John Saunders has sent the editor the following comments on Paul’s article entitled “Modern Petrol and Classic Cars – The Manchester XPAG Tests”
With regard to the article by Paul Ireland in TTT 2 for Feb. 2017, and specifically about the graphs of Percentage Evaporation versus Temperature for current petrol compared with that from the 1930s and 1960s.
The values quoted in the diagram are very interesting, and important. I have long realised that modern fuel is more volatile than petrol of the 1950s and 60s for instance, but the magnitude of the difference in fuel volatility (or percent evaporated) is surprising and on its own is quite enough to explain the reported overheating problems that some users of XPAG-engined cars are experiencing with modern fuel.
Paul does not reveal the source of the information but the numbers, if valid, strongly suggest that current petrol will burn much richer, and thus more slowly, than fuels used when the XPAG was first designed and produced (i.e. 1937 to 1955 and beyond), if the original design settings, or close, for the ignition and carburettors are employed. A slow combustion will cause significant engine overheating as the fuel/air mixture “end gas” will still be burning when the exhaust valve opens.
According to the diagram, over a range from 15% volatility (percent evaporated) to 95%, the extra mixture richness (difference between the red and blue lines on the graph) varies from 122% at 15% volatility to 7% at 95% volatility, with a mean over the range of 53% and a midpoint of 34%. This is a very substantial offset of an engine variable that should optimally be controlled to within 2 to 3% of the mean. The carburettors are capable of this accuracy.
Note that, on ignition, the entire bulk of the petrol/air mass from “2% evaporated” to “98% evaporated” will burn virtually instantaneously (within 3 milli-secs) and all at once. The graph notations “front end components” and “back end components” are meaningless in this context.
Proposed modified settings (MG TC)
1). Static ignition point set ——– 5 to 7 degrees BTDC
2). Maximum ignition centrifugal advance ——– 32 degrees at 3,600 to 3,900 rpm
Note 1 — Changes to the centrifugal advance mechanism are not simple and involve new lighter distributor springs. This is a longer term and more permanent improvement than item 1).
3). Change carb needles to FH, or better FK to lean out the mixture from standard.
Note 2 — FH is a little rich at piston positions 2, 3, and 4, compared with FK.
4). Discard the carburettor blue springs, these only serve to increase the carb richness still further and are counter productive.
5). Discard the acceleration damper in the piston rod, or at least pour away the damping oil. It is not necessary and only causes more unwanted richness.
6). Weight the carburettor pistons to be equal and within the range 175g to 185g. The standard aluminium piston weighs around 110g. Do not go below 165g.
Note 3 — Pistons of 175g to 185g are 15% leaner than the standard piston plus blue spring combination which gives 240g maximum on each.
Original MG TC design settings
7). Static ignition set ——– TDC, or 0 degrees BTDC.
8). Maximum centrifugal advance ——– 32 degrees at 4,420 rpm (too slow for modern petrol).
9). Standard needle ES, far too rich for modern fuel.
10). Blue carburettor spring added effective weight to the piston; when piston closed an extra 73g, when full open an extra 130g. This is counter productive with modern fuel.
Note 4 — Total piston weight then is 110g + 130g = 240g, 15% rich compared with a recommended 175g to 185g on the square root relationship which is required by the physics.