| Villiers Handbook | |
Authors
| |
|
Introduction
| |
|
History
| |
|
Crankcases
| |
|
Crankshafts
| |
|
Pistons
| |
|
Primary Drive
| |
|
Gearbox
| |
|
Ignition
| |
|
Induction
| |
|
Exhaust
| |
Silencing
|
Pistons & Cylinders
Pistons are many and various, but watch out for changes in the ring peg positions as some less sporting types may place the pegs into the larger ports of a tuned cylinder. The original circlips used to retain the gudgeon pin are adequate for normal usage but have a tendency to break at the eyelet holes during continuous high speed, and should be replaced with ordinary wire circlips. The circlip grooves being recut to a half round profile to suit. To further prevent mishaps the ears of the wire clips are cut off after fitment. If the Seegar clips are to be retained they should be replaced each time the gudgeon pin is refitted, and positioned so that the sharp edge is facing the cylinder wall.
During its lifetime the Villiers engine was fitted with a large variety of pistons that sported an even larger variety of rings. originally they were manufactured with thick rings, which became progressively thinner as power outputs and rpm increased. With its single ring the high domed Omega piston is without doubt the best available for a sporting or racing 197. The 197 range of Omega and TKM pistons, which have large transfer cutaways, are available in steps of 10 thou, up to a maximum of 1.75 mm over the standard 59 mm bore. Having a high silicone content, these pistons can be run with a low skirt clearance, 3 thou for the Omega or 1.5 to 2 thou for the TKM variety. (these are advisory clearances specified by Invader engines), and apply to well cooled alloy barrels. For cast iron barrels the clearances should be doubled. To eradicate gudgeon pin problems these pistons feature larger circlips, which stay in their grooves even under racing conditions.
Many 250 pistons are still available, from the original Villiers with the fat rings and single peg, through the thin ring types with two equi-spaced pegs, to the Dykes ring racing types. The latter being available from Terry Silvester and Peter Hepworth, which are believed to be of Italian ASSO manufacture. The gudgeon pin supplied with these pistons has a large diameter hole through the centre, and is not really suitable for racing as they have a history of breaking under stress, and should be replaced with one that has a smaller diameter hole, which should be sealed as discussed earlier. By modern standards the gudgeon pin looks very thin at 0.5 in, and can be replaced by a cut down Greeves Oulton 16 mm pin, a tactic used by Brian Woolley on 250 Silverstones to alleviate the problem of the phosphor bronze bushes moving in the piston. Gudgeon pins for the 9E are not without their problems, for in the early nineties a batch were produced that were too hard and fractured under racing use.
Clearances
Piston and ring clearances are critical if the engine is to run correctly in the normal or competition role. Modern pistons are not round or cylindrical when cold, they are wider at the bottom and thinner across the gudgeon pin holes (see Fig 7 ). When hot, the piston is hottest at the top and front, it changes shape - but so does the cylinder. To counteract this a specific amount of cold piston skirt clearance is given, depending on the job in hand. On a modern liquid cooled motor, that suffers less cylinder/piston distortion than our air cooled type, a skirt clearance of 1.0 to 1.5 thou can be used, but this would be a disaster on a 9E, resulting in frequent seizures. A clearance of 5.0 thou is specified for a standard tune motor, which should be increased to 6.0 thou for racing. The reason for such a large amount of clearance is the heat which distorts both the piston and cylinder. Modern cylinders are not uniform in their construction but are thicker in certain places and thinner in others according to the heat profile of that cylinder, which equals out the distortion, this ensures that the heat retention and transfer to the casting is controlled and distortion minimised. The Villiers cylinder was not manufactured with this idea in mind.
Piston rings also demand a specific clearance to ensure that they do not nip up in the cylinder when hot. Don't be tempted to fit a new set of rings without first gaping them. Always check the rings in the bore, the gap being 0.5% of the bore diameter. The gap specified by the makers is 8 to 12 thou, with a maximum clearance of 30 thou. Another critical clearance is the ring clearance in the piston groove. Given as 4 to 6 thou this clearance should not be exceeded or damage will occur, and it must not be confused with the dreaded "ring flutter". The bigger the gap, the harder the rings will hit the groove faces, resulting in wear which means even bigger gaps and even more wear. This will eventually result in compression loss and ring breakage.
Ring flutter
To understand the piston ring flutter phenomenon we must discuss piston speed and acceleration, which are related to conrod length and stroke. While the trend today is for square or over square motors, there is much to commend the older long stroke two stroke, as they are able to utilize a more compact expansion chamber design which makes for more efficient combustion and lower thermal loading on the piston, and the smaller bore in relation to swept volume presents a smaller piston crown to absorb heat. The long stroke engine usually concedes to the short stroke design because of the short strokes ability to rev much more freely.
The biggest problem in crank train design is piston acceleration and the effect it has on piston rings. It is often thought (without due consideration) that the ring seals against the bore by virtue of its spring pressure. With a combustion pressure of over 700 psi it would easily over come the ring spring pressure. It is the gas pressure on top of the ring which pushes the ring hard down to the bottom of the groove, and it is this pressure on the inside of the ring which pushes it out to the cylinder wall. At very high accelerations the forces lift the ring off the bottom of the groove which allows gasses to get under the bottom face of the ring (see Fig 8 ). Now that there is no differential pressure sealing the ring, the full force of the combustion gasses burst past the ring on both sides. This is all over in an instant and at the bottom of the stroke if not before, normality will be restored. The short blast of high pressure high temperature gasses have heated the skirt of the piston and burned away the oil film that was lubricating the piston, a situation which obviously favours a seizure. The initial research was done by Paul Dykes (of Dykes ring fame) who determined that thinner rings are less susceptible to flutter than the fatter rings. Based on his research he designed an "L" shaped ring which could never suffer from flutter. The onset of ring flutter is determined from ring width and the maximum piston acceleration during its cycle.
For example :-
| 3 mm | ring flutters at an acceleration of | 40,000 ft/sec squared |
| 1.5 mm | 80,000 | |
| 1 mm | 140,000 |
| at | 6000 rpm | maximum acceleration is | 58,700 ft/sec squared |
| 7000 rpm | 80,000 | ||
| 8000 rpm | 104,000 | ||
| 9000 rpm | 132,000 |
As a postscript to this section, it is worth mentioning an early idea to overcome piston problems (see Fig 9 ). Radical as it was the reasoning was sound, and it came from the drawing board of the Cotton Motorcycle company, well known for it use of the Villiers product. By far the strangest piston ring arrangement ever devised was that fitted to the Cotton Cougar announced in October 1961. It comprised of 2 spiral wound rings that interlocked with each other, and when fitted to the piston formed a 7 stage gas seal. The piston was of the semi slipper type with extra high carbon steel rubbing pads riveted to the skirt, and ran in a liner-less light alloy cylinder, few if any actually remain.
continued....
These sections appear in the book
Cylinders
Alloy barrel identification
Lifting the barrel
Exhaust port shape
Transfer port shape
Very little is usually written about the transfer ports themselves, except that they are largely responsible for the swirling scavenge action which takes place in the cylinder. We take this to mean that little is known in the motorcycle tuning world about swirl, certainly the authors are unaware of any good 2-stroke tuning guides published since 1983 (or containing data after 1983), and hence the good work undertaken at Queens University Belfast has been largely unreported outside of the academic world.The normal advice given is to "leave well alone". We would modify that advice to be "leave well alone unless you have the machining ability, or you give it to a machine shop which has the ability to cut metal accurately in a very confined space".
The published books will contain only the classical theory of swirl, which is based on acoustic principles. Acoustic principles deal with infinitesimally small waves in an elastic medium. This hardly seems to describe the situation inside a 2-stroke engine. The classical theory was perhaps finally laid to rest by a paper from Dr. Gordon Blair in 1988 where he concluded that
"... only the worst scavenging [test] cylinder comes close to the theoretical prediction, but even that has a steeper scavenge ratio graph than is theoretically possible. The significance is that the real scavenging performance of all test cylinders is superior to the classical scavenge model. In short, the classical theories of scavenging underpredict the time behaviour and efficiency of two stroke engines"The process of swirl is better described by the theories of unsteady gas dynamics, and QUB have devoted a great deal of time since 1988 to developing and refining their ideas. They have already published a number of computer programs which model parts of the two stroke process, and the authors are following their progress with great interest.
The work by Blair, Kenny and others at QUB showed that the velocity of the gas across the width of the transfer port was virtually constant, but that the direction of the gas stream varied at all points in the cycle.
For a test cylinder, having a transfer port height of 12 mm the results were
| Port open mm | Angle one (Horizontal) | Angle two (Vertical) | |
|---|---|---|---|
| Fully open | 12 | 10 | 14 |
| 10 | 10 | 16 | |
| 8 | 14 | 19 | |
| 6 | 18 | 26 | |
| 4 | 21 | 33 | |
| 2 | 25 | 34 | |
| 0 | - | - |
We see that the flow bends upwards and forward towards the exhaust port as the cycle proceeds and the piston rises to shut the port. In their experiments, Kenny, Blair et al tested six different transfer shapes, the best two are presented here. Shape "A" (Fig 12 ) gave the best performance at all rpm, but was only marginally (2% - 3%) better than shape "B" (Fig 13 ), which is very similar to the shape of most Japanese ports, and is more easily achieved adapting an existing cylinder than shape "A".
The authors have condensed some of their work into the following description of what happens during the cycle from transfer port open to transfer port closed. The description is based on a transfer port which opens at 60o BBDC and closed 60o ABDC.
| Degree open | BBDC | Action |
|---|---|---|
| +0 | 60 | Nothing happens! |
| +10 | 50 | Jets from the individual ports are forming but here is as yet no flow above the height of the port |
| +20 | 40 | Jet front extends up the rear wall nearly to the cylinder head but some short circuit of the new gas out of the exhaust port has already started (concentration of new gas is above 1%). It is really depressing how soon that starts! |
| When a jet of gas penetrates an almost stationary medium, its velocity changes in accordance with Newton's law of momentum. The wave front takes on a mushroom like appearance, and when the overhanging edge curls round and enters the exhaust port then short circuit waste of new gas starts. The short circuit generally starts in earnest about 25o after port open, and continues for the remainder of the port open period. The minimum effective distance between the exhaust port and the transfer port has been determined at 6% of stroke (about 4.5_mm on our 72_mm stroke engines). This is some 50% larger than the rule of thumb passed down from an earlier era. Changing the primary compression ratio, or the main compression ratio, has no measurable effect of the swirl-scavenge action. | ||
| +30 | 30 | New gas reaches the cylinder head |
| +50 | 20 | The cylinder is now about 45% new gas by volume. At the cylinder head and at the back of the cylinder the new gas has achieved a concentration of 80%, but the new gas concentration at the exhaust port has risen to 30%. Jets are still in full flow, and the charge of new gas is starting to loop over the cylinder. |
| ABDC | ||
| +90 | 30 | The loop is clearly defined. The cylinder head, the rear and sides of the cylinder are now over 80% concentration, and the exhaust is at 50% concentration of new gas. A core still exists at only 60% concentration within (and trapped by) the looping gas jets. |
| +120 | 60 | Exhaust concentration is now 70%, hopefully with gas rammed back into the cylinder by the exhaust. It is surprising how much old gas still remains from the previous firing cycle. The enclosed core of 60% new gas and 40% old burned gas is still very evident, and is about 10% by volume of the cylinder. |
| True compression starts now | ||