Making the Power-4

A short while ago I was contacted by a Bantam racing enthusiast who was considering the conversion of an existing piston port 125 cylinder barrel with the installation of a reed valve assembly. Any help I could offer that could simplify his task would be greatly appreciated, and in particular, how could he calculate the tuned length of the reed valve equipped intake system?
It was at this point in our conversation it was clear this was another one of those exercises that begins with a question which takes two minutes and requires an answer of two hours! However, as I still have the notes and drawings from when I made my water-cooled barrel and head, plus subsequent modifications, the task for him was a little easier, aided by the fact that he was starting with an existing air cooled barrel!
The following is a somewhat generalised look at the inlet side of two-stroke engines.
Mid-way through last century the acoustics of organ pipe theory was all the rage, formulas existed to tune them so why not inlet tract length, then there is the Helmholtz-resonance theory. Dr.Blair, as he was back then, of QUB fame released his work on exhaust pipe theory complete with a formula, and some enterprising types adapted this to inlet lengths, but when it was realised that a tract length of 2/3 the exhaust length was needed the whole thing was dropped. None of this works because the variables are almost infinite, indeed the crankcase volume is virtually dimensionless in that it is only at TDC, where the crank is briefly at rest that the case has a constant volume where there is actual inlet flow taking place and that in itself is variable in dwell time by the geometry of stroke and con-rod length. A further complication is that at the business end of a reed system things get far more complex with reeds being open, closed, and partially open thus presenting a variable length and area within one engine cycle.
The conclusion then is that calculating tract length based on wave motion is pretty pointless; but we`ll have look at what happens inside the engine anyway!
When the transfer port start to open, exhaust pipe diffuser action sucks mixture from the crankcase, shortly after BDC case pressure starts to fall below atmospheric. A suction (negative) wave next moves from the case towards the open air. When it reaches the open end of the tract it reflects and returns to the case as a pressure (positive) wave where, hopefully it arrives just in time for the next engine cycle! So in one crank revolution this wave should travel once up and down through the inlet tract once!
Now comes the trickier bit where numbers are involved, actually I may have done this before but it won`t hurt being repeated here. Just for illustrative purposes I`ll use an inlet speed of sound of 300mtrs/sec for the cool mixture, and 12,000 rpm for peak engine speed, these values give nice whole, round answers.
1 crank revolution will take (60/12,000) .005 seconds, so during this time period duration the wave will travel (300 x .005) 1.5mtrs; once up and once back, so the tract length would have to be 1.5/2 = .750 mtrs. or 750mm. That 750mm length might be theoretically correct but is of no practical use, reality might suggest a suitable reference tract length of .125mtrs (125mm) then .750/.125 shows that the wave will travel each way 6 times. One thing that can be assured is that after a total number of 12 end reflections the wave will have no energy left to work for us! All is not lost however, instead of trying to harness the wave action that travels through the mixture; the mass movement of the mixture can be looked at, and that does involve the Helmholtz-resonance factor.
The mixture mass itself moves much more slowly with its mean velocity about half the speed of sound. So the bottom line is: keep the inlet tract short, doing this has the added bonus that there is less inertia in the standing mixture so is far more responsive when the pressure difference allows mixture movement towards the case at inlet opening.
One thing is for certain however, the type of Helmholtz resonator depicted in reference books, and some two-stroke tuning books, is not one to be found in any engine! A racing two-stroke engine is a compound resonance system because all of the ports are open simultaneously at one time or another, and flow through the engine is determined mainly by the effectiveness of the exhaust system in any case. A lot of other important factors also come into play and you certainly can`t separate these out and compartmentalise or separate the inlet-cycle part from any other.
Some of the features that influence Helmholtz frequency are: a large case volume will lower it, a long inlet tract does likewise, but a large diameter tract increases the frequency.
The large case volume will need more time, and a large carb, to fill it, but more time could mean a later port closure and that will have adverse effects at low engine rpm giving poor carb response. Two–strokes are nice brain puzzle, aren`t they, or as I categorise them, a perverse conundrum?
It seems that the two most intensely talked about topics involved with Bantam race engine preparation are, crank case volume and exhaust dimensions. As mentioned in previous articles these two elements are connected by the transfer ducts and it is the quality of flow efficiency of these ducts that can therefore determine the geometry of both case and pipe.
Case volume (or compression ratio) is a far more complex subject that one might at first think it to be, however a few lose rules of thumb can apply.
The higher the BMEP of an engine then by definition, the higher will be the delivery ratio and scavenge efficiency. For this to be able to happen the transfer ducts have to have a high coefficient of flow, so a large case can store a larger volume of fresh mixture that in turn is available to those free flowing transfer ducts. So it might be said that by default, the quality of transfer port and duct geometry determines both the case volume and pipe dimensions?
In order to take advantage of a large case volume (1.3ccr) the fundamental needs are: a large bore carb to fill it, together with a long inlet duration. Critical, are free flowing transfer ducts of adequate cross section, short column lengths and large time/areas for the port durations! If any of these elements are absent or even restrictive then you will have to reappraise you targets. The WORST flow coefficient of any transfer duct is when the inner wall has any sort of parallel relationship with the cylinder axis, and this commonly seen profile is when the outer diameter of a cylinder liner alone, forms that inner duct profile.
Conversely, a smaller case volume (higher ratio) will provide for greater pressure rise in the case as the piston descends; this will far better suit the compromised transfer ducts that have little or no inner wall shape as it forces the mixture to flow though and around the ducts quicker. But this scenario requires careful choice of pipe geometry that matches those compromised ducts. If the exhaust diffuser sucks too hard on crap transfers leading to excessive velocity of the mixture loop, loss of directional control of the ingoing mixture streams and consequent massive short circuiting mixture loss to the exhaust port becomes inevitable! And, if piston velocity does not match transfer stream velocity you will have real trouble.
We have another Catch 22 situation here in that we need a small case volume to speed up flow, giving a responsive engine, but this potentially limits stored mixture, we do need the pipe to pull the maximum quantity of mixture through the transfers around BDC but not at the expense of flow control and of achieving high scavenging and trapping efficiency.
So what case compression ratio (CCR) is best for a Bantam engine, or is there one? If you look at CCR within the modern 2T race engine context of what is now acknowledged as being “normal” then the range is accepted as, 1.4 being high and 1.3 as being low.
Not much difference you might think, just .1, but .1 of what, is that a lot or a little and will it make any difference? Just having a reference decimal fraction doesn`t really convey very much but if we had whole case volume numbers then comparisons give context and relevance.
By using the formula below the maximum case volume (volume at TDC) can easily be determined.

Case volume at TDC= compression ratio / (compression ratio-1) x cylinder volume.

Once again for the purpose of keeping the numbers uncomplicated a cylinder volume of 100cc is used.
A case ratio of 1.3 gives………. 1.3/.3 x 100 = 4.333 x 100 = 433.3cc
A case ratio of 1.4 gives………. 1.4/.4 x 100 = 3.500 x 100 = 350.0cc
The difference between these totals clearly is 88cc, but this represents 88% of the swept volume of the 100cc engine and a 20% difference between the two and as such, become very significant numbers!

Something else to ponder on: whist the pressure differences in the case move more or less at the local speed of sound, the piston moves at about 10% of that speed, confirming that that the greater portion of mixture movement is down to pipe action and not piston movement. As good duct/port geometry with a high coefficient of flow does not directly rely so much on case pressure, then the greater pressure of the small case may not be required to initiate the required mass movement of mixture.
An omission so frequently seen in “homemade” barrels such as the Bantam is that the porting may have all the raw time/area that on paper is needed, but if the basic rate of mixture flow, due to the limiting effects of rubbish ducts, is not there, then power won`t be either. This limiting effect comes mainly from flat sided ducts with a right angle bend leading into the cylinder that simply can`t flow anything like the predicted numbers would indicate.
Remember the mantra: No flow, equals…… No go!

The following article was written by Prof Gordon Blair and concerns the testing of a 65cc chainsaw engine and the results from that testing at QUB can be found in books and technical papers. This particular short article contains observations taken at the time and with the engine running at 9,600 rpm, and achieving almost 4bar bmep.

THE INTAKE SYSTEM BEHAVIOR
“The intake system is controlled by the piston skirt opening the port at 75* btdc, and provides a relatively constant delivery of 0.5 over most of the speed range.
The system works well in that the delivery ratio increases progressively from zero to its maximum with no back flow, aided by a modest ramming behaviour in the inlet duct prior to port closure. The short intake duct provides those pressure oscillations from the reflection of the initial suction wave, some five or so, and the fundamental frequency of the sound propagated into the atmosphere can be seen to be four or five times the natural frequency of the engine speed, i.e., between 640 and 800 Hz.
The crankcase pressure has already dropped to 0.8 atm by the time the inlet opens, sending a sharp, i.e., noisy, pulsation as an intake wave into the inlet duct, which peaks out at about the tdc position. The ensuing crankcase pumping action raises its pressure to about 1.5atm, aided at that juncture by the higher pressure cylinder backflow into the scavenge ducts.
The air in the crankcase never drops below 60*c, whereas most of the air in the inlet duct oscillates around 40*c. During induction into the crankcase, the combination of incoming air at 40*c flowing into an expanding volume containing air already at about 65*c, decreases the crankcase temperature somewhat to about 60*c. This seems curiously insufficient as a drop in temperature; however in terms of the mass of air already within the entering quantity is quite small. Put crudely, the DR (delivery ratio) value is .5 or about 32.5cc in volume terms for the 65cc engine. The crankcase compression ratio is 1.5 so its maximum volume is 195cc, therefore the entering quantity is only some 17% of the total in residence.
As the air is gulped into the intake duct the temperature around the tdc point briefly drops to the atmospheric value. The peak temperature of the crankcase air rises to 120*c and can be heated in the cylinder up to 600*c during the early stages of scavenging. Not surprisingly, the largest majority of fuel vaporisation occurs within the cylinder”.

Trevor

Despite having Jan Thiel on board for a short time, even he despaired of the thing and decamped to Bultaco to win more championships

Trevor

Found it, actually I think this picture would make a good candidate for the Wacky files!

With the disc valves angling the inlet charge in exactly the wrong direction and the huge standing waves in those long cavernous tracts small wonder the machine was a monster to obtain consistently clean carburation. I did also read that the toothed rubber belts under the castings driving the discs were prone to stretching and thereby adjusting the valve timings from cylinder to cylinder. In all a sad tale of a miss guided pursuit for Grand Prix glory.
Trevor

The following is a sort of post script to the original questions prompting this topic, and arises from subsequent queries, putting a little more flesh on the bones so to speak!
Six petal reed blocks allow for enormous scope in juggling with reed types to optimise inlet flow for individual engines. For instance, you can bias flow to the top or sides of the block using different reed material and thickness, by substituting the metal reed stops for stiffeners (back-ups) and even juggling the stiffener thickness and length, and trying spacers under them. The permutations are endless, some will be felt to work on the track or seen to work on the dyno, and equally, a lot may make no discernible difference at all! One interesting set I have seen for a specific set up was where a single blade petal with double back up was on the bottom of the block, with three reeds with a single back-up on top, and it worked very well indeed! Just goes to show that mix-n-match may be the way to go for individual engine needs.
It may also be the case that reed experiments require exhaust pipe adjustments to optimise any benefits, for it is overwhelmingly pipe action that influences reed motion.
So what can go wrong with an induction system that is controlled by reed movement? Not a lot really, there is are broad range of frequencies that allow for efficient operating within each engine cycle.
Too low a frequency (Helmholtz) will prevent all of the kinetic energy of the inlet mixture flow converting to pressure energy before the next cycle begins. Conversely, too high a frequency and flow direction can reverse before maximum case volume at TDC is achieved!
Between these two extremes it all works pretty well.
The choice of reed block assemblies is quite vast and some can be very expensive, if it is of any help to other would be reed “convertees”, my w/c Bantam engine used a I25 CR Honda block in conjunction with standard RS Honda plastic reeds, over time we ran through the whole gamut of available reed types and combinations but the RS jobs did it overall for us.
During 2015 Steve tested a number of exhaust pipes on his RS during an extended dyno test session. The best performing pipe was then paired with a brand new Vee-force block and reeds replacing the standard items. Whilst the hype surrounding these after-market assemblies suggested improvements were to be found in their use, the reality of the test session showed otherwise, no worthwhile performance increase was delivered, which both surprised and disappointed us!
Getting back to Bantam engines, it should be continually born in mind that they produce only modest power compared to other competition engines of similar capacity. This being the case, it is necessary in development terms, to visualise the whole package and not simply apply certain aspects of a far more powerful engine. For instance, on a good 12/14 bar race engine exhaust pipe action can pull pressure at the exhaust port down to around half an atmosphere, which relates to about ¼ bar at the transfer ports leading to the crank case. If your Bantam cannot replicate those pressure values then the reeds will respond quite differently, simply copying the reed system from that engine will never produce similar results. An entirely different regime may well offer better results.

Trevor