Do Bantams really lay chocholate eggs?

Design Analysis of a Bantam Race Engine
Part 5.
The Cylinder Liner                            
The production of any new, or as in this case, a replacement liner, is without doubt the most crucial task of any in creating a new Bantam race barrel, at least, that has always been my enduring experience. One mistake can ruin many hours of intense work.
Reactions to perceived performance levels are progressive, some seemingly elementary details are initially omitted, but may become more significant with increasing experience and expanding knowledge. This then was the background, from gentle, distant nudges to an insistent clamour for a new liner, enabling a remodelled porting regime, thereby reinvestigating hitherto spurious old rejections, to substantially embracing and promoting new concepts and developments, especially, with the transfer ports!
Fundamentally, the transfer ports cut through the liner wall enable the flow of fresh mixture into the cylinder, guiding that mixture inwards, up over the piston crown and thence filling the combustion chamber. This inflow also helps to push remnant, redundant hot exhaust gas remaining from the previous cycle out from the cylinder into the exhaust pipe by way of the exhaust port. Regrettably, but inevitably, some of the new mixture will combine with the old stuff and be taken into the exhaust duct and the start of the header pipe. How much and how far is determined by the energy of the pressure pulse and exhaust port and pipe influence. The physical dimensions of the combined exhaust passage (duct and header) exert considderable influence, excesses in both are always detrimental to overall performance. Critical to the effectiveness of the liner ports are the pairs of transfers designed in conjunction with each other and with the ducts that feed them from the crankcase, and so back to atmosphere at the carb bell- mouth.
Before any of this can happen of course, we have to have a liner to hand, my preferred course of action is to make our own from cast iron tube or bar. Some while ago we managed to secure several off cuts of liner sized billets of `Meehanite` which is the trade name for spheroidal graphite cast iron. Ordinary grey cast iron has its grain structure in flakes, but the addition of a tiny amount of alkaline silicide’s to the melt creates grain structure in spheres, that is stronger, more wear resistant but aids machinability.
As an engineering exercise the manufacture of a bare cylinder liner is quite straightforward, after all it is not much more than a basic tube with a flange at one end. As is customary, I made a suitable drawing, dimensioning it to all the finished sizes but left 0.010” on and in, for a final finishing skim after all the ports and cut a-ways were milled in. Meehanite is a very stable material not prone to distortion during machining, but leaving a final sizing cut seemed prudent, particularly so when considering how many hours of effort was going in to making the thing.
I liberally painted the outside of the liner and the flange top with engineers blue, marking out lines are clearly visible using this medium. The first lines to be established were the centre lines including extending them across the top face of the liner. BDC was also marked on inside the bore and out, that is stroke plus .5mm squish clearance and the .010” final skim prior to liner installation.  The liner was initially made over long and a short 10mm long step was machined to just slip into the barrel bore, this enables the two sets of centre lines to be checked for co-axial alignment, which happily they were. Picking up on these centre lines provided basic datum`s for the ports to be machined from, thus ensuring positional symmetry to a high degree of accuracy.
Radially asymmetric ports cause mixture streams to swirl around inside the cylinder bore, some mixture will find the exhaust port and disappear, declining to take part in the subsequent combustion party! It also follows that the larger the export area is, the greater the opportunity for mixture short circuiting. An exercise here then, in proportion, balance and compromise.
The liner I shall describe in the following article is the up-graded replacement of the 1990s original, featuring the larger flange diameter`s contact area with the new cylinder head and modifications to all of the cylinder ports.    
When eventually the decision was made to go ahead and replace the original liner with an improved version, the first assessment was to re-establish the engine`s power potential against the engine specification. What makes it perform as it does and what we think holds it back, what are its strengths and weaknesses, are there any non-productive, `neutral` areas, or what could be described as the weakest link in the existing power delivery?
There can`t have been too much at fault however when that original engine specification is set against a history of a lot of race wins and some lap records. In spite of that I felt there were a number of incremental porting changes which could be made to hopefully liberate a touch more peak power and broaden the low to mid-range performance, I have always tried to bias power characteristics in favour of power band width and not absolute power numbers at excessive rpm. The much later change from Motoplat to an RS Honda retarding ignition proved to be literally a” liberating” experience such was the release of midrange performance.
The first port to be scrutinised was the exhaust, after much analysis and number crunching the plan of action was for an increase in blowdown area, but not by a timing increase. An area reduction of two mm before bdc to help minimise charge loss during mixture transfer, was also planned. It is now accepted that the main exhaust event is all but over by the completion of blowdown.
You can have all of the appropriate numbers in place but if yours is a single port then the flow area, down to blowdown`s angular rotation to commencement of scavenge flow, will be restricted, as rpm rises adequate blowdown can become a struggle for a single port. Some of the recognised benefits of reduced port area below blowdown can be characterised as follows; better blowdown flow coefficient, stronger exhaust pulses due to less disruptive turbulence losses, less short-circuiting of fresh mixture and improved flow potential for returning mixture, there are others but I feel the case made, is quite compelling. It is the blowdown period which counts, not the total exhaust port area and both the exhaust port window and duct passage shape and volume interact with the exhaust pipe!
It is the quantity of exhaust gas which defines how much time/area needed to evacuate the cylinder, if the transfers open before the cylinder pressure has dropped below the scavenge pressure then the open transfer ports will be happily used by the hot exhaust gas as just more exhaust ports and enter the transfer ducts. We then have a compound problem, subsequent transfer from those ducts is delayed and what fresh mixture was waiting is in part contaminated and pre-heated. Reversing that unwanted back flow cannot begin until the pressure delta equalises, all the while the pipe`s negative effects is trying to suck the good/bad mixture out into the header and beyond. Blowdown then, has a huge effect on the scavenging regime, and as rpm rise so the exhaust port time/area reduces and the situation deteriorates and eventually power falls of the cliff edge.  
By way of an example, I calculated the blowdown area of a typical single port to compare to my water cooled bridged port in its later `improved` spec. I used the same timings for exhaust and transfer on both examples, the single port width was a generous 70% of the 54 bore dia at 37.8mm and the top corner radius was 10mm. The calculated area came out at 450 sq mm, I repeated the exercise but with a smaller corner radius of 8mm, this second area calc came out at 467 sq mm. The water cooled exhaust blowdown area is 590 sq mm, a difference of 140 sq mm, and 123 sq mm respectively. It is also important to remember that until the top corner radius has been cleared by the piston crown edge the maximal flow area will not be realised. The bridged alternative has top corner radii of 5mm, which I think speaks for its self.
Mass fluid flow is proportional to the open area, twice the open area gives twice the flow, providing the flow coefficient remains substantially constant. In principle we want as much mass flow as possible during the blowdown period so that all the old hot, useless gas is removed before fresh mixture starts to fill the cylinder. Pressure release is a critical function of blowdown, it would be nice if cylinder pressure at transfer opening had dropped at least to the equivalent of case pressure, but sadly, it rarely happens, flow delay from and reversal into the transfer ducts and crankcase is the unfortunate consequence.
 As the piston starts to open the single port we have the maximum pressure delta over the smallest area so the initial flow can go sonic, but is short lived as the cavernous volume increase of the duct expands and reduces the energy of gas particle velocity and wave speed. Interestingly, the only time the gas particles and wave move at the same speed is during sonic flow.
I achieved the blow down area increase of the bridged port by widening each ‘half’ by 1mm, and reduced the previous corner radii to 5mm thus achieving a useful area increase in blowdown. There is always the temptation to go carving great holes in a barrel but as we know, a great big port also allows the good stuff out as well as the old junk! The same applies to big, pipe diffuser angles and diameters that have such a significant, negative suck on the cylinder which can destroy any attempt at transfer flow symmetry and large quantities of fresh, over scavenged mixture disappear out of that large, accommodating exhaust port, never to be seen again. This is particularly true for poorly designed transfer ducts that do little to control flow movement.
Taking an imaginary line around the roof of the `A` transfer ports adjacent to the inward curving exhaust port edge, the approach was to reduce the port area below that line and increase the area above that line.
It is tricky to definitively declare that the exhaust port modifications made all the difference in finding the extra performance we achieved, but in the overall package of up-dates there can be no doubt that they played a significant role.  
The inlet port had a bridge included for additional piston support with an area increase to compensate for area loss of the bridge. It seemed to work well for after running, a clear image of the bridge could be seen on the piston skirt. The rear (C) transfer port was not bisected by the bridge so at port opening incurred no flow restriction. With all inlet timing controlled by reed action reacting to pressure influences within the barrel and atmosphere the actual port is reduced to little more than an access hole, indeed a port as we know it is not even necessary, as in the case reed RS Honda engine. And of course the reeds will be open for a far longer period of crank rotation than a piston port could ever be, with exhaust action pulling mixture into the cylinder via the (C) port shortly after bdc, directly from the carburettor upon re-opening of the reeds.
The above is part of the liner article but in its entirety it will be large so this first 2,000 word instalment is just the opening gambit!
Enjoy, catch you all later, Trevor.