Colin Hall Engine...

Colin ,
Thankyou for all of that input , it just goes to show what a small world the enthusiastic twostroke fratenity is , in that Derek and i were discussing , at the recent Midlands meeting ,the images on page 191 0f the CDF paper and bemoaning the depressing way that the flow vectors indicated such clear losses to the exhaust port !
I suspect as well that image 4 on the SAE paper indicates a similar tendancy , but i do wonder how all of those details would look if they included the significant effects of the exhaust system ?

Trevor

Trevor,

You can split the boundary condition two ways.

The simplest is to think of the engine as a series of volumes i.e. inlet duct, under piston volume, transfer duct, cylinder and exhaust pipe. The air/fuel mix moves as a block from one volume to the next but is modified due to heat transfer, frictional losses and so on – the ‘filling and emptying’ method. Any short circuiting charge is mixed with exhaust gas residues within the exhaust port resulting in further charge dilution and/or loss of charge.

Considering wave phenomena would require the method of ‘characteristics’ and hence the need to mesh the volumes to calculate energy transfer etc and some rather tricky sums. This I guess you know.

The problem is that the reflected exhaust pipe wave is only a pressure field; this is not going to recover poor cylinder scavenging. It might be more useful to think of the reflected wave as a mild form of supercharging i.e. increasing the trapped charge density. Hence, my concerns regarding aggressive vertical fresh charge velocity up the rear wall - too much charge is transferred into the exhaust. With velocity measurements from a motored engine one can easily determine the time of flight of the free gas column and see how this lines up with the exhaust port timing.

We ask too much of the exhaust pipe?

Colin

Hi Colin Trevor,

we have been hooking up a vacume to the exhaust port bolting the cylinder to a container with a smoke pellet, rather crude way, but its very effective in helping you see how out of balance the transfers are and all you to balance/opening/direction of incomming smoke, (I know- no thermal effect/lower speed no sonic waves) but its a very effective way to show you how the transfers work or dont by going straight out the port, Trevore much like your CFD we were discussing but Ihave to say a much better way of seing the smoke going straight out the port, its really interesting when you fit a smaller valume outlet amazing what it does to the smoke, but simply it so effective on helping you, never had any improvement in top end, while trying to do this, this is perhaps to do with the fact "I dont really know what to try and achieve and improvment up or down,"
its very good picture we open the exhaust port by pulling on a rod connected to the btm of the piston through a long sealed shaft through the smoke filled container, its difficult to see while doing this so it has to be filmed, ! then played back and whatched slide by slide. just dont have the space to hook the engine upto am electric motor or the time, but Im sure this would only copy what Im doing.

I think like many looking in, I do not really posses the knowledge or maths, or experience with what I should do to make a power improvment, but I'm making progress, and only by applying araldite covering plaster to the transfers ducts, then dyno-ing the engine on a rolling road, I have jumpped almost 10hp in little over 10 months, the next 10 months will be really intersting.

I wish I could fully understand all what your talking about, anyway

many thanks

Derek.

and Colin WOW "wall wetting" what an interesting topic and full of really helpful info, I believe I understand why my bike starts ok, them after starting it becomes difficult to start and easy to flood.

mind you since I reduce the choke size this has vastly improved, but also reduced the Torque over 8,500 rpm.

I find it so stressfull in the paddock, just before you go out to practice or race and it will not start". so I have a system I follow to avoid stopping the engine once started, and get stressed if there is a delay while in the collecting area, or the Under plug temp goies over a certain temp.

regards Derek

Trevor,

I have been trying a very weared cylinder head shape/position of the chamber, that worked straight away, increasing both torque and top end, maybe its because! its reducing the mixing of fresh charge with the exhaust gases going out, so perhaps we have a better trapped fuel mix, as no other changes, smae everything, almost 2hp difference and all through the rev range.

its an intersting shape, sadly not mine, but my idea to try it, copied from "Bosalgnia" Kart/Marine power boat engine tuning Book for the 60's.

Derek

Colin .
Employing the Method of Characteristics to duct flow pressure wave analysis is mathmatical perversity in the extreme ! Fortunately there are some pretty good simulation software packages that are fully capable of dividing ducts into equal meshes , solving the problem and of writing coefficient of discharge values . And , as one who would prefer the computer to do the iterative work for me , i embrace it all with grateful enthusiasm and , not a little relief , as the computer makes no computational errors , my maths frequently does ! Using such packages enables a computer image on the screen to be realised in minutes , indeed a whole engine analysis is there for the viewing , at any chosen rpm .

It may be worth considering , that at around BDC in a firing engine , the exhaust pipe will send a negative pressure wave return to the exhaust port , bringing cylinder pressure down , to perhaps .7atm , thus enhancing effective transfer flow and velocity and introducing the important element of temperature ! It is difficult to see a motored engine replicating that at the 11,000 plus rpm that 125 Bantams frequntly see ?

I fully share your anxiety over rear column velocity , however with multi transfer port arrangements , of varying timing , axial and radial efflux angles and duct area all playing a predetermined role , column pressure and charge density can now be coreographed to clear spent gases and leave fresh charge , of high purity , to be combusted , and if piston velocity is factored in , high values of power are available .

That is the ideal scenario , however , what we amateurs can in reality achieve in the garden shed , is another thing altogether !

Trevor

Derek,

Sucking smoke through the engine or as I have seen, flowing water through the transfer port with a garden hose, provides an accurate guide to how the transfer ports operate with smoke under vacuum or water from the garden hose. What it does not do is provide any meaningful indication of how fresh charge will flow into the cylinder under pumping at high speeds. If you apply vacuum to the exhaust port why would the gas drawn through the transfer ports bother to head up the back wall of the cylinder – it is guaranteed to short circuit.

Part of the problem is that not all fluids are equal and the same fluid will not behave in the same way under all conditions.
For example when fresh charge moves slowly through a transfer port it is the viscosity of the fluid that dominates the fluids behaviour. Because it is moving slowly, the inertia of the gas ‘column’ is small. The column of gas flows in a laminar fashion with no turbulence. When the velocity of the gas column is high, the inertia forces dominate and the flow becomes turbulent.

The Reynolds number (Re) is the term for the non-dimensional number that indicates the ratio of viscous forces to inertial forces. As a rule if the Re is <2000 the flow is laminar, >4000 the flow is turbulent and between the two transition flow. To calculate the Re number is very straight forward, however, you need to know the fluid velocity. Fortunately, you will probably not need to calculate this. A fag pack guess at the numbers suggest a Re in the order of ~50 000 so well into the turbulent zone.
Yes, have a play with smoke or whatever but to ‘see’ what is really happening to the flow you need to run the engine at working speeds and flow rates, motored or fired.

Colin

Trevor,

I will get back to you when I have a spare moment but until then method of characteristics is an example of a one-dimensional approach, which can return fair results.

Colin

Cheessuss! You blokes are rattling on -- I am still way behind on SAE 960366 paper and have to ask did this really apply to the Bantam? and IF, then HOW SO -- to your back-to-front engine Colin?

I liked the term, "Cup handle" transfers... and that has me thinking(???) would they also apply to any Bantam? The roofs of the transfers are horizontal and that suggests a flat-top piston. And all the Bantam heads I have seen on here have a partially domed convex shape which I had assumed was why the tranfer port `roofs´ were sloping to match the angle of the piston edge. Trevor´s wonderful pictures of the statures of creatures from Mars had sloping roofed transfer ports... didn´t they??

With concentration on exhaust emissions the domed piston was out because of the increased surface area of dome versus flat -- any comments?

Are there flat-top pistons in racing Bantams -- or am I spouting a load of crap?

The only Quiescent diesel combustion chambers I came across were with wide diameter, flat bottom shallow combustion bowls and multi-hole injectors (goes with no swirl...) ... on SLOW speed applications.

Cheers!
John-Boy!

COLIN thanks,

"yes agreed" short circuiting, but thats "if" you pemanently hold the exhaust open, but a better picture can be seen
applying small velocity, from transfers, this we have done by partly presurinsing (with an air line/pressure) to the smoke "prior to opening of the transfers" by droping the piston dowards, all happens very quickly, but when slowed down on film you can see a very effective process. I have not done this for some time, (I use an O'ring IN THE PISTON RING GROOVE not a piston ring, when I do this, as too much air leakage past the ring causes shorting out the exh)

I also tried the water method by introducing AIR bubbles, not as good though, need tiny bubbles / difficult to achieve and the same type of issue with shorting, I found smoke to be best to see a picture of what is kind of happening.

I understand the running motored under power -measuring I think "actual volume" of air, in and out,

increased/ better breathing means and a bigger volume ?, but this does not acuratly say what actually was trapped when it was fired !, but I think one assumes if the volume pumped increases so does the trapped volume, or it should be better, I have no experience of ever seeing or recording this, its way our of my depth.

but I have recently foundout there is no substitute for actual cut and try methods on a dyno, for sure the maths have to be about right, thats where Trevor is of great assistance, but the tinyest of changes make masssive differences (all compaired on a accurate/consistant???? rolling road).

cheers Derek

John

As I guess you may well know, a problem with Diesel fuel injection is to get air into the spray cone (fuel droplet pyrolysis is a current research area) and aggressive squish with a re-entrant combustion bowl is a common form. Swirl is kept modest so that individual spray cones do not merge into one another. My comments regarding quiescent combustion was very much directed towards racing engines, not Diesel.

Diesel piston crown geometry is all about controlling air movement within the combustion chamber. Get that right and emissions go down and fuel economy go up. I am assuming that the domed piston crown on a two stroke is for structural reasons – the increase in surface area would be small so yes, a small increase in wall quenching but on a race engine who cares.

I guess the ‘ideal’ combustion chamber would be a sphere providing maximum volume to surface area ratio and the spark at the centre for minimum flame path. Perfect air/fuel homogeneity and no air movement - sort of ‘burn it as fast as you can and get rid of it’.

I have a sneaky suspicion that in the old days all that swirl business was to promote fuel/air mixing within the cylinder, as carburettors were rubbish.

Colin

Trevor,

I have fond memories of struggling to write a method of characteristics program for the Bantam back in the 70’s in FORTRAN. I ran it on a DEC PDP11 which with the peripherals would fill a small shed.

My point was and still is, is that any modelling software is only as good as the raw engine input data. I would be very impressed if someone is running a fully instrumented Bantam engine on a dyno. One of my clients logs cylinder pressure every 1/8th degree crank angle to obtain accurate heat release data for their combustion model. Another example; piston blow-by is a sensitive variable and I have no idea how you could measure this in a two stroke. There is a whole host of things that can go wrong in accurately modelling combustion up to the start of blow-down and if that is not right, the rest is compromised. Taking data from one ‘similar’ engine and using to input into a different engine is asking for trouble.

I have no knowledge of proprietary exhaust system software so an unable to comment. Anything that provides a starting point and perhaps a guide to direction is worth having but I would treat all such packages with a fair degree of caution.

The other problem is that a lot of engine development testing and modelling is done steady state, not a common occurrence with racing engines. The world is littered with engines that worked on the steady state dyno but failed on the racetrack.

Colin

Spherical balls Colin -- the Ricardo Comet pre-combustion chamber with pintle injector and no concern about poor fuel consumption (losses in the throat) because its for racing -- NOT possible on a Bantam???

-- take no notice, I´ve forgotten what I was talking about, forgotten what day it is, forgotten the wife´s birthday, wedding anniverary and cannot find my teeth...

-- our re-entrant toroidal, DI was better -- BSFC, emissions & overall -- than GM´s IDI but we didn´t getter the Military contract because we had a Square Head....²²

I liked the bit about not trusting too much modelling and having instrumentaion for piston-blow-by on a 2-Stroke and stll wondering if that could be simulated in some way....??
Cheers!
JayBee...

²² German company pretending to be Canadian.

Derek,

I am having a little difficulty unpicking your last reply but if I am reading it right then I need to explain the Reynolds number further as it should help you get a handle on your flow simulation.

A handy thing about Re numbers is ‘Similarity of Flows’. If you hold the Re number constant, for a geometrically fixed duct you can characterise a flow condition that will hold true across a number of different flow conditions. A classic case is using an exact scaled down model of an aeroplane in a wind tunnel. You achieve the aerodynamic ‘performance’ of the full size plane.

To get meaningful results from your flow tests you first need to calculate the Re for each target engine speed at operating conditions. For that, you will need to know the fluid density, kinematic viscosity, dynamic viscosity, its velocity and the mean transfer duct length.

See:
http://en.wikipedia.org/wiki/Reynolds_number

As you really don’t want to get into further complexity of multiphase flows it is simpler to ignore the fuel element of the air/fuel mix. This is only one reason why you would not use a water/air bubble mix. Equally, the dramatic change from flowing a gas like air to a liquid would push the limits of the Reynolds number (I don’t know this but it’s not something you would normally do). Remember that just in terms of density alone you are looking at air ~1.3kg/m3 and water 1 000kg/m3 - not insignificant.

So, keep playing with the flow simulation but don’t get trapped into assuming that what you find emulates the real world condition. The closer you can get to a real running engine the greater the quality of the results. So even a motored engine at modest speeds gets closer but even then, there are other factors that will shift the results.

John,

Ah, yes I remember the Ricardo Comet very well. The throat had a tendency to thermal cracking and the Lucas CAV DPA pump was guaranteed to produce maximum cylinder-to-cylinder and cyclic variation. Was that the Ford Dorset?

Colin

Hi Colin thanks,glad your able to follow my post, some highly technical responces from you, "comming back down to earth" what do you think of the std 175cc formula, and if you were starting again, would you build a 125cc shortstrokeor or go for the more limited 175CC (allowed to use a 64mm piston-that are readly available, taking the bike upto 190cc)? (in the spirit).

regards Derek

Colin! -- it was the York. As I remember it every engine using the Ricardo Comet IDI system (including the Japanesse) had their throats cracking. Since the fine cracks didn´t do too much to overall performance not much fuss was made about the "dífficulty" and at Servicing time the dealers were quickly changing the inserts.... We tried ceramic inserts and ceramic coated DI chamber & piston-crown surfaces and they all failed miserably -- but it was fun finding out -- at Henry´s expense. Our instrumentation in those days was terribly crude. I measured piston temperatures with thermisters and a linkage -- connected to the con-rod -- supplied by Associated Engineering on the York which lasted 6 hours of dyno-test time. However we had a Bosch Injection Rate Analyser which gave us valuable & worthwhile information.

Wish now I´d done more hands-on-tuning with the 2-stroke petrol engine but the (day job) 4-stroke diesel was bringing in the bread-&-butter and my time for the Bantam was such that it was limited to practice and racing and I even gave Colin Aldridge my Icarus-2 barrel to shape the ports. He used the water-flow from the kitchen tap to check his workmanshup and although this was harshly criticised by well known stars of the Bantam Racing Club ... Icarus-2 did rev-on to thousands more rpm more than Icarus-One. Remembered well is that at Llandow it was going like the clappers -- really quick -- but making a dreadful noise and the rev-counter said something like 11,500 just before the piston broke into two pieces across the gudgeon pin bosses. Yeah! I know, BSA standard were no good for that applícation!

Cheers!

JayBee.

Derek,

Having no experience with two stroke engines since I gave up Bantam racing over 20 years ago, I’m not sure I am in any position to comment. What I would say is that I would tend to go towards the more mature formula as there should be fewer bugs to sort out - but if you like a challenge?

Actually, with no intention of racing or building again I would go 125 if only because it’s unfinished business.

John,

We did have a go at instrumenting the throat insert but to no great effect, I seem to remember. The Puma and Lion are wonderful to work with in comparison. Are you at Aachen or Cologne? I get involved with Dunton a fair bit…………….

Colin

Hi Colin! -- Residing at Bensberg 15km from Cologne. Ì left Ford Dunton and joined Deutz Diesel Canada (Montreal) as KHD Cologne´s attempt to penetrate the North American diesel market -- in `74. When Airdiesel Canda-- KHD Canada Incorporated -- lost the military contract for the engine for the replacement vehicle for the Jeep (won by GM, of course!) KHD Cologne wiped us out. One day R&D manager, next day unemployed -- so it goes -- yet I had some marvellous experiences in those ten years. Was wined and dined at McGill and Concordia universities as if I were a Professor -- well I did run a Component QC/QM Cost Reduction exercise at McGill and assisted in Concordia, which made feel as if I were a proper academic -- which I ain´t ....

Met up with some of Dunton´s Testing Operations (`Resting Operations´ we called it...) ex-staff last September. Some of whom were horrified at me decribing working at Ford Dunton as a most boring period of my life. Maybe you knew -- or know -- some of them. Andy Boyle (retired from Ford a couple of years back) who currently has the GP5 Royal Enfield being ridden in VMCC events lent me his 250ABS ( the fore-runner of his GP5) in 1971 convinced me that the Bantam & I were incompatible regarding winning any other than Intermediate races and that the 250ABS with 5 gears and plenty of torque meant I could really compete against other 250s. Not that racing the Bantam was not fun --it certainly was in Intermediate races -- particularly when it rained. I think the extra weight helped there.

Did I get something wrong with looking at the Jante drawings -- I thought the transfers were supposed to be directed towards the back?

Cheers!
#
John-boy!

Colin Hall wrote:

Derek,


Actually, with no intention of racing or building again I would go 125 if only because it’s unfinished business.
Colin

Hi Colin

I read with puzzlment this comment, " I for one feel or doubt you have any unfinished business! especially in Bantam Racing, you won the championship!, i remember the year well, this was a feat that eluded many and me, even in my finest riding/racing tuning years.
ther are many more great rider racer tuners of that period and before who never won a championship, perhaps you would like to explore this comment.

Derek

John,

I never worked at Dunton but did spend a very fruitful year at Jaguar R&D at Whitley working on their 4.2l V8 direct injection gasoline engine. A happier bunch of guys’ than at Dunton…..

No, you have not missed a trick on the port layout in the Jante paper, the exhaust port is at 3 0’clock on all the diagrams.

Derek,

You never finish a job in R&D it just goes somewhere else. There is always more you can do and that was true with the Bantam with the exception that for me it was turning into the day job. I guess I was coming to a dead end despite running my own dyno and being able to borrow some very expensive kit from Imperial College I was not going to be able to explore in greater depth the areas that interested me.

I guess if I was starting again I would build a transient test bed as I think there is plenty of fun to be had there. Rapid prototyping in clear resin opens up a number of possibilities and coupled with a fast CCD camera you could have a look at transfer port level flow dynamics. As usual, it’s just a matter of time and cost and for now all of those commodities are taken up with dinghy racing – so perhaps when I retire?

Colin

Colin you are right about R&D that you never finish. It´s like learning anything -- you get into any new subject and think you have learnt a lot (and enough) and are now quite clever at it -- and then find someone else has either done it, or is ahead of you -- and what you´ve learned is not enough. You have to go on... like a Bruce spider.

Speaking of Jaguar I visited their Refigerated Air-Conditioning R&D section at the end of fifties when they were selling the XJ... whatever it was -- to the Californians. It was like being in the back yard of a Bantam rider´s place: one engineer who was his own fitter, mechanic and artisan. Well, at that time, the British car driver had only just got used to to having a car with heater as part of the car ... and Jaguar XJs were popular in America except for having no air-conditioning.

Because its there, I have kept all my old text books, I am still playing with maths cum intelligence questions. An example is, "What time is it when the hands of a clock are contra (directly opposite) between 3 and 4 PM?" I said that at a SAE hospitality suite in Detroit and it was akin to the "Geneva Top"²²² epic ...-- such that everybody and the President´s brother were trying to produce a formula for the contra hands question and forgetting all else, including the Hospitality Suite boozing!!

I suppose its a bit like the fast Bantam lads at the moment saying, "Now what do I do to make my Bantam go quicker -- take notice of Trevor or Colin? -- or should I first try and work it out what they are both getting at and compare that with so-and-so´s recent revelations -- remembering that he might have purposerly left something out.

²²² Before the second world war a week-end sémina -- in Geneva -- of the world´s top scientists was brought to a halt in the first hour because some nut spun a Top on the table which eventually turned itself upside down and spun on the peg used to to first spin it by the fingers. This Top was a sphere with a chord cut off, in the centre of which was the peg. Spinning it with the peg uppermost and vertical has the top spinning for some time on its round base. Then as it slows the peg goes horizontal for the top to finally stand on the peg -- upside down -- stll spinning.

Both examples have a formula for solution and a scientific explanation.

To All Bantam lads -- Have a good time at Mallory!

Cheers!
JayBee.

John,

I hope no Bantam tuner would take my musings as anything other than that. Therefore, my vote would go to Trevor. However, if anyone is happy to bend the rules a bit I have a couple of Eaton superchargers being shipped over from the States that would provide a performance boost, although you will need to drop the compression ratio by more than a tad.

More power to your slide rule,

Colin

Power to my slide rule, Colin? I have two -- one normal and the other circular -- but I got back into playing with logarithms, recently, and amazed the missus by showing how calcs can be made easy with logs. Takes a bit longer but the wonder & logic is exposed wheras some of society consider Higher Arithmatic a form of magic they´ll have nothing to do with...

I´d go along with a bit of supercharging of a 125 Bantam engine -- blowing a few to bits would be a long-lost contentment -- and to get one that would increase max power to somethimng like 35 -40 bhp would be worthwhile -- and it would give the chubbies a chance....!!


Getting back to slide rules and Log-Logs & that dreadful past: Now that everything is done by Information Technology, Personal Computer and Mobile Phone, youngsters cannot do any form of mental arithmatic -- but why should they?

We had a lecturer at Bristol Uni who used to show off at how well he could add and subtract -- in his head-- 6-digit numbers and I often wondered what use that was to us struggling students who´d rather have a hint (from him) at what might come up in the exams. What a poseur!

And at Dunton we had Engineers who worked more as Technical Clerks -- another bit of Ford `Boast of Intelligentsia,´ I guess!! My engineering talents (whatever they might have been...?) were wasted on justifying costs of purchasing new forms of test equipment.



Cheers!

JayBee -- for John-Boy.

It has been an interesting experience revisiting two-stroke engines after a quarter of a century but before I stop interfering with your fun, I have a few further comments that may be of interest.

As I guess, we are all aware getting more power means more air and fuel. The hard bit is getting more air through the engine. A figure of 55BHP at, I think 13000rpm has been quoted for the Aprilia engine. To achieve 220BHP/Litre some degree of charge boosting will be required. A standard calculation can be made to get a handle on what this engine is doing.


Engine speed 13000rpm
Engine speed 217rev/s
Air density 1.5kg/M3
Engine swept volume 0.000125M3
Air swallowed volume 0.0271M3/s
Air mass flow 0.040625kg/s
AFR 14
Fuel calorific value 42000kJ/kg
Fuel mass flow 0.0029kg/s
Heat release 122kW
Shaft Power 41kW (55BHP)


I have adjusted the air density to achieve the target output power.

Atmospheric density of air is 1.3kg/M3 so modest boost lifting it to 1.5kg/M3. Perhaps of interest is that if we look at the rest of the heat balance something of the order of 45kW ejected down the exhaust pipe.

I guess we are all happy with how the expansion chamber works and how the relative lengths determine the timing of the pressure waves. However, the amplitude of the pulse is a function of the blow-down energy and it is the amplitude that drives the scavenging process. It does not matter if the pipe lengths are correct, unless you have good combustion and therefore maximum heat release (in this case 122kW) you will not achieve high pulse amplitudes and therefore high levels of charge density.

Unlike a four-stroke engine, the two-stroke is very sensitive to the previous combustion event. If there is a slight miss-fire in the previous cycle then the energy available in the exhaust to scavenge the next cycle will be low resulting in further loss of power. All engines suffer a degree of cyclic variation, it is good mixture preparation, and cylinder filling that will keep this to a minimum.

That’s me done then, hope I have not repeated what you already know.

Have a good season,

Colin

Colin ,
Not sure if you are aware , but the 55bhp is taken from the gearbox output sprocket on the dyno , crank power must be up a bit on that if transmission losses are factored in ? Damned impressive power figures considering that development effectively ended in 2008 . Oh , by the way , to lease a machine from the factory would , then, have cost over 160,000 Euros . Stick to Bantams then !

Trevor

Colin Hall wrote:

It has been an interesting experience revisiting two-stroke engines after a quarter of a century but before I stop interfering with your fun, I have a few further comments that may be of interest.

That’s me done then, hope I have not repeated what you already know.

Have a good season,

Colin

Hi Colin
sorry that comment "thats me done" is just not allowed, ! but seriously "I do hope thats not it from you".
I like many have found your contributions really interesting, in trying to develop my understanding of the very complex world of the often said "simple little two stroke"!!
please make sure its not your last and that you look in and post from time to time.
Very Best regards Derek