plug gap / ignition timing/ carboration

Courtesy Of Trevor Amos.

Carburettor bell mouths don`t seem to get a lot of press, trying to compete with the more popular (dominant) tuning topics seems to leave it out in the cold. If truth be told it actually does matter quite a lot, when the journey of the ingested air starts badly it never makes up for lost ground. So to start off correctly seems a sensible situation to be in. The mean reference point outside of the bell for reflecting pulse waves is normally taken as .5xD, where D is the carb nominal bore diameter.

The first illustration is of a pair of beautifully crafted FPE cnc produced carbs to suit the parallel twin replica of the Rotax style kart engine using Aprilia barrels. These are copies of the Dellorto specials fitted to the 250 and 125 Aprilias. Of interest is the position of the progressive stepper motor, electric power jet and the imposing PTFE bell mouth. The bore size is 42mm and the huge profiled bell mouth illustrates the full area from which a carb draws air from. If you should fancy one, it is available for a mere 2,000 euros, plus vat!!!!



The CFD simulation images below illustrate particle velocity flow into a typical bell profile with the elliptical shape proving to be superior to all others. The multiple pictures are fairly self-explanatory, and the spit back potential of three differing shapes does describe a feature of design that is practically ignored by most tuners.

Taken from information published in the original technical paper, the measured flow rate from the worst, at 30.023 grms/sec, to the best at 36.15 grms/sec, at first may not seem great. Measure that against the number of seconds of a typical race duration and the potential benefit becomes very significant. Apologies for the indifferent reproduction here but the originals were tiny and indistinct, any larger and definition begins to deteriorate dramatically.



The elliptical shape of the lower bell shows both the most efficient inflow and minimum subsequent reverse-flow spit back, a real win, win! The middle pair with no bell is a situation no one should encounter, but it forms the basis from which the remainder can be assessed.

Violent reverse flow spit back can clearly be seen in both top and middle simulations! The optimum profile shown represents an exponential curve and that does a very good job of damping out unpleasant shock waves. Furthermore, the danger of a Vena-contracta forming at the entrance to the carb is all but eliminated. In very extreme cases, boundary layers and Vena-contracta restrictions can reduce flow potential considerably.

The other area in the Bantam gas flow regime where constrictions occur is the rear cone to tail pipe junction in the exhaust system, in all cases the situation worsens with a rise in rpm driven gas speed. In the last two pictures there is just a ghost image of the zone of draw area that the inflow reaches out to, a good enough reason to avoid frame tubes and the suchlike. Providing a large capacity Plenum containing cool, turbulent free air for the carb to breathe from could very well lift power more than just a touch, cool is always good!



Taking in a wider context to look at entry contours one need only examine as examples of good practice, the efforts of the design engineers of both jet aircraft engine nacelles and any duct entry of a F1 car body. In both of these examples huge amounts of corporate energy goes into finding optimum entry profile with minimum turbulence and drag.

So it should follow that Bantam inlets deserved to be afforded similar consideration, drag is not really a problem but the basic aerodynamics and velocity profiles need to be optimised and improved performance will follow as a natural consequence. The bell mouth drawing above provides just about every detail necessary for manufacture of a state of the art carb entry bell. If the overall size is too large to accommodate then a slightly scaled down version will probably be ok.

Under the pair of images  is a computer generated image of the FPE carb, its internal components, and the location of the power jet motor. The floats shown here rise and fall on fixed vertical posts and not by the conventional pivot method.

Bravo Trevor! From reading your contributions on here I´ve got to know now -- that I missed a lot. That´s because I left  the tuning to my mentor and the engine-man of our little (3-man) team in early days of Bantam racing . BUT I didn´t miss out on the bell-mouth aspect and I personally made a couple to my old Hydraulics Lecturer´s calculations for my earlier race machines. Those were wonderful times in terms of machining equipment avaialble. That was  in Walthamstow Tech´s Engineering Department where I was also being paid whilst doing my `homework´... I missed all that later...

Must be very costly for Bantam Racers to get machining and welding jobs done... Anyone willing to publish a typical machining or welding cost?  

Cheers!

John,
You may regret that you missed a lot, but we all did, in those days there was precious little in the way of reference material available. Most of the stuff relevant to Bantams was more or less a re-hash of stuff already out there, so it was more a case of suck it and see, they would call that empiricism these days!

Like you I had access to unlimited engineering facilities and as most of my engineering life then, was in an Aerospace environment, quality and accuracy came naturally. An unfair advantage I acknowledge, but when the formula became a free for all one would be foolish not to participate?

Wonderful times for sure, but when I discovered that stripping a Bantam engine came a poor second to stripping with my statuesque muse of the time, dedication certainly began to waver!

We do have today a fantastic source of fundamental research information available to us that we could not ever dream of back then, and it is just the interpretation of this that I try to do to assist Bantam enthusiasts to improve engine performance.

Very many thanks to you Ed, once again, for enabling the posting of this stuff!

Trevor