RR.

I take 90% of valve diameter because it is really port velocity that seems important although valve diameter minus 1 mm is also sound depending on valve size being analysed.

For a 45 degree valve seat there is no improvement in gas velocity beyond opening 25% of valve diameter though your observation that most good valve grinds go out to 28 or 29% because it allows optimum gas velocity to be acheived for a longer percentage of time through the valve opening cycle.

Regards


Joe B

[Joe Bosworth]

Canuck

Oh by the way, for a four valve cylinder use half the cc per cylinder!!!

Regards again

[Canuck]

I'm sorry Joe, you've lost me a bit. So, what you're saying is rather than approaching lift as a derivative of valve size, lift becomes a function of the desired gas speed?

[Joe Bosworth]

Canuck

Let me fully explain since there seems to be a small bit of confusion creeping in.

Let me start with where I am coming from. I started modifying engines circa 1949 and continue to this day. By about 1955 I was chasing national championships or such as they existed at that time in drag racing. Chasing performance I was changing a cam shaft a week at times trying to find a "sweet spot".

At that time I had been exposed to very good university courses in therodynamiccs, fluid flow and machine design that augmented a bit of practical experience in engine building. Lest this lead you into believing that I am a dynasoar I am still building the odd good performer or two.

On of my early discoveries was that a good performing engine had to match components that all were aimed to a single destination. No sense in having 5500 rpm ports, a 6500 rpm induction, a 7000 rpm camshaft and a 5000 rpm exhaust all driving a gearing that never peaked.

A second discovery was that it was far cheaper and quicker to work from the inside out on an engine.

I also started a life long quest for engine data from those far smarter than me to find out what really was working in practice.

One of my early discoveries that remains verified to this day is that almost every really good engine finds a port gas velocity of 260 fpm - 80 mps +/- a little bit just behind the valve at best HP.

Port for that and you won't go wrong - hence my earlier message.

Then select a cam shaft that meets your criteria for valve lift - hence my message in further detail. Valve timing is another subject!!

Working outwards you select suitable carb and venturi sizes, carb inlets, exhaust diameters and exhaust lengths and voila, you have an engine that makes good sense. All that assuming that you set sensible parameters in the first place!!! These details are easily found if you know where to look.

I hope that this diatribe clears the air for you.

Regards

[Canuck]

Originally posted by Joe Bosworth
there seems to be a small bit of confusion creeping in.

That's generous to say the least

Originally posted by Joe Bosworth
Canuck
... Lest this lead you into believing that I am a dynasoar I am still building the odd good performer or two...

I also started a life long quest for engine data from those far smarter than me to find out what really was working in practice.

Regards

Despite what seems to be my nature here, I come with hat in hand to sweep the floor for free so to speak. I'm not dumb and know my way around an engine, but I'm ignorant (as in lacking knowledge). Dynosaurs like yourself and McGuire , and the rest of the incredibly smart folks are why I come and ask questions. It's refreshing to be exposed to knowledge with more depth than a magazine article. All the theory in the world is nice, but without experioence to back it up, it's just theory, not power. The in-depth knowledge brought about by 50+ years of doing what your doing is not to be discounted. I am extremely grateful to individuals like Greg Locock, McGuire, Fat Boy, Engine Guy and even the paranoid PhantomII ( ) for taking the time to correct me and guide me along the path. Far and away the greatest benefit coming here is nothaving the answer handed to me on a platter, but being told where to find and how to learn it. I have a great deal of respect for the knowledge in here, despite however I may come across. And that's enough ass-kissing for fore-seeable future.

Then, to get my optimum valve size, I would re-arrange your formula to:
Valve Diameter = [SqRT{rpm/(23.275*80/cyl cc)}] * 1.11

If I use numbers that I know - Indian stuff, I get the following:

Valve diameter - 1.94"
Cyl. Volume - 50.12"^3

This puts the RPM for 80m/sec at 4400. So then, if I want maximum RPM to be 6000, I then need a valve that is 57mm to feed the same cylinder at the same velocity correct? I don't want to sound ignorant (though it's not new), but that's a 2 1/4" intake valve which seems awfully large (though still within the realm of .5 - .65 x bore diameter.)

You mentioned that you built drag engines. Assuming that we're talking about a street or road-course car, how does that influence the RPM design?

[Joe Bosworth]

Canuck

I was hoping that others would pipe in by now. If they don't maybe we should go private for the discussion so as to not bore others to death.

Yep, you are using the equation correctly and 2.25 inch intakes calculate to a rev of 4400 for max HP on your 401 cu inch engine.

However 2.25 is really quite small for your application. For years 2.02 was standard for anyone building a 5 liter Chev. That calculates to 6350 rpm. we are now finding those who once standardised on 2.02 are now using 2.08 's. That brings us to 6735 rpm.

your 401 cu inch needs to be using big block type technology if you are looking for the best. I guess best has to be defined though as everyone has their own idea of what they want to accomplish.

My definition is what works best on the track.

Though I started drag racing and still keep tabs on what the guys are doing there I moved to sports car then open wheelers very early. Except for a couple of tries of roundy-rounds in another's car.

It is interesting that road course, stock body drag and short track engines all very nearly share technology. Their needs are very similar needing good torque after shifting in the case of drag and road engines and getting off corners in short track.

I can be reached privately at doveratarach.net.au

Regards

[Ray Bell]

Originally posted by Joe Bosworth
I was hoping that others would pipe in by now. If they don't maybe we should go private for the discussion so as to not bore others to death.....

No, Joe, there are readers here even if they're not adding their confusion to the discussion...

So what valve size will the Poly need to optimise a 4" bore, 3.91" stroke and an intended safe rev limit of 6,500rpm?

[hydra]

Just to add my $0.02, Intake velocity can be between 80-100m/s @ peak power, depending on how peaky/torquey you want the power curve to be, and on how efficient the intact tract is. The more efficient your intake is the higher the "sweet spot", i.e. you trade off less peak power for midrange power when you increase velocity.

Also, I'm inclined to think that the reference area used should be the average area from bellmouth to valve head, or simply intake tract volume divided by length. This can be anywhere between 0.8-1.05 Dv (the latter being for a racing 2v engine)

[hydra]

Another thing Ray and Canuck,
Valve size is only a very small part of the overall picture, you need to be looking at the intake tract as a whole.

Using the 80-100 m/s rule of thumb for a 2v engine we end up with

Di = B(sqrt Vp)/(sqrt 80-100)

Where Di = average intake diameter, B= bore, Vp = piston speed in m/s, and 80-100 is our target intake tract velocity.

now, optimum Di/Dv (Dv = valve diameter) should vary with L/Dv ratio, valve seat angles, port efficiency, and a bunch of other stuff. Its all very application specific really, but the sweet spot is typically around 0.85 or so...

[McGuire]

What Hydra said.

80 m/s is not terribly efficient IMO...depending on the methodology I would like to be more in the neighborhood of 100 m/s and 120 m/s on the exhaust... and some applications are more velocity sensitive than others... are we talking about 2V or 4V? IR or plenum? What is the target use? Road or track? Indy 500 or Lime Rock? ...and of course here the figure is just a reference value in a calculation... if the value tracks a given approach that's fine and if it achieves the desired result even better.

In an engine there is no such thing as mean or steady-state airflow velocity anyway. Really we have turbulent airflow in finite amplitude waves, accelerating and decelerating every valve cycle from zero effective velocity closed to 300-400 fps in the valve bowl at lift... bumping Mach 1. The column of air is like a big old spring. A zero-dimension calculation is a very rough guide... after all, even the most sophisticated modeling software provides at best an approximation.

As Hydra said, valve diameter/curtain area is just one piece of the puzzle.

[hydra]

Thanks for the heads-up McGuire, it really means a lot to me coming from you


Another thing I'd like to add, and I've had several people disagree with me about this before, is that Valve lift is a GOOD thing! Allow me to rephrase, one should aim to maximize an engine's L/D ratio as much as possible within valvetrain (geometry and longevity) restrictions. Better yet, one should aim to optimize the intake tract to make use of high L/Ds by favoring the high-lift flow regime over low-lift (using relatively large valve bowls and steep valve angles). A very respected race-engine builder mentioned that the ideal L/D for an all-out performance engine is around 0.41 for 2V, and 0.36 for 4V, and everything I've seen or read tends to agree with that...

Another thing is it allows you to use less duration/overlap for a "fatter" torque curve, albeit at the expense of valvetrain stability

[McGuire]

Valve lift is good. I think confusion arose only because there was a time when many engines could only handle so much... back in the day there were some applications where more than .620" or so would only bend the hardware... for example the rocker stud bosses in the head would flex. So naturally the more aggressive lift profiles would make less power and break a lot more parts, and there were all sorts of valvetrain harmonics that could be fixed by backing off on max lift. There were no good valve springs by modern standards and anti-harmonic flank profiles were a cut-and-try type of thing. But today the materials and the technology are so much better. 9500 rpm all day in 2V NASCAR engines... that's how they did it.

[Joe Bosworth]

The discussion is starting to get more interesting with more comments coming in. I’ll re-confirm that my years of compiling data keeps coming back to the 80 m per sec average inlet gas velocity in the port immediately behind the inlet valve seat. I started to keep this data because it has been relatively easy to get data on inlet valve diameter while it is really tough finding meaningful data on port shapes and lengths.

The most recent information piece I have is for the Honda CBR1000RR which of course is the basis of the Honda superbike. I have to presume that Honda has designed without restrictions and they end with nearly 160 BHP per liter, (real and measured independantly) and they conform to the 80 mps I quoted in the second reply to this thread.

It is interesting that almost every un-restricted engine that I have data on stays under 85 mps and most very close to the 80 figure. These include quite a wide range of formula 1 engines although I freely admit that such engine data for the last 10 years or so have been tough to find.

I readily admit that there are some engines that are restricted by regulations that go to 100 mps or even slightly more. Certainly, the most dramatic of these will be the Nascar engines. I have never had a peak inside of a Nascar engine and would tear down doors and walls to have the opportunity. I am sure that they are marvels.

Working forwards from what I do know I suspect that they are running inlet valves of between 2.15 and 2.2 inch diameter and are undoubtedly runing inlet velocities of 100 or slightly more. I suspect that they are also running lifts of about 36% of valve diameter.

Interestingly they are “only” pulling about 130 to 135 BHP per liter. This being imposed on them by regulations that among other things limits their single 4 barrel carb throat diameters severely. Their big oval engines also don’t have to be overly flexible and I suspect that they really aren’t. I strongly believe that their short track engines will be quite different again.

These regulations mean that they have to be very very inventive on how they make power and revs and they clearly succeed within their limits.

It would be great if some of those in the know would/could publish some real data on Nascar engines and/or others that make serious HP per liter at high gas flows.

[Canuck]

Originally posted by hydra
Another thing Ray and Canuck,
Valve size is only a very small part of the overall picture, you need to be looking at the intake tract as a whole...now, optimum Di/Dv (Dv = valve diameter) should vary with L/Dv ratio, valve seat angles, port efficiency, and a bunch of other stuff. Its all very application specific really, but the sweet spot is typically around 0.85 or so...

Right. See, I know that I don't know how to sort that out. At least I know that I don't know. In trying to take things a piece at a time, it's rather apparent that the entire intake tract as a whole must be addressed at once.

I'm awaiting the arrival of Scientific Design of Exhaust and Intake Systems but I suspect like Taylor's work that while full of solid information that I don't want to miss, is still half a century behind current knowledge. Can anyone recommend further reading on this topic?

[desmo]

Here's a graph output from WAVE for port velocity at 18krpm that I was sent, using assumptions that are pretty generic for F1 engines, can't recall the specifics I posted them once here.

http://members.atlas.../F1intake-V.bmp

The simulataneous differences between velocities at different port loci hinting at the pressure gradients locally inside the port and the flow reversals at 79mm really jump out at me. As well as underlining McGuire's point that it's really all transients in practice. A mean flow velocity misses a lot of the picture.

[Canuck]

2.9 and 79mm what?

[desmo]

Sorry. Those are the distances from the valve seat.

[Joe Bosworth]

guys

Thanks for the additional info.

I have integrated the inlet velocities from Desmo's curves and, voila, the curve for closest to the inlet valve for the period over the 0 to 180 degrees of the inlet stroke averages 77 mps, which is directly similar to the 80 mps average that I have found for all F1 and high performance engines since the FWA Climaxes.

This has been consistent through data collected for the 1.5, 2.5, and 3 liter formula ones as well as most motorcycle and othe race engines for a long- long time. The exceptions have been engines set by regulations that limit valve diameters by various means such as Nascar; these push towards 100 mps by necissity and innovation that don't necassarilly end up with efficient race engines, only ones that work well within the constrictions.

Interestingly, the numbert stays consistent for my latest, CBR1000RR Honda as well as the second most recent, working on a Matchless 500 race bike

The problem that I have found using computer programmes such as Wave quoted earlier is that they provide for too many variables and nom base lines to work from.

My quoted 80 mps sets a base line for port and valve diameter from which you can use your neat programmes to optimize for inlet and exhaust lengths and cam timings to end up with well matched systems.

I have found, not surpriseingly, that the best results come from matching all of the elements towards a single objective.

Rgds

[Canuck]

Originally posted by desmo
Sorry. Those are the distances from the valve seat.

Given that there is a pressure wave (as McGuire noted), wouldn't it stand to reason then that an inlet port with pressure sensors along it's length, they resulting information would represent the motion of the wave? We know (even I know ) there is a pressure wave in the tract, so the it should stand to reason that there are going to be differences in pressure at different loci for a given CA...shoudn't it?

[Greg Locock]

Sure. Use a microphone to plot them out. Mind you, interpreting the results is a bit of a brain tease.

[Canuck]

Hey Joe - your figure of 80m/sec - I'm a little fuzzy there. I mean, I understand the formula you gave and follow our discussions, but I'm a bit lost on the 80m/sec.

I wanted to understand the potential power difference if one took an engine with 2-valve heads and used a 4-valve design, so I 'wrote a paper' for myself based on the following: According to Taylor, if Z > .6, Ve begins to fall so if we use Z=<.6 as maximum, you can determine maximum piston speed using the formula Z=(Ap/Ai) x (s/Ci x a) where Ci is valve flow co-ef, and a is velocity of sound. Further in Taylors discussion on Z it's stated that values between .4 and .6 were shown to be optimum. If a = 366m/sec that puts 80m/sec well below .4, down around .22

Am I missing something?

[hydra]

Z represents what is called the "inlet mach index", or "gulp factor", and isn't at all representative of the actual Mach number, whether peak or average. Therefore, you can't use average port velocity/a to find Z.

Just for the record, Z is usually around 0.4 for a peaky, all-out race engine, and between 0.45-0.5 for a decent street machine.

[Canuck]

Ah - that's in essence what I was wondering - Z is in fact a different measure than the gas velocity figure in the formula Joe posted. So what then is that difference? Mach Index is a still a relevant calculation or measurement and it certainly appears the 80m/sec is also relevant. Is it an actual measured quantity or a representation of relative mean speed or ?

Saying Z isn't representative of the actual Mach number...if Z=.6 is not 60% of the speed of sound in the intake gases, what is it?

[hydra]

"Z", as given in your formula, is an imaginary construct. Z=0.6 does NOT mean that the actual mach number is 0.6.. Sure its very relative to engine performance, but it has no physical real-world equivalent. It is totally unrelated to average port velocity, as you can have two engines @ 80m/s, but if one breathes better than the other (higher Ci), the other will have a higher Z, higher pumping losses, and will make less power, all else being equal.


To make things even more confusing (I didn't want to go there but you asked for it), ACTUAL PEAK Mach number does have real world significance. Basically it is a measure of how restrictive the intake system is. The lower the overall pressure drop, the higher the peak Mach number that can be attained. This directly translates to increased volumetric efficiency. State of the art racing engines can reach a peak Mach number of 0.55 (this is on a live running engine) and a VE of 125+%

This isn't something I would worry about, as it only enters the equation when you're doing high-end engine development, not building a weekend cruiser...

[Canuck]

Originally posted by hydra
"Z", as given in your formula, is an imaginary construct.

Ah - excellent. Okay, I can work with that. The Actual Peak Mach doesn't cloud things - I get it.

So if I use the Inlet Mach Index formula posted above to obtain max. piston speed, and take that number and plug it into P = Ap x BMEP x s / 132,000 - am I still within theoretical reasonableness or am I blowing smoke?

[Joe Bosworth]

Canuck

To respond to your query of 9 August on the 80 mps issue:

The biggest value of using that formula is to make sure that you are coordinating port and valve sizes with RPM expectations. As I have mentioned a few times on this thread it has provided an accurate relationship between port/valve size and the RPM at which peak HP is found for virtually every high horsepower engine for which I have reliable data for about 45 years including F1s through Superbikes and road and drag engines.

Certainly no use in finding out what HP you are going to make but it sure stops one from porting or camming for 8000 RPM when you have no chance of getting there for other reasons; and vica versa.

When one has a suitable library of BMEP data It can be used to get yourself in the right HP range but that is another thread for another time.

As I think that I have said before, the very best engines are those that coordinale all of the inlet and exhaust elements to work in concert together. The 80 mps is just one of the tools that help you get there. Also very useful for narrowing the range of things that one can burn hours on when using any of the computer programmes that allow you to vary the elements, sometimes in non-validated ranges.

Regards

[Canuck]

Originally posted by Joe Bosworth
Canuck
As I think that I have said before, the very best engines are those that coordinale all of the inlet and exhaust elements to work in concert together.
Regards

Thanks Joe. I understand where I can use the formula you provided, I just wanted to understand why optimum Z and 80mps were so far apart. Hydra has done a nice job clearing that up.

Regarding your statement above - If I tuned my camshaft, intake and exhaust systems to peak at a theoretical 8000rpm, I would expect to see a very definite peak at 8000rpm. 99% of the motorcycle engines I build are street-oriented, even if folks fancy themselves as 'hard' riders. We know the vast majority of time in the city is spent between 2800 and 4500rpm and on the highway in an even narrower range between 3200-3800. Obviously I want to concentrate my peak torque in there somewhere, but what then is the best way to spread it out? Despite what they might say about the numbers, I'm pretty sure my clients would be happy to trade 5-10lb-ft of peak torque for 5-10lb-ft more average torque within that range.

[jdi]

Joe, under what conditions is this 80 m/s number measured? Flow bench at 28"?

Also, is there any difference in what you'd consider an ideal inlet velocity between naturally aspirated and forced induction engines?

[Joe Bosworth]

jdi and Canuck

Let's see if I can adequately respond to your questions:

For jdi; the 80 mps / 365 ft per sec is simply a derived value for the average port velocity over the 180 degrees of intake cycle. It is derived as per the formula that I previously published. I first hit on it years ago when I was building a lot of engines for drag and track use. Back then things were not as well documented nor were there the tools that are now available. But I did have access to a few race designed engines with known pedigrees. I started to look for common elements such as bore/stroke ratios, rod lengths, port configurations, ring widths, piston accelerations, max piston speeds, swept bearing areas and the like. Once started I have kept up the information search and compilation.

The 80 mps has been a constant over the years for virtually every high performance engine, (unrestricted by artificial engine spec rules), and one I have continued to use for the engines that I have built or advised on.

It has nothing to do with a flow bench. Interestingly, I have never been able to find a good correlation between flow bench numbers and where an engine makes power. May be that I haven't looked hard enough!! IMHO the bench's main value is in determining the value of doing things such as raising ports and configuring the turns within a port. I guess the first guys to profile valve seat angles also used a flow bench but I don't know that for sure as I was doing that in a less scientific way before flow benches came into existence as we know them.

For Canuck; There are motorcycle engines and then there are m/c engines. My interest in such is primarily track ready engines as opposed to those that have a primary street use. Having said that I get regular road use on a Ducatti 999, Yamaha R1 LE and a Honda CBR 1000 RR all with engines developed for superbike use with a hand full of simple mods, some of which have been applied to the above.

Also having said that I fully recognise that a decent race engine needs to have a suitably wide torque band to turn lap times. This is even more prevalent with M/C applications than race car applications as you can almost always stay in a torque band with cars these days while bike racers deliberately short shift and come out of corners below the point of max torque; sometimes much below. So a fat torque band is a good thing.

In my experience with 4 strokes the point of max torque is at about 80%, plus or minus a little bit, of the max HP point. There doesn't seem to be much that you can do to influence that factor. Interestingly, with 2 strokes you can lay max torque point right on the max HP point but that makes a very awkward bike to ride indeed; I did it once!

Yes most M/C are street ridden in the 3200 to 3800 rpm range. I'll ride a 2 valve Ducatti there all day just to sense the vibes but that is another story.

When we are talking such riding points we are talking riding at abot 30 to 50% of max HP point. This needs what I refer to as a fat torque band.

In my experience you can tune in a fat torque band without harming HP much by inlet and exhaust tuning. I first concentrate on header and exhaust pipe diameters, then lengths of ditto, then using a X pipe. On the inlet side carb size seems to have the most affect on fat torque though I have also used carb inlet pipes to affect torque points up and down with good affect.

Keeping my 80 mps gives you the HP point that you want then the above is used to influence the "fat".

Of course I did mention using the 80 mps to optimise your engine and optimisation is meant to be meeting the objectives of the builder. No sense in building a 8000 RPM screamer for a 2 ton road car.

Regards

[jdi]

Thanks for the explanation, I think I get it now.

[Ben Wilson]

Originally posted by Joe Bosworth
I take 90% of valve diameter because it is really port velocity that seems important although valve diameter minus 1 mm is also sound depending on valve size being analysed.

Joe - Under what conditions is diameter minus 1 mm sound? For any valve over 10mm it makes a substantial difference in the results.

[Joe Bosworth]

Ben

To give you any black and white answer would take an enormous amount of space to get into the detail.

Experience lets you you know when each alternative is appropriate.

Your example of a 10 mm valve is not really real unless you are talking some rather really small displacement 4 valve engines. I am half a world away from my long term records but I suspect that the 50 cc 2 valve Honda I was in to years ago was even bigger than this. Back calculating I suspect that the valve size was about 16 or 17 mm.

Part of the experience thing is knowing what you build vs what you think others have built when somebody gives you a valve size and you are working backwards to find what velocity they are using. Remember that race engines typically use pretty sliver wide valve seats compared to road based engines that are meant to last a couple of 100 k's.

Regards

[McGuire]

From what I have seen of recent wet flow research a lot of the current thinking will soon be headed out the window.

[Ben Wilson]

Originally posted by Joe Bosworth
To give you any black and white answer would take an enormous amount of space to get into the detail.

Any chance of a hint?

Don't get me wrong, I mean no disrespect here, I'm finding this discussion fascinating and I'm just trying to get my head around the intricacies of the information.

As you said a 10mm valve is ridiculously small, but it's the only value where your two equations correlate. I would have assumed that race engines where the valve seats don't have to last would have used thinner rather than thicker seats to capitalise on valve area, which would move the data further from the V-1 mm figure.

[Joe Bosworth]

Ben

Very hard to get less than valve diameter -1mm dimension for a port diameter when you are trying to get two valve seats diametrically opposite to one another plus a multi-angled transition from port throat to seat to outside diameter of valve.

Once upon a time we used a sharp edge transition from port diameter to valve seat to maximise port throat diameter for a given valve until we woke up to the fact that more flow resulted from a multi-cut transition.

Regards

[vvillium3]

McGuire,

What kind of current thoughts do you speak of going out of date??? Also, what advantages does wet flow testing have over the flowbench, and how did it lead to these new trends????

thanks,
Jason

[McGuire]

Not to be too oblique about it, I hope to say more later but the problem with dry flow testing is real flow is wet. Fuel is much heavier than air so when you add fuel to the air, the air suddenly takes a lot more work to move. Also, the A/F mixture is far less homogeneous than assumed.

[Greg Locock]

Once upon a time somebody made a perspex intake manifold.

Ever since then, everybody who saw the photos dreamt about cyclone separators.

[McGuire]

And then they discovered that as long as you didn't give it too much heat or vibration, you could run the SLA model on the engine just as though it were a real intake manifold. What fun. You can't see much but it looks cool.

[Canuck]

I've been trying to convince the owner of our machine shop that we really need an SLA/SLS system that can run both plastic and powdered metal. I haven't quite figured out what we need it for, but I have all sorts of ideas for my own projects.

[vvillium3]

not to pry, but if the air becomes much harder to draw efficiently into the motor. Then what other methods are there to make the mixture have more momentum to get into the combustion chamber??? More of a pour into instead of draw in situation?? I am getting at what you are hinting, or am I way off????
Also, McGuire. What do you do for a living???? Kind of new to this forum, haven't caught your profession yet...

jason

[hydra]

Originally posted by vvillium3
not to pry, but if the air becomes much harder to draw efficiently into the motor. Then what other methods are there to make the mixture have more momentum to get into the combustion chamber??? More of a pour into instead of draw in situation?? I am getting at what you are hinting, or am I way off????
Also, McGuire. What do you do for a living???? Kind of new to this forum, haven't caught your profession yet...

jason

Its called supercharging

[Canuck]

Originally posted by vvillium3

Also, McGuire. What do you do for a living???? Kind of new to this forum, haven't caught your profession yet...

jason

I KNOW I KNOW I KNOW!!! McGuire is the Crew Chief for Franklin R's Propster and full-time director of studies at the Prop-Derived Thrust Driven Supercar Design University

[NTSOS]

I KNOW I KNOW I KNOW!!! McGuire is the Crew Chief for Franklin R's Propster and full-time director of studies at the Prop-Derived Thrust Driven Supercar Design University.

Ah ha........just as I suspected!

I had a feeling that their supposed battles were pre-scripted and were designed to increase forum foot traffic in order to draw attention to their Propster project. They are in reality, really good friends.

OTOH.......Mac, does not the implementation of FSI circumvent the need for a wet flow bench?

I have always kind of had a slight problem with flow benches in general, as they are plugged into a wall socket as opposed to an engine.....I guess you just have to be carefull in interpreting the results and that probably comes with experience. I know, I know.....master of the obvious!

Juanito

[Canuck]

FSI? Fill me in?

I've often wondered - considering the engine's dynamic state and the express desire for a static state in the flowbench, how representative is it? OTOH, folks seem to have used them to find power. We just need some in-port and in-cylinder high-speed cameras You know - typical backyard mechanic stuff.

[McGuire]

Originally posted by Canuck


I KNOW I KNOW I KNOW!!! McGuire is the Crew Chief for Franklin R's Propster and full-time director of studies at the Prop-Derived Thrust Driven Supercar Design University

No, it's even worse than that. :

[hydra]

The state of the art uses high-speed laser cameras to capture in-cylinder motion and study combustion. Typical backyard mechanic stuff, just like you said!

[NTSOS]

Hi Andrew,

Yes sir.....that would be Fuel Stratified Injection (FSI), as developed on the unbeatable Audi R8 ALMS cars and now used on various production vehicles.

Since it's a form of direct injection, there is probably no need for wet intake port/manifold analysis.

Maybe all we need now is high-speed laser cameras to capture in-cylinder motion and study combustion as referenced by you and hydra.

John

[Ben Wilson]

Originally posted by McGuire
From what I have seen of recent wet flow research a lot of the current thinking will soon be headed out the window.

I'd assume that as most of the data which has been discussed on this thread is not based on flow bench results, but rather on the overall package that at least this bit of current thinking will remain valid.

From what I understand of wet testing, I believe it may be more inherently flawed than dry testing. I'm assuming that you're talking about spraying a fluorescent dye into the inlet stream and using a blacklight to view the flow paths.

While flowbench testing uses steady state flows to simulate a very variable situation, it has been proven to give a reasonable overall approximation. Injecting a dye into a steady state flow and attempting to analyse the results strikes me as attempting to refine a system which is known to be inherently inaccurate without correcting the flaws.

I've never played with it, and I can see the inherent value it should have, but without finding a dye which replicates the specific gravity and viscousity of fuel, I would be worried about making too many assumptions about fuel distribution based on the results.