Edward Irving

Edward I thought you might like to have a look at this free-hand sketch I did as part of the general discussion at the recent Midlands Meeting. Shown are the relevant exhaust pipe parts set against crank position, case and cylinder pressures. No specific engine is represented just a general trend towards the train of events. It may seem a little confusing at first, but all will be revealed with a longer article showing how to calculate for yourself the various pipe components and some other stuff as well! You might draw your own conclusions as to how case comp ratio factors into all of this?

Trevor

Thank you Trevor that's great, your right I am going to need to study the drawing,

Thank you again

Edward

You`ll get there Edward!

Bulk flow occurs around BDC when the pipe effects are at a maximum with the biggest depression seen at the exhaust port face. The sketch reflects this with the pipe diffuser working through half transfer port open to half closed, at transfer opening gas flow is zero with a positive pressure ratio across it`s face. Transfer closing is with mixture flowing with induced velocity that converts to pressure in the cylinder.

Case comp ratio is geometric calculation, not volumetric!

Trevor

I hope I will get there!

Firstly thank you so much for your time and the sketch, there is a lot of information there.

I have been trying to work through your sketch in my mind and by sketching it out again. I seem to understand things more quickly when I draw them for some reason.

Not sure if I understand the transfer flow reversal? the Transfer port opens, the inlet port is closed and the exhaust port is open. The piston is moving downwards in the cylinder and as the piston descends the new charge is expelling hot exhaust gasses from the chamber. Why is there a transfer flow reversal before BDC? is this because of the pulse arriving at the exhaust port, I don't think so this would be at or after BDC? and if so why does the case pressure reduce after transfer flow reversal? if piston is moving to BDC and the flow rate in the transfer passage has been reversed pressure should go up?

Mindful that I have included a lot of ? I think I am going to continue studying the diagram. lol

Edward,
When I first started racing a 125 Bantam back in the mid-60s of last century, I always removed the head and barrel after every meeting. Another ritual was to pull a clean white cloth through the transfer ducts, the cloth was always smeared with black deposits. At first I took no notice, it was just one of those things, it always happened. When I finally woke up and asked why black, when the mixture going through was clear, the only culprit was combustion deposits, but how did that get into the transfer ducts? Flow reversal was the only way it could have. The cylinder had insufficient time to blow-down (reduce pressure) before the transfer ports opened. So the transfers acted like extra exhaust ports until case pressure eventually became higher than cylinder pressure. That was then! The sketch shows reversal, or back flow, as a small duration pressure spike. As a graphic example, at full bore and 13,000 rpm the Aprilia has to blow-down from 62 bar cylinder pressure at 18* ATDC before the transfer ports open and needed a massive blowdown area to do it!
Within the bounds of what we call normal, and no matter what the porting arrangement is , there will always be more pressure in the cylinder, when running in the engine`s power range, than in the crankcase. Thus there will be reverse flow from excessive blowdown pressure at initial port opening.
As the piston continues to drop, the pressure ratio finally goes negative as the pipe diffuser starts to do it`s intended work with the piston rolling through BDC. It has also been conclusively determined that there is very little case-flow into the engine, it is overwhelmingly that mixture that resides in the transfer ducts that gets into the cylinder. This is why I depicted the diffuser working from partial transfer opening to closing where the bulk flow is concentrated.
You suspected that the reason could be a returning pulse arriving at the exhaust port? That won`t be the plugging pulse, or at least it shouldn`t be if you pipe is correctly proportioned, for that returns just prior to exhaust closure to stuff mixture back into the cylinder. The only exhaust effect needed at the exhaust port around BDC is the negative, or, suction wave that is created by the diffuser.
Crankcase pumping only serves to start the engine and bring it to the power band. Once in the band you can forget all about pumping efficiency!
Keep asking those questions Edward, every one who reads the Forum contributions benefits that way.

Trevor

One thing to remember is that nothing flows in until The pressure above the piston in less than crankcase/ transfer duct pressure, a small point to think about and makes the case for aux exhaust ports plus a tapered top face to give a small signal up the pipe to start with plus pressure drop from piston port blow back robbing the transfer advantage

Thank you Trevor that makes total sense now you have explained it.

Dan just one question, when you say "tapered top face" I am a bit confused. Is this the shape of the port so it increases in size as the piston moves down?

Thank again guys, so helpful.

should really have said a curved top of exhaust port, not only for gas reasons but ring life too

Edward,
There are two equations to be worked through that will then provide a tuned length for the retuning plugging pulse within the prescribed exhaust duration for an exhaust system. It is from this one figure that all the other lengths of the component parts of the pipe are proportioned.
As speed of the acoustic pulse (speed of sound) is proportional to the square root of the temperature, in Kelvin, so temperature selection has no need of being considered an ultra-critical number. Indeed temperature varies in the pipe length over 100*or more, being lowest at parallel section and yet peaks again at the tailpipe entry!
Some of the data for inputting into these two equations are constants so I have calculated them to a single input value making the calculations much easier, the first one is….393! I will explain how this constant is arrived at later.
The first value needed is the acoustic velocity of the exhaust pulse in mtrs/sec. for this we need a temperature reading, in the absence of thermal gauging a temperature of 480*c can be used. As we need temperature in Kelvin, to convert simply add 273 to the 480, this then becomes 753*k.

So the first equation is: SQRT (393 x 753) = SQRT of 295929 = 544 mtrs/sec This number is then inserted into the second equation.

The next equation works out the tuned length from the piston face to the reflection point at the junction of the rear cone to the tailpipe, in mm, and uses another numerical constant namely…..88.
Within this equation the data from individual engines is factored in for the example I have used the exhaust duration of 197* and engine rpm of 9250 and carried over the velocity value of 544mtrs/sec

T/L = 544 x 88 x 197/ 9250 = 9430784/9250 = 1020mm.
T/L = 544 x 88 x 197/11000 = 9430784/11000 = 857mm.

The lower of the two sets of results would suit a 125, I can`t see a big Bantam engine peaking at 11,000 rpm, and serves to illustrate that juggling the variables within these two equations can offer some insight into the effects of data changes in arriving at a tuned length figure. Only the time element is changed for the 125 pipe with the increase in rpm. Inserting a different value for the wave speed from 544 mtrs/sec will alter the tuned length yet again! It is well worth taking time to mix and match these numbers to get a handle on what is happening within the pipe length and the length changes they produce.

Overriding the conclusions arrived at by working through this mathematical exercise is the caveat that, all of this is only an approximation, just to help the tuner on his way and into the ball park!
To see why this is so we can look at the speed of sound; it varies with engine rpm, engine power, mixture strength, ignition timing and compression ratio. Then it varies for each degree of crank rotation. Then it varies at each junction and expansion/contraction along the pipe length. Then finally, initial exhaust pulses are supersonic (faster than the speed of sound), decay to sonic and return as subsonic!
No way can all of this be factored into any proprietary pipe formula, you have to make a pipe, optimise the engine settings, then experiment further and so on, ad infinitum! And if all of that wasn`t bad enough, at 2,000rpm lower than peak the pipe isn`t anywhere near as efficient, so perhaps, as we only have 3 gears to work with, a pipe that works somewhat better at lower than peak rpm might be overall a better option?

The speed of sound in an ideal gas (air for instance) is a function of the stiffness of the gas divided by its density. Since these are incalculable for our use they are replaced by suitable equivalents from the ideal gas law, these being specific ratio of heats, and a universal gas constant. They are 1.4 for air at 20*c and 287j/Kg-K* respectively. As the gas passing down the pipe is clearly not pure air and is way hotter than 20*c, then these values are adjusted to more closely reflect exhaust gas, namely 1.35 and 291. By multiplying these two values together we arrive at the constant I referred to above of 393!
Given that we have this constant established for our equation, the speed of sound only then varies with temperature; but, there is always a` but` with Bantams, pipe temperatusre within a race context is always varying so even that is subject to change.

Simple things, Bantams?

It case you are wondering about the remainder of the pipe body, that comes another day, I have been ignoring others in my life!!

Trevor

Thank you again Trevor

I understand now why there is not a perfect design, or at least its not possible to design a perfect pipe because of changing conditions.

I can also see why you might want to use a different pipe at Darley Moor to the one you would use at Lyden Hill?

I will have a play with the numbers over the weekend and I will let you know what I come up with.

Thanks again

Edward

Edward,
You will learn a great deal from working through possible data combinations, throw in some crazy stuff as well, you might be surprised by what is thrown up!
One more task for the weekend, work out the area of your exhaust port, you will need that to determine the exhaust duct outlet diameter, that in turn can decide the exit of the header pipe at the diffuser junction, and so on.

Trevor

The idea of changing pipes for circuits is new to me -- although the idea does have some merit when I think of the annoyance at Brands in having to change the rear wheel sprocket between practice and race....

-- wouldn´t it be easier to change the gear ratio by the gearbox (G) sprocet and  (R) rear sprocket sprockets??



Someone in the club  can supply you with the list of preferred combinations of   G & R teeth for each circuit....

CheerS!

JayBee...

I'd say that changing the gearing to suit the circuit is essential. If you can alter the engine's power characteristics in a way that gets you round a particular circuit more quickly, that would be a bonus. Presumably, you'd have to change the carb settings to suit the different exhausts, which would make things rather complicated, particularly in view of the limited practice available at race meeting these days.

Having said that, on my own racing Bantam I never altered the sprockets. It wouldn't pull higher gearing, and was always flat-out well before the end of any straight. Cadwell, Oulton, Mallory, New Brighton Promenade, Perton, Snetterton, Rhydymwyn, it was all the same.

Good point John... hadn´t thought of the carb....

I changed the rear sprocket at Brands by teeth one larger and surprised myself... Unforgrettable.... I was actually racing...

Edward,
Taking someone else’s theoretical recommendations at face value can often be a dubious leap of faith, like being persuaded that the earth is in fact, flat? So to give some credence to the equations in these articles we can retro-fit published data from the Aprilia exhaust pipe into the sums and evaluate the results.
Available data for the Aprilia “Tubo 102” exhaust pipe drawing indicates a tuned length of 803 mm.
The known exhaust duration is 202*
The known peak power rpm figure is 13,000.
As before in the previous calculation we need the wave speed first, and this is given as follows:
Let…..w/s equal wave speed in mtrs/sec.

T/L = w/s x 88 x 202/13,000 as we already know the T/L of 803, we can transpose the equation for w/s

w/s = 803 x 13000/202 x 88 w/s = 587 mtrs/sec
This is a velocity value I have seen before, or very close to it, in some of the various forum around the world I look in on.

To find pipe temperature, we can insert the wave speed into the second equation.
Let T equal the temperature in degrees Kelvin.

587 = SQRT: 393 x T….. Rearranging for T
T= (587^2)/393 T=877, as T is in Kelvin we need to subtract 273, T now become 604*c

For an engine delivering nearly 60 crank hp and 16 bar bmep, more than double that of a good Bantam, the two values are very credible and correlate well with known parameters, therefore the method of working through pipe lengths as described, can be performed with some confidence!
I am always mindful that it might be suggested, that in theory there is no difference between theory and practice; but in practice, there is!
Just in case anyone should wonder why I often cite the Aprilia engine for an example, it is simply that there is just so much detailed information available from the two men instrumental in taking it to the world beater that it was, and all just for the asking!

Trevor

Hello all,
ive recently read an interesting article with regard to reducing spark voltage at high rpm by some 50%, which enabled the writer(Luc Foekema) to surmise, amongst other things, that his exhaust pipe was indeed to long for his optimum power/rpm engine design. Im not sure where this reduction in spark voltage was achieved (possibly at the coil feed terminal) but would seem to be good check for correct exhaust length... this of course is only my understanding of the article.

The old fashioned ignition system was about 6,500 volts at the plug, Nigel,  so just revert to coil and battery ... but I cannot see the point at all....Transitorised  ignition gave about 15,000volts and high-voltage, capacitor must have been higher still ...and with the latter engine-management can be achieved....

Maybe I've missed the point relative exhaust pipe length -- is it something to do with combustion temperature or rate of burning -- I think I must have missed something somehwere -- where´s the connection of high voltage ignitionj to exhaust pipe length????

JayBee....

Nigel, John,
Have to say the proposition seems counter-intuitive, re-reading Luc`s article there seem to be a number of unsubstantiated assumptions made! Luc`s proposition was debated at some length on Pit-lane a couple of years ago and was praised and rubbished in equal measure, I`m not an electronics engineer so reserve judgement, but remain to be convinced that pipe length per-se is determined by adjusting spark intensity. There are far too many variables to conclusively state that the technology is a precise factor.

Mind you it is always good to have us doubters proven to be wrong, but no other independent work has been done, that I know of, to prove or disprove Luc`s findings, so maybe the jury is still out?

Trevor

Trevor, John,
it seemed to me to make sense... lower spark intensity... lower combustion heat ... low exhaust temp...therefore slower pressure wave...increase in engine power... exhaust out of sync with engine port tuning... could be fairly easy to recreate this phenomenon.just need some kind of variable resistor dial in line with the 12v power supply to the coil.

Hey Nigel, do you think that spark intensity could be another factor to incorporate into the TorqSoft pipe calculations, making it a more comprehensive and realistic package?

Trevor

hi Trevor
, i would think that any extra information that could be deemed to have an effect on exhaust gas temp would be a good thing in being able to get closer to the sweet spot, regarding tuned lengths. Use of dyno/data equipment is very much outside the limitations of your average garden shed, bantam development, research departments,
i wonder if primary and secondary voltages of the various coils in use are documented anywhere and if there has been any research in to lean/rich fuel delivery with coil voltage manipulation at various rpm levels, which have effects which can be used to improve 2t exhaust design/engine performance?

Isn`t it a strange irony, way,way back Bantam racers were forced to endure contact breaker ignitions producing weak, intermittent sparks. That was then superseded by contactless systems that produced a big, fat spark in limitless numbers that could even fire up an oiled plug, and in the case of my engine gave a performance boost as well, then came retarding and then programmable stuff and now as Luc has apparently shown we are back to weak sparks?
Might be a good question to pose to Wob and Frits on KB Nigel, I can see that one having a good run!

Spare a thought then for the Bonneville Boys, that are regulated to use contact breaker ignition on their record breaking Bantams, as I know the self-limiting affect that can impose, they do very well indeed!

Trevor

Hmmm not convinced -- we surely need numbers and perhaps PV diagrams or something that shows -- have these people used dynos and good electronic  measuring equipment*** to prove their point. Last of all if  if small improvements in performance on dyno are being quoted, are  these being compared with corrected results corrected  to a Norm or  Correction Standard for ambient temp and pressure -- plus humidity...? If all the testing is done on one day it is still possible that in the late evening the test results are effected by the ambient changes to as much as 3%....

And subjective `feel´ for performance improvement needs numbers to verify it conclusively although I must admit Icarus-! often went better on a cold damp day ....

I cannot get on Facebook or where ever it is that this is being chatted....

*** like properly calibrated  pressure and temperature guages.