Peter Lynn ARCHIVE
Gas fired beta, 159mm bore, 125mm stroke, 2.48 litre swept volume, unpressurised.
15.1 was air cooled using external aluminium finning. It had a displacer seal (like 11), annular transfer passages at the hot end and multiple grooves at the cold end.
It first ran in November ’13 with a chain driven geared-up flywheel and an epicyclic crank system that slowed the piston at points in the cycle to allow longer for heating and cooling. After development and further changes- including to a conventional (for LSMs) single crank side flywheel (15-2), maximum unloaded speed remained unsatisfactory at less than 300rpm, and start-up was very slow. Various heaters were tried- including a powerful (and noisy) double venturi burner. Heating and cooling areas at 1000sq.cm. each were less than for the successful 12.3 and 14.1 engines (both ~1200 sq.cm for cooling), but not by enough to explain the major performance shortfall. Nor was poor sealing or high friction a likely cause- sealing was excellent and friction low.
Air cooling was then replaced by a water jacket- without noticeable improvement.
Dead space became suspect- high because of the external transfer passages and substantial displacer clearance. It was also possible that until 15-3, the large transfer passages were not generating sufficient turbulence for good heat transfer (there’s a balance to be found between heat transfer and flow losses that unpressurised Stirling engines are sensitive to)- but the successful 12.3’s flow cross section is greater.
From 2017, 15 was re-built twice more with greater heating and especially cooling area (~1200sq.cm), less dead volume and various linkages for better transfer. The final 15-4 “star” linkage (2020) does appear to provide superior transfer with less of the air working fluid moving to the cooling space during the latter part of expansion.
With these changes, performance improved a lot –better than 11.3, though still a little shy of 12.3 and 14.1. Time from lighting to sustained running (a good test for Stirling engines) is now close to 1 minute, as good as for any LSM.
The remaining deficit is probably because the barbecue ring burner currently used cannot supply enough heat – speed drops off with time after starting.
Using a steel piston in a steel cylinder also limits how fast it can be run (by occasionally seizing)- but this is not a fundamental issue- fixable by fitting a cast iron piston.
15- 4 is a quiet, easy running reliable engine. To get it to run at exactly the right revs for operating a rocking chair – which is very sensitive and partly a function of the sitter’s weight (and to prevent overrevving), it uses the Watts governor originally fitted to 12.2 rigged to progressively vent the working space.
PETER LYNN, ASHBURTON, NEW ZEALAND to MAY 2025
