Colin Hall Engine...

Hi Nick have some questions ref the Colin Hall motor (replacement
made by Tom).

I know Tom rebuilt it as close as he could to the original ref size and
direction of ports etc. I can't see an o-ring grove for the cylinder head,
is there one fitted in the corner of the step? (Does that make sence)
Also what piston is used? It must have a short skirt on the inlet/exhaust
side?

Mike

Good evening Mike,

I think Tom Did a more basic copy/design of Colins original. I believe Carl asked Tom to reproduce the motor with more reliability in mind and Carl really wanted to keep to same layout of the cylinder for the good of bantam racing to show people how varied all the bikes are.

I remember seeing the original cylinder at Carls house and think it was a fair bit different.
From memory it had more transfer ports but was beyond repair after many years of racing,
Water leak problems i think.

The latest spec bore is 53.75mm running early an TZ 250 piston with the dreaded centre rear peg. Tom then machines the piston to suit the bore as per the Snowy TZ 350 piston.
Unfortunately i have not got the drawings and notes anymore as this information was passed to Rob the new owner who i believe is busy with his own updates to the bike. I hope this will encourage him to post as he has been improving lots of stuff.

Come on Rob let us all know what your doing.

Cheers Nick

HI NIck what's it like down south not been home for two weeks hope to be home Friday then I will tell you all what I am up to if the spy have been about you will no I have been to lydden 2x in the last month let's say I should be the first to race ????

JB have you noticed the holes/slots in what was an air-cooled fin(top) on this back to
front motor so more water flows round the L/H side of the cylinder and just a wee
bit round the R/H side...

Yes Mike -- I didn´t want to show my ignorance by asking why slots one side and holes the other?? I hazarded a guess that initially there were all holes and whomever made the slots made them because he felt their was insufficient flow to the head in those areas.

In opposition to that -- a number of people I know, on bigger than Bantam engines, have water-cooled the barrel and left the head air-cooled. Andy´s 250 ABS was like that and it was very very quick.

I wonder if anyone has ever tried to find out how much a racing, ported two-stroke cylinder distorts in action...?

Or how much barrel distortion there is by just the tightening down torque?

It has been done with aircooled, 4-stroke, 4-bolted-down barrels where the discovery was of cylinder walls bowing inwards and, of course, the temperature expansion, at the top, could be calculated. KHD Deutz used a cast iron Cuff -- the depth representing the cylinder head deck -- with a hole large enough to allow the rotary hone to enter and with the 4 bolts tightening the cuff down to normal torqe the barrels were given a final `lick´ with the hone and a few microns of relief at the bottom third of the cylinder so that when in heated action the barrel was parallel (well, nearly!!). Deutz carried on this production procedure right up to the end of their production of air-cooled engines even though "Experts" opined it unnecessary.

Cheers!

Hi John & Mike,

Interesting observations and comments,
John my guess is that there is a certain amount of distortion happening with all our engines.
But by how much? maybe some of our development men could let us know(Tom Peter Trevor Mick Ted Ned)

Mike i really liked looking at the barrel because you could easily see how it was built.
It would interesting to have Tom or Trevor describe the manufacturing process.

All the best Nick

What do you think nick?

Good evening Rob,

Happy to see your pictures on the welding and new bearing housing.
As you said to me you wanted the cases to look like a factory job ,i think you have
done a great job, better than my efforts of melted saucepans at college.

out of interest did you vapour blast the crankcase, i dont think i have seen a set so clean!

A much better radiator to. Should be a bit safer being more inboard if you have an off.

Keep up the hard work.

Cheers Nick

Yes the cases have been vapour blasted with aluminium oxide and then dry blasted with glass beads to give it a shine. I washed everything thoroughly and lapped all of the cases, then washed them again. The radiator is off a Honda MTX 125 and seems to fit pretty well. I welded on new brackets and moved the top water pipe to the correct side to suit my bike. With the radiator being taller I have got to relocate the water pump as there is no longer space for it underneath. Think that's going on the other side. Glad the exhaust is well out of the way haha.

An email from Les Eggs prompted me to have a dip into the Forum, which I had not done for some time. I noticed that Tom Miller has rebuilt my ‘back to front’ engine. What is not obvious to anyone looking at the original barrel is how it came to be so. The following is a very brief outline of the thinking and practical development which may be of some value to current engine work.

My primary aims were to:

1. Maximise cylinder filling with minimal mixing between fresh and spent charge, and minimal short-circuiting/fresh charge down the exhaust pipe.
2. Reduce the pressure losses through the transfer ports by increasing port width i.e. reducing peak gas velocity.

Without incurring the pressure losses of a reed valve, the back to front layout freed up a big chunk of cylinder wall for a pair of extra ports. To give the piston ring a fair chance the main ports were split giving four main and two boost ports.

Driving an engine with an electric motor (motoring) is a standard engine development technique. For loop scavenged two stoke engines this can be readily done in the home workshop. The idea is to motor the engine at constant speed with the cylinder head removed. The gas velocities are measured across the top face of the barrel with a simple vertical Pitot tube. The measurements are taken on a fixed grid (say 5x5mm) so that a contour map can be drawn of the velocity profiles.

The first task is to ensure that the main transfer ports are ‘balanced’ and to do this the boost ports are blanked off. Asymmetric flow will corkscrew fresh charge up the cylinder and this will be seen in the measured values as an off-axis rotation in the velocity contours. Plasticine can be used to modify the transfer ducts to achieve a balanced flow field.

The next task is to address the velocity profile with a notional transverse line at mid piston reading around zero velocity and with velocity uniformly increasing towards the rear wall (in a conventional engine). With the combustion chamber in place, this will result in an ordered tumble with minimal mixing. The central core of the gas is folded over whereas at the back wall the gas is on the outside of the tumble and will have to travel at greater velocity to arrive at the exhaust port at the correct time (think of a washing machine).

To control the quantity of fresh charge escaping down the exhaust port the main transfers are kept ‘flat’ to create a stagnation point towards the rear of the cylinder. This effectively kills the velocities and provides a useful density gain.

The next step is to repeat the motoring tests with just the boost ports and again refine as necessary.

The final step is to repeat the tests with all ports open. The boost ports are now controlling the vertical component to the main gas flow – tumble velocity. When you are happy with this replace the Plasticine with Devcon or similar and repeat the entire process.

Some random notes:

This engine never seized nor suffered from detonation and produced useful power from 8,000 to 12,000rpm.

I ran a handful of combustion chambers with varying squish geometry and found no great difference. Squish and other dead volumes are very important with regard to engine out exhaust emissions but I am not so sure with a race engine.

I was involved with a project studying flame propagation in a four-stroke, four valve, pent roof, spark engine with full optical access. Of interest was that the initial flame kernel is very fragile and can be torn with modest rates of air movements – swirl, squish and tumble. The critical time is when the first 5% of the mass fraction of fuel is burnt (the percentage of fuel in the cylinder). For every engine cycle with poor initial flame kernel growth the combustion never recovered – peak cylinder pressure was down and the crank angle between ignition and 95% mass fraction burnt extended. Further work indicated good air fuel ratio at the spark plug for all engine cycles good and bad. Good initial flame kernel growth always resulted in robust and stable combustion.

I am still of the opinion that:

Two-stroke race engine combustion systems should be quiescent and with minimum flame path lengths - compact combustion bowl geometry. Low loss filling of the cylinder with low levels of tumble and no swirl. Minimum mixing between fresh charge and exhaust gas residues (minimum internal EGR). Homogeneous air/fuel mixture.

Much left out but hopefully still of some interest.

Colin

Hi Colin,
I'm so glad Ive been allowed back on to make an active contibution to your post, but sadly Im sure like most Ill still be trying to understand is when I hang up my leathers, fascinating reading some of it, you never spoke about droplet size this must have been an important factor in the combustion system,

I actually remember the year you did this (off season)you were reciting this in th paddock at Snetterton, during one of those talks I use to gatecrash and listen to, or should I say, I remember you talking about what you did and the many long hours you spent doing it. this congregation, of JS,R.I.P Martin Roper, Bob Trickett and myself. "If I remeber correctly" you were asked by JS what difference it made to the performance, or could you feel it, your answer was "NO you laughed and said "nothing, well not a lot", if anything a slight improvement under peek power, but not really noticable".

I was I'm sure about 19/20 years old, but I do remeber it was a reliable and failry quick motor!, well only when you never said this".
it was for sure a lot more reliable, than it has been since it was passed on to those who have used it since.

please pass my regards to Elain, and perhaps you guys would like a morning in the future at one of our midlands Bantam Racing "friendly" meeting, im sure your locals or only an hour or so away.

kind regards
the ever young" Derek Betts

I guess a few basics first.

Liquid hydrocarbons do not burn - the fuel has to evaporate. This is a phase change, so heat transfer has to occur (the latent heat of evaporation). The lighter volatiles will phase change quicker than the heavier end (which includes lubrication oil). Time is of the essence with high engine speeds, coupled with a short path distance from carburettor mixing tube to the cylinder (residence time). Aerodynamic forces on the droplet help break it up but as the droplet is soon travelling at the same velocity as the air the slip velocity is very poor.

In an ideal world, a fully evaporated, fully mixed charge, at the correct air/fuel ratio will fill the combustion chamber at the point of ignition.

Can you determine the life of the fuel droplet?

Detail analysis of the air movement, fuel mixing, combustion and emissions are obtained using non-invasive techniques such as Laser Doppler Velocimetry (LDV). These techniques rely on the fact that fuel droplets are present in sufficient quantities to allow the back scattered light to be detected and so velocity and direction recorded. Techniques such as Planar Laser Induced Fluorescence (PLIF) target specific species within the fuel so the different evaporation rates can be determined and just as important, where they are. To do this you also need an engine with optical access.

This is a wonderfully expensive business and I have seen very little published work on two-stroke engines. However, a quick trawl across some of the most likely suspects; here is a paper from Lund Institute of Technology that is worth a look:

http://lup.lub.lu.se/luur/download?func=downloadFile&recordOId=538521&fileOId=625826

This was not an optical engine and the data is limited to the transfer period. The ‘metal’ engine used does not allow the velocity and direction in the third dimension, so the recorded velocities will be under estimated and the complete spatial resolution of the process not fully resolved. However, of interest is that they had sufficient fuel droplets to seed the flow during the transfer period (~80deg crank). I can only suggest that the chance of achieving a homogenous mix at the point of ignition is anybodies guess.

Also of note is the peek, in-cylinder, velocities at ~300m/s. Fig10 shows consistently high velocities throughout the transfer period. The modest vertical velocities (Fig8, @+9mm, 15-30 CAD after BDC) indicate that the transfer velocities have been effectively killed at the rear of the cylinder.

Related transient engine problems.

One racetrack issue concerns wall wetting. Fuel will drop out onto cool surfaces within the engine and form a film. The thickness of the film is, in part, a function of the local airflow. After periods of closed/part throttle running, the film thickness will build to a stable level. Sudden opening of the throttle body rapidly increases the air mass through the engine and increases local air velocities. This scours the liquid off the walls and dumps it into the cylinder. The net result is that the engine is now super rich and misfires until enough engine cycles have occurred to stabilise the wall film thickness at the new condition and clear the excess before normal air/fuel ratio is achieved.

And, just for the fun of it below is an optical single cylinder research engine before being hidden by the laser kit. Incidentally the rather nice CNC machining was done by Colin Freeman at Harris Performance – so I am not completely out of touch with bike racing.

PS. Had a long letter from Sue Sawyer at Xmas, all very sad.

PPS. I noticed an error in the original piece which I have corrected - I hope this was easy to spot!

Evening all ,
Anyone wishing to have a look at the paper Colin is refering to might do better to Google SAE 960336 . interesting stuff but a little restricted for race applications , and the duct profiles are pretty elementry !

Trevor

Sorry guys that number should have been .......... 960366

Yours in embarrased humility , Trevor

Trevor,

Hopefully I was trying to answer, to a limited degree, Derek’s question about droplet size. For all the limitations of the Lund paper it did show that in a fired two-stroke engine, there was a significant droplet population during the transfer period.

I can well remember the pretty awful cyclic variation the first time I ran a modified spark plug to measure cylinder pressure on the dyno. Achieving consistent stoichimetric air/fuel ratio at the sparkplug at the right time is almost job one, regardless of whether it is a racing engine or not. Any insight that takes even a small level of guesswork out of what is going on ‘inside the box’ is worth having especially if the tools to measure directly are not available to you. In fact it’s a bit like running the engine on a dyno – it tells you where you are and not necessarily what to do next.

Has anyone published any optical work on racing two-strokes?

Colin

Hi colin

Thanks for that, and NO sadly I never spotted the mistake,!!!!!!!! but jolly interesting stuff all the same, the bit about, wall fuel build up was very interesting, but im sure this would only be in "out of resonant phases" outside of the power band/curve, also I coul imaging that a rich engine would produce the same effect, the same nomina i would imagine when an engine boggs down, "you have to hold it open until it clears" perhaps because of the same issue.

more importantly what could you do if you were experienccing this phernomina in the paddock or off the start, that would help avoid this, Im soure its not as simple as an adjustment to the jetting, or perhaps a change in trace length or extra packing under the barrel to change the timing phases, ? certainly over my head.

p.s could you please forward sue's full postal address, please private message it to me.

regards Derek

Derek,

Wall wetting started life as something called shunt and the explanation at the time (1970’s?) was that it was due to the sudden increase in pumping load with increase in air mass flow. However, shunt was not as noticeable in single cylinder engines or multi/single-cylinder Diesels. We did some work at Imperial College measuring wall film thickness in a four-cylinder spark, single point injection engine that pretty well nailed it. The move to inlet port fuel injection almost removed the problem although there is still a film on the back of the inlet valves and I have seen it dripping off the valve lip under transient speed conditions. With direct injection, shunt has disappeared, as long as you don’t have wall wetting within the combustion chamber!

It is not an in or out of resonance issue, you are stuck with it. The worse thing you can do is to try and lean out the carburetor, otherwise as soon as the engine runs clean it will be too lean. The only suggestion I can offer is to note where on the circuit you have a noticeable problem and open the throttle a little more gently so that the film thickness is not striped off in a big chunk - sometimes being slower is faster.

Alternatively, you can do what we did when turbo-charging Formula One engines a few decades ago. When the driver lifted off the throttle, nothing happened to the throttle body, it stayed wide open but the ignition was massively retarded which kept the tubo at full boost - hence the flames out of the exhaust. When the driver pressed down on the throttle, the spark reverted to its normal operation. Oddly enough, this is a common technique for automotive spark engines on cold start to rapidly light off the catalyst. This would keep the film thickness down………

Colin

Colin, thanks again, and please check your private messages,

regards Derek

Colin ,
To my knowlege , optical investigation of , exclusively , two stroke race engines is tiny , and now that current race engines are all four strokes it isn`t going to happen anyway . Those investigations that i do know of , almost all are validations of simulation software results compared to dyno output , but then why would anyone finance expensive research on discontinued engines , at least at GP level , for that is where the money is . What the big Japanese companies got up to is probably going to remain secret to them , which is a shame , for as you so rightly say , any little insight helps .

As Bantams are " garden shed " endeavors it will probably require a certain expertise and fundamental knowlege to be able to accurately interpret what is being shown by technical papers and not all can , or indeed want , to aspire to that .
LDV , CFD and such like techniques have been employed for years , but they concentrated , of late, on clean air and environmental issues and not race engines , i dread to think of the amount of harmful emissons pouring out of the average Bantam tail pipe ! So perhaps , and taking a wider view , our legislators have a valid point ?

So , Colin , carry on with you technical input for it is both facinating and insightful and provides a window into a world few have the opportunity to experience first hand !

Trevor

Hey Trevor & Colin, thank you. I am much appreciative of your inputs -- or is it outputs? -- and have learned more about two-stroke engine tuning on here than from anywhere else before... And it is pleasing to know the Bantam Racing Club lives on... Not that the information is of much use to me now. It is that the holes and gaps are getting filled ín -- which is somehow satisfying. I have worked with non-Brit people (some called themselves `Engineers´-- must not be racist and say Chekky and German...!) who were not backward in plaigiarism when it came to presenting a Paper....

The use of the pitot-tube -- for measuring airflow -- with `motored´, headless engine was really cute and I wonder what really happens to the fuel-air mixture as the swept volume get squashed into the combustion chamber...? It would be wonderful if movies could be made of that happening...

Keep posting...

Cheers from snowy Germany. Nearing end of March and still very cold -- but I do remember a 6th March, Brands meeting being cancelled because a massive snow storm filled up the valley between Paddock & Druids...

Stay cool!

Rduesbury wrote:

What do you think nick?

Hi Rob for the record I think you have done a great job cleaning it up, and restoring.

well done "it looks a credit to you".

Derek,

Cold starting is a problem for any engine (for the automotive calibration engineer that means down to minus 30C). Wall wetting plays a significant role.

Typically the engine will have been shutdown from hot and after a short period the wetted surfaces will have evaporated away or, the engine has been rebuilt, ether way the engine surfaces are dry. Therefore, for the first few engine cycles, fuel is ‘lost’ from the fresh inlet charge as it drops out onto the cold surfaces. Once the film has stabilised more of the fuel in the charge is available within the cylinder. Without fuel enrichment during the first few seconds the engine will not fire as the air /fuel ratio within the cylinder is too lean.

After the wetting phase a smaller amount of fuel enrichment is still required to compensate for cold charge temperatures, due to poor evaporation of fuel droplets. In automotive calibration, this can be achieved by splitting the extra fueling into a ‘fast’, wall wetting phase followed by a ‘slow’ low charge temperature phase. This is tricky to do unless you can part choke the carburettor.

Prolonged excessive fuel enrichment past the point at which the wall film has stabilised risks flooding the engine – don’t keep running up and down the paddock with full choke.

Incidentally, you run a race engine slightly rich of stoichimetric not because you get more power, you don’t, you have run out of oxygen , but because the flame speed is higher. Given that the engine should run ok a little either side of ‘normal’ rich, you do have some wriggle room with carburettor settings.

In the days of push starts, as soon as the engine was cut, the fuel film starts to evaporate off and the longer the starter takes the greater the risk of the engine hesitating. Given long enough this becomes a ‘warm’ start – sort of half way between hot and cold - so loss of wall film but good evaporation.

Alas, you cannot work round wall wetting but you can increase very cold starting performance by preheating the engine to help evaporate the fuel that has made it to the cylinder. You should also ensure that the fuel is fresh. Fuel ages, don’t keep old fuel.

Colin

If you wish to pursue the motoring method to evaluate cylinder filling/scavenging the original paper that I used was by Jante (SAE 680468). After a quick trawl on the internet, I found a 1999 paper based on Jante’s work using a 125 Yamaha. Of note is the large transfer port width indicating maximum effort in reducing mean transfer velocities - hence the high axial inclination to achieve correct timing of the scavenging gas flow.

http://ledob86.free.fr/NouveauSite/Documentations/porting.pdf

For a fairly comprehensive computational fluid dynamics analysis (CFD) with some interesting comments try:

http://www.freidok.uni-freiburg.de/volltexte/2096/pdf/TwoStrokeEngine.pdf

See Chapter 2.7.2 for Jante’s diagrams of the four characteristic flow patterns. You can skip the CFD chapters and go to Chapter 10. It would have been a better paper if the CFD model were anchored to real engine data. It is always good practice to run mathematical modelling in parallel with a fully instrumented engine for validation– some of the assumptions are just that. It is still my view that there are considerable gains to be had in cylinder filling and mixture composition. Beware the file is some 46MB.

So, what started as a few notes on my back to front engine of thirty years ago has grown a little. My work since then holds little relevance to two stroke racing engines. So I shall stop here, however I am happy to answer any questions if I can.

Have a good season,

Colin