Nigel

Hi Nigel,
You didn`t seem to get much of a response on KB so here is something about what I feel is happening to transfer mixture flow during detachment. Due to its inertia effect the flowing mixture has no option but to closely follow the outer wall profile. But that is not to say that just any old shape will work as well as a smooth, neat set of radii. However, the inner wall shape is super critical when trying to maximise the flow coefficient that influences how much mixture makes it into the cylinder. In most ducts the inner wall radius tightens as the duct gets narrower towards the cylinder port outlet. As a result the flow velocity increases, and so also will the tendency for flow to detach from the inner wall. If it does detach it will continue on its way, resisting the need to turn into the cylinder port window, and collide with the outer wall stream in a violent, turbulent fashion, consequently bulk flow is badly affected. So in theory, the inner wall should have a larger radius towards the port window but that is very tricky to incorporate into a standard cylinder casting. The compromise is incorporate a more constant radius to the inner wall that compliments the outer duct wall profile in a coherent, constantly tapering section ending at the port window.
This is a massive topic and in the Bantam world we are just scratching the surface. I tried to design in some improvements my w/c barrel back in 93. One thing we did find was that running the engine with the inner wall inserts in place produced more power and in a more user friendly way.

Trevor.

Hi Trevor,

I found that very interesting, and it reminded me of the differences in transfer duct design between the TZ250 and 350 (early ones) - As the bore sizes were 54mm and 64mm respectively, but cylinder spacing was the same (common crankcases), on the 350 there was much less depth available for the transfer ducts. I assumed at the time that this was why the specific power output of the 250 was higher? But they both had good inner wall designs.....

When I was playing around with the RS125/250 in 1994 I recall it was noticeable how Honda had taken advantage of the greater space available for their transfer ducts.

Good to see this forum is still active!

All the best

Pete

Trevor, thanks for giving a more detailed answer.

(edit) Trevor, transfer port duct length, if a scenario of different length transfers appeared in a barrel, maybe A ports 50mm long and B ports maybe 60mm long what sorts of issues would that cause and could this be dealt with.

Trevor, the entry and exit diameter of the transfer port.... i seem the think its a factor of 1.3 for some reason? eg 10mm enrty 7.6 mm cylinder face exit?

Nigel,
             Recently I stumbled across some old data sheets, one showed where I taped off the transfer ports in the cylinder wall and measured the volumes of the 4 main ducts in the barrel of my w/c engine, and was surprised at the figures. If the sum all of that mixture had been able to be transferred and retained then the power of the engine would have been hugely more than was actually made, I wish! There are two realistic options that could explain this puzzle, one that the residue went directly out of the exhaust port, without making any contribution to combustion, hence power, or the ducts were never that full in the first place? I tend to favour the latter.
If the delivery ratio of a race engine is examined and evaluated against the swept volume mixture flow of those ducts plus the extra flow from the 5th `C ` port you also come to the conclusion that the mixture sitting in the ducts is more than going into the cylinder and then trapped there.
If then the ducts are only partially filled there can be no firm conclusion made as to their required physical size, so respective ratios of in and out are not critical.
The mass of mixture in a duct equals the specific mass x duct length x duct cross section area. To be able to move that mixture into cylinder in the limited transfer time available requires a certain flow velocity, and the velocity is inversely proportional to the cross sectional area.
The case for an optimal ratio of duct entry area to port window is rather inconclusive, the `B` ports in my w/c barrel have in fact a ratio of 1.3, but the B ports in the more recent Aprilia barrel have a ratio of less than unity where the cylinder base entry is 304 sq mm and the window 324 sq mm. a ratio of .935!  Can they both be correct in two engines of the same capacity, these figures seem strange, even contradictory.
Perhaps this is why the current practice, particularly for lower bmep engines, is that ducts have been getting progressively smaller. The stored mixture mass has less inertia to overcome and flows to the cylinder more quickly so for the same time duration and the same pressure ratio across the ports, more mixture can be transferred.
I`ve seen ratios from 1.5 to .935 in different 125 cylinders, of different bore and stroke so there is no rule that establishes an optimum ratio, however if you are pushing 14bar then you will already have got things right!
Trevor

Trevor, thanks for that information. whilst im laid up with back problems again, im re-visiting barrel characteristics. Theres so much mis-information out there it can be confusing at times to decide what to believe. It seems that only info from the horses mouth of the builders/developers should be taken on board. Can you confirm that the pumping function of the engine is only required to start an engine and once fired up the exhaust pipe action is only then required to keep the engine running (reed valve engine) did you also model your w/c engine on the honda rs 125 barrel of the day? bridged exhaust port/oval stepped exit?
thanks ,nigel.

Sorry to hear you`re crocked up with the old bad back again Nigel, take it easy and recover soon. Mis-information, dis-information, absurd Bantam lore and down right alternative facts and much more! One reason a revised Bantam tuning manual would have been so useful, where the absurd could be separated out from the facts.
I have to drive into Wells today but will get something done a bit later.

Rest easy, Trevor

Nigel,
Lawn mowers rely on case pumping to get started and to keep running, but they usually are running at a fairly constant speed and have a totally non-functioning exhaust being required only to act as a silencer. A race engine is in a somewhat different situation in that its resonant exhaust pipe will only fulfil that function at the rpm that it was designed for, but then that’s what we have gearboxes for. Bantams have only half the number of gears that a contemporary racer has, so from initial start up to the revs that the pipe beginning to work to any degree there has to be some extra assistance to move fresh mixture from the case. That can only come from the case acting as a pump, once the engine begins to resonate effectively the case influence diminishes. The lawn mower has a very low bmep engine and within the context of two stroke race engines so is a Bantam engine. It comes as no surprise then that Bantam clutches take over the role of a low bottom gear ratio in getting revs, if not road speed, up to the point where the pipe starts to work.
The smaller the crank case volume the greater the pressure rise in the case as the piston drops, this is a help in engines with poor transfer duct shapes and low flow coefficient as it forces mixture to flow though the transfer ducts. But go too far with this and the stream velocity is so high a lot of mixture goes straight out of the exhaust port, acute pipe diffuser angles and big diameters can create so much suction which also pulls mixture out of the exhaust port. In this extreme scenario a larger case volume helps to keep that mixture in the engine, but then the signal to the carb becomes `woolly` the bigger the case volume becomes, compromises and finding a balance.
I started the w/c engine in 1993 after an enforced break of some years from all things Bantam. I didn`t really want to get involved again but my good friend of many decades, Steve Hall, persuaded me to do one last engine, I resisted for a while but eventually gave in and agreed to give it a go. As usual I got the whole design on paper with drawings, sketches and calculation sheets, Steve sourced the off the shelf hardware that we agreed to use and I made the rest.
We were ready to go in 97 (yes, it did take me that long) testing revealed a water leak on the cylinder head but a skim of just .005” for extra clearance solved that. Mark entered the Lydden meeting that June, won all the races and established the existing lap record. Had we have had a retarding ignition available then performance would have been so much better, that did not come until 2007 when Steve had enough second hand Honda bits available to enable me to graft one on, but what a difference!
During the period from 1964 when I first became involved in Bantam racing I have built the following engines, original spec 52x58, a 52x58 with a shorter rod, 54x54 with a 116 rod, various 56x50 air and water cooled engines and the final 54x54 reed valve w/c engine.
Over the passage of those years and differing engines you learn a lot within the Bantam context about what tends to work well and what doesn`t.
It was also during my `dormant` period that Steve borrowed the drawings and paperwork of one of the 56x50 engines adapted them to w/cooling, built a complete new engine jointly with Colin, another engineer friend. Steve, Mark and Colin then went on to win three championships with this new engine.

So with this back ground experience I was fairly confident as to what was not working too well for a Bantam race engine. Most crankcase hardware conversion is straight forward, common-sense engineering, and very luckily, I had extensive workshop facilities at my disposal together with the tacit cooperation of my employer.
The barrel however is an entirely different matter, but here, for me I had one great advantage, for all of those years away from the strictures within the sphere of Bantam competition and influence of its personalities I was also unimpeded by established Bantam concepts.
One thing was for certain however, it had to be water cooled!
I identified three major impediments to good gas flow, the inlet system, transfer and exhaust ducts. For me a cylinder reed valve is a total no brainer, an extra, fully functioning transfer port being fed directly from the inlet is dead easy to include along with the potential of having 260+ degrees of inlet duration, and being pressure responsive makes for a well behaved characteristics.
The transfer ducts in a standard D7/D14 barrel fall into the category of `rubbish` for efficiency in a race engine, perfectly fine for the task they were designed for but not for ours, they had to go. Gas flow objects to being contorted by sharp edges, twisting bends, and small radii, these just create energy sapping, disruptive turbulence, smooth free flowing generous curves always do best. With this in mind I prefabricated new ducts that were influenced by but were not copies of the RS 125 Honda. Fundamental to a high turbulent free flow rate is the critical profile of the duct inner profile. I turned up an alloy ring with the profile machined on, cut segments off and fixed these to the liner inner wall. This method of construction ensured that both sets of the transfer ducts were equal in all respects.
Next was the exhaust port and duct. The natural offset to one side always presents problems, particularly when using a bridged port, flow becomes asymmetric whatever you do with the bridge and asymmetric flow is also disruptive and inefficient, so that also had to go. It was replaced again by a prefabricated steel duct, but where the transfers were straight forward the exhaust was not. One primary consideration was stability of the final barrel structure, so I opted to have all of the new steel ducts that were to be affixed to the bare barrel to have 3mm wall thickness. Trying to make a new duct which tapered from the complex profile of the port window to the circular/oval outlet was quite tricky. CNC was not an option as the machines were in constant use, so cutting, milling, welding, milling again and hand grinding were used. At one point the fabrication was welded to a base plate at the roof angle to be able to clamp it whilst milling. The outcome is a duct that has dead straight symmetrical flow out and down both sides of the bridge toward the exterior pipe flange over a distance of some 55mm, only then does it angle away to avoid the frame down tube. The exhaust pipe then slips over an external adaptor that is internally profiled to blend to the header pipe diameter with no steps. Incidentally, the cooling water is pumped to the exhaust duct and can flow completely under, over and around the exterior duct wall, enabling the over-scavenged mixture in the duct to remain cool, ready to be returned to the cylinder by the timed plugging pulse.

Hope that clarifies things a bit, cheers, Trevor

Trevor, many thanks for your reply. its clarified the things im looking at . i had hoped that the modern exhaust was capable of starting the engine from dead with a little suction enticement to open the reeds via the exhaust from an external source, But as you say there needs to be some pressure usually from the pumping action.i was hoping that all the transfers could be fed from the internal side of the reed block area and the exhaust/ cylinder pressure drop action would draw the mixture through those transfers into the cylinder.Like having 5 "C" ports.Then the crankcase would be just an oil sump for the crank along with a 4 stroke type piston in the cylinder with splash lubrication. All in the hope of reducing turbulence / keeping the fresh fuel charge nice and cool before it enters the cylinder. so a pump required cylinder side of the reed block.ok(motorised big syringe with a return spring)

so, was there much difference between the 56x 50 w/c engine and the 54x54 w/c engine in performance? my 56x50 is very lively and sounds very nice.. easy set up.... when using ready made cylinder/carb/exhaust/ignition.

Nigel,
How on earth do you dream up all of these concepts you float out on us, I like this new one, don`t know If it could work but it all sounds very plausible. Are you sure you`re not on some powerful mind bending stuff for the back pain, or just plain old alcohol doing it for you?

I don`t think it is really fair of me to compare engines that were conceived ten years or so apart. What I would say is that if Mark was racing both machines against each other he would lap himself when riding the 54x54 reed engine machine, it is just so superior in every way.
Back in the day, Yamaha stuck with their 56x50.7 engines and watched the Hondas blast past them. When they finally swallowed their corporate pride and built a 54x54…250, they took the world championship, a lesson learned!
A short stroke engine can never match the port area of the equivalent square engine, a longer stroke has greater port areas again but it will always be more highly stressed and is rev limited.
One way to get a view of the respective stresses at 11,000 rpm is to compare piston speeds between the various strokes, whilst bearing in mind that friction losses increase at the square of the speed.
50.5………18.5mtrs/sec        54……….19.8mtrs/sec         58………..21.3mtrs/sec
Looking at the numbers here shows how the longer stroke is incurring significant losses compared to the short stroke alternative, you would have to push the short stroke engine up to over 12,500rpm to match the piston speed of the long stroke, but it will be so blowdown restricted that it wouldn`t work any sense up there. All things considered, the square engine is the better overall compromise. You could then bring combinations of con rod length into the equation, their relative angles and dwell time at top and bottom dead centre and rates of piston acceleration?
And so it all goes on!

Hope you are on the mend and are able to get around ok now Nigel,

Trevor.

Wonderful stiúff chaps!

.. but there has been no mention of Friction losses relative sngine speed .... i.e. Mechanical Efficiency.

   And I consider that aspect as equally important as top-end tuning.
   
Under No-Load condition an engine will rev to its maimum No-Load speed at which speed the power output equates only the  engine friction-drag power -- consequently an engine with high drag friction will never achieve the theoretical shaft power output speed no matter the exotic shape of ports, combustion system or exhaust specialities.

.Merkle in his engine performance  book reckons piston, ring and con-rod bearings can represent 60% of total engine dragging friction but, of course, he was NOT speaking of a Bantam.

What about:  "How to reduce piston and  piston-ring drag** to a minimum; getting weight of con-rod and other bits of crank machanism to a minimum  of out-of-balance forces***. What of gearbox oil viscosity, chain type and how tensioned**** crankcase seals that are still effective when hardly dragging, plain bearings instead of ball....

**The Walsh Bantam had one 60thou thick ring ... Alpha 250 used a single  Dykes with very low tangential force ...

*** improperly balanced flywheels, heavy con-rods -- including the small end bearing, of course -- out of balance increase friction-drag....  

**** I have heard of differences of prop shaft efficincy to chain being as much as 3% difference  and some chains having a lower drag friction but never heard how much for chain differences also when using the slipper style of priomary chain tensioner how much drag loss there might be  there. Even tyre tread pattern ...???

Perhaps the Bantam racers have arrived at the optimum for these things mentioned -- if so i apologise for keeping readers awake....

                  Yaaawwwnnn!.

Cheers! and back to kip....

JayBee....

yes John frictional losses, power killer, massive percentage of potential power gone in the heat of the moment.!!

yes Trevor, all pill^d up and back on my feet some people have commented that i should take the pills even when i have no pain as it makes me more sociable..

[quote="nigel breeze"]Trevor, many thanks for your reply. its clarified the things im looking at . i had hoped that the modern exhaust was capable of starting the engine from dead with a little suction  enticement to open the reeds  via the exhaust from an external source, But as you say there needs to be some pressure usually from the pumping action.i was hoping that all the transfers could be  fed from the internal side of the reed block area and the exhaust/ cylinder pressure drop action would draw the mixture through those transfers into the cylinder.Like having 5 "C" ports.Then the crankcase would be just an oil sump for the crank along with a 4 stroke type piston in the cylinder with splash lubrication. All in the hope of reducing turbulence / keeping the fresh fuel charge nice and cool  

well im down to this, need to get the sleeve to push fit and press fit for the top of the sleeve. cut transfers and configure some horizontal transferports to feed directly  from the reed cage area. is pumping required.??.. got an idea to supply this from the high pressure forming in the crankcase when cranking. https://i.servimg.com/u/f11/17/51/65/49/sleeve10.jpg