The first Douglas aero engine starter?
While flipping through a box of WW1 aero documents at the National Archives in Kew, I caught a glimpse purely by chance of a familiar looking engine. Sure enough it was the distinctive 2-3/4hp engine; or it had been at one time. The document was a report by the R.A.E. (Royal Aircraft Establishment) on development of a suitable engine starter for installation in aircraft, dated August 1920. The experimental engine was installed in a Bristol Fighter powered by a Puma 6-cylinder inline engine. It seems the Douglas engine was used as a matter of expedience rather than as development of a new application. Later tests in the same report involved a specially made 2-stroke engine.
The rear cylinder has been replaced by a fabrication that is used as a compressor cylinder. Rather than just pumping air, it compressed and pumped an air/fuel mixture fed by the same carburetor (and original inlet manifold) that supplied the other, unmodified cylinder. It ran as a single cylinder engine solely on the front cylinder. The compressed mixture was lead via a relatively small pipe to a rotary distributor driven by the aero engine camshaft, with a pipe leading to each of the Puma’s six cylinders via a non-return valve. The pressure mixture was applied to the cylinder on the firing stroke, causing a pressure unbalance and the engine to turn at a rate of 3-4 revolutions per minute. A bleed hole (about 1/16 of an inch) supplied surplus compressed mixture to the aero engine induction pipe so that the next cylinder on the intake stroke received a charge of explosive mixture. Further, the magneto lead from the front cylinder of the Dougie engine was taken to the center distributor terminal of the aero engine magneto, via a suitable high tension (H.T.) voltage switch that can short-out or ground the circuit when closed.
The procedure for starting up as described:
“The pilot turns on the petrol to the main engine, opens the throttle slightly and switched on. The observer starts up the pumping set (with H.T. switch closed) which immediately begins to turn the main engine around. He then opens up the short-circuiting switch, thus supplying spark to the main engine. The engine usually starts immediately on opening the switch, but if it fails to do so it will invariably fire after turning over on to another cylinder. In fact, the main engine will turn slowly, firing all the time, even when it is switched off and the throttle quite closed, both the fuel and air mixture and the spark for ignition being supplied by the pumping set.”
A further test experimented to see if the small diameter tube carrying the mixture to the aero engine was an impediment. A length of ˝ diameter copper tube thirty feet long was imposed. Left overnight about 2 ounces of fuel condensed out, but the apparatuses started the engine normally. Also exhaust valve timing of the main engine was a factor. If it opened too early not as much turning torque was developed to overcome compression in the other cylinder(s). This was just barely a factor on the six cylinder, and not expected to be a factor when applied to twelve cylinder engines. Nor was it a factor with test while airborne, as the wind milling effect of the propeller assisted the starter engine.
The weight of the Douglas starter apparatus was not given, but the slightly more powerful 2-stroke engine that replaced it was specified as adding 31 pounds. The induction pipe was arranged so the pumping cylinder got a somewhat richer mixture. Since no more is heard of these remote starting engines, it is presumed that Coffman type starters and electric starters quickly superseded them; pony motors (with mechanical drive) only surviving on the largest engines of WW2. For that Douglas manufactured some prototype pony starter engines for Armstrong Siddeley and Napier in water and air-cooled versions.
The first photo shows the modified 2-3/4hp engine mounted to the outside of the observer’s cockpit of a Bristol Fighter.
Same view, but with the streamline cover installed. The tube supplying the pressurized fuel/air mixture, as well as the high tension lead can be seen heading forward to the main engine.
The view from inside the observer’s cockpit. The gauge is calibrated in pounds per square inch and registers the pressure in the pipe to the main engine. Presumably there was a sprag clutch built into the sprocket on the flywheel hub.
References: Air Publication 791; R.A.E. report No. E.2381; Summary, Results of Tests of the R.A.E. engine Starter. National Archive asset No.: AIR 10/809
D. Kephart, Glen Mills, PA, USA
First published in the NCR, Sep/Oct 2015
