Developing Curves

Developing those Curves

This article is ostensibly concerned with ignition timing relative to the point of crank rotation and its rotational speed where combustion is actually initiated.

It is always helpful if an actual engine of known performance is used to illustrate the significant points, one can immediately identify those with one`s own engine`s performance.
So I will use my own Bantam engine, a brief overview shows it is water cooled with reed valve induction and 54mm square dimensions. The 96mm o/d crank has 25mm dia stub shafts integral with the crank discs and runs on larger 6205 main bearings. Initially the gear box fully ran on needle rollers for the lay shaft and the main shaft sleeve gear also had needle rollers fitted. The clutch body, using case carburised, hardened plates, was made from an EN 25t billet, fully lightened along with the skimmed up chain wheel and the whole assembly was dynamically balanced. So, a pretty bullet proof bottom end that aimed to minimise friction and flexure losses, and maximise reliability. Somewhere amongst my forum posts are images I posted of the barrel components and drawings of barrel/ liner construction, I always make working drawings, the universal engineering language. I started all of this on paper in 1993 and we hit the track in victorious fashion during 1997.
So with all of this success what could possibly need altering? Nearly all of our close competitors had the advantage of retarding ignition systems, we did not. Back then, these things were not cheap and the overall costs involved in building a new engine are considerable so to eke out our meagre resources the existing Motoplat unit was used but it had the very real handicap of no automatic advance/retard facility.
What then are the benefits of an ignition system with a varying ignition timing? The simple answer is, a fuller power curve is created. That then is also the short answer.
If nothing is changed on the engine the maximum power produced remains the same, in the over-rev area performance drops off and at around two-thirds of peak torque rpm there is a gaping hole in output performance.
It is in trying understand why those two events, separated as they are by several thousand rpm, happen as they do that we start with scavenging of the cylinder contents. However, all has to fit and be synchronous so ignition analysis must follow a little later.

It has been a more complete understanding the physics of how it`s various dimensions of length and diameter function that makes the exhaust system of the modern two-stroke race engine into the powerful force that it has become., it is the interplay of positive and negative pressure waves that achieves this. Back in the 60s it was thought that sky high crankcase pressure was the answer to mixture movement, yes it was, straight out of the exhaust port resulting in a razor thin power band! The exhaust pipe now takes care of that function most effectively.
Exhaust pipe action sucks on the contents of the cylinder during the blowdown period. Hopefully by the time the transfer ports open all combustion gases are out and fresh mixture flows from the transfer ducts into the cylinder. Some of this fresh mixture also gets pulled into the exhaust duct and header pipe. If the engine rpm and the speed of sound of wave action all match, the pressure pulse changes the flow in the exhaust duct and that, over scavenged mixture, is pushed back into the cylinder. Eventually the exhaust port closes trapping the augmented mass of fresh mixture in the cylinder.  
There are a couple of instances where engine rpm, exhaust pipe length and the speed of sound become out of sync with each other.
These are; if the rpm is too high or the pipe is too long or the speed of sound is too low. Although mixture is still pulled into the exhaust duct the exhaust pulse begins far too late for this speed, the exhaust port closes before the fresh gas in the duct and header can be returned to the cylinder. It is this late action that affects over rev performance so badly and power fall off rapidly.
Conversely, if engine rpm are too low or the exhaust too short or the speed of sound is too high, the engine responds poorly. Again the cylinder may be completely scavenged, and the exhaust duct may contain enough mixture to charge the cylinder, but back flow begins far too early for the rpm, when the transfer ports are still open. The overpressure from the returning pressure pulse returns mixture into the cylinder, straight into the transfer ducts and the crankcase. When the transfers eventually close there is little pressure in the cylinder, but there is a higher pressure now in the crankcase, the subsequent inlet cycle has to battle this high pressure, with predictable consequences. Back flow has not only started much too early it then come to a standstill and again reverses direction, so what little mixture remains in the cylinder is sucked out, eventually the exhaust port closes. It is no small wonder then that at low rpm there is a major shortfall in torque production.
There are a couple of ways to influence exhaust flow at both high and low engine speeds: change the exhaust length or the speed of sound, the former is impractical and mechanically quite complicated, the latter is far easier and works by modifying the exhaust gas temperature.
In GP level 125 engines, combustion chamber temperatures can reach 2300*c but with the effects of expansion and cylinder wall proximity the gas has cooled by the time the exhaust port has opened. The rate of expansion can vary, but begins immediately after combustion has finished where cylinder pressure is at its maximum and continues until the exhaust port opens. The quicker that combustion is completed, the greater the subsequent expansion and the cooler the exhaust gas is when exhaust out-flow begins.
The duration of the complete combustion cycle depends on two major factors, namely the ignition timing and the burning speed. Burn time depends on the quantity, too much or too little fresh mixture, the quality, how much residual exhaust gas lingers to dilute the charge, the ratio of fuel to air, rich, lean or just right, and turbulence from squish action.
Combustion must always be as rapid as possible, which implies a compact combustion chamber and an effective squish action with minimum piston to squish band clearance at peak rpm.
Influencing the exhaust gas temperature enters the realm of ignition timing, it has taken a while but we have finally arrived back at ignition timing!
As an over view, at low rpm if we need to reduce the speed of sound, thus reduce exhaust gas temperature we can achieve this by advancing the ignition. Conversely for high rpm where the exhaust is too long we can compensate with a later, retarded ignition timing.

In 2004 I invested in a copy of EngMod2T which is today the most comprehensive engine simulation software available at an affordable price. Using this facility I simulated my Bantam engine running the Motoplat ignition starting the sim at 6000rpm where the engine peaked at 11000rpm. In 2011 Steve had accumulated enough second hand bits to assemble a 125 Honda RS ignition system. After a fair bit of engineering including a new timing side crank main shaft, it was all fitted to the Bantam and subsequently ran extremely well. Not knowing the actual timing points from the CD box I made an appeal on FatBac and had a terrific reply from an American karter who ran RS engines. He had access to some electronic gizmos that produced, in 100rpm increments, a full timing curve for our year and part numbers for the CD box. The obvious action to take was to initiate a new sim with the RS timings, all other parameters remained as for the Motoplat. I superimposed the two power curves on one readout. The difference was startling, at 12hp for both curves the RS was 1200rpm lower, both outputs coincided at 11000rpm but whereas the Motoplat flat lined for 500rpm then fell, the RS gained 1-3/4hp and another 400 rpm and then power fell away less steeply. On the race track Mark was able to confirm the sim findings in the extra breadth of the power band citing much better acceleration from the start and from low gear corners. The power gap that occurs at around 2/3 of torque on a straight line ignition is very clear to see. If you are so inclined you can divide the base rpm, 7,000rpm and peak of 11,000 curved and the answer comes out at 63%; pretty much on the money.  

Should anyone like a copy of the print out and the RS timings then send me your email address by a PM or on the open forum and I’ll send you a copy, makes interesting scrutiny and perfectly illustrates the assistance that variable ignition bestows to our gear starved Bantam engines.
I used to post a lot of images on the forum but Photo bucket demanded $400 for the continued privilege and so I declined.

There is another episode to this article but comes in the form of a series of vignettes which if included in the main text would create a huge amount of work and string out the main thrust of the post. I`m sure with a little bit of thought these short `bits` can be mentally slotted into the relevant passage in the above text.

Take great care out there, this evil bug has not finished it`s pernicious work and lies in wait for the careless public.
Trevor.