From the Rolls-Royce experimental archive: a quarter of a million communications from Rolls-Royce, 1906 to 1960's. Documents from the Sir Henry Royce Memorial Foundation (SHRMF).
Testing various schemes for a starting carburetter, detailing issues with fuel flow and starting times.
| Identifier | ExFiles\Box 27a\2\ Scan147 | |
| Date | 20th April 1933 | |
| -2- (2) This arrangement was tried and it was found that the starting carburetter would suck petrol up to its jet in approx. 8-10 secs., but when the engine fired it would only run for a short period - say 5 secs - before stopping. On restarting it, it would again stop, this continuing for some period. It was thought that this irregularity was due to some pulsing effect of the pump on the starter carb. flow, or due to air trapped in the pump being sucked up with petrol - in all probability a combination of both, resulting in the starter carb. being starved of fuel at one instant and choked with air the next. As a result of this test it was decided to try the third scheme. (3) Pipe from starter carb. was now taken into main pipe line from tank to pump at the point where inlet union screws into pump, and this arrangement was found to be perfectly satisfactory - after running dry, the time necessary to run on starter motor in order to start on starting carb. was 7 - 9 secs. while a further 20 secs. running on the starter carb. gave sufficient petrol level in the floatchamber to open up on the throttle. No irregularity in running on the starter carburetter was experienced with this arrangement. As a final test of this scheme, the engine was run on starter motor (with ignition switched off) and the floatchamber level was observed by means of a gauge glass - in 1 mins. running the floatchamber level had risen to within 1/4" of its normal height - this is a considerably higher level than would be needed to start off throttle-edge slow-running scheme, so that even if the starter carb. could not be used for starting owing to the engine being hot, it should be possible to start off slow-running device in 40-45 secs, using starter motor only. Notes:- The effect of lowering the tank on the last scheme was tried - it was found that while a drop of about 6" did not appreciably affect it, a drop of about 15" caused the running on the starter carb. to be very erratic, with frequent stops - indicating that the starter carb. suction was not much in excess of that required to lift petrol from normal tank level. Hence it was thought that the extra resistance due to having to suck through both pump valves, mentioned in scheme (1) would alone render that scheme impracticable. | ||