Re: Starting large piston engines

Date:         17 May 97 15:15:51 
From:         Steve Lacker <>
Organization: applied research laboratories
References:   1 2
Followups:    1
Next article
View raw article
  or MIME structure

Carl Peters, M.D. wrote:
> How were large radial pistons (eg. on a Connie, DC-4,6) started? The
> starter would be electrical, but what was the source - the onboard
> batteries vs. an external generator cart?

My only firsthand observations involve ex-WWII era military planes, but
many use the same engines or predecessors of the engines used in the big
passenger liners. All that I've observed are started with internal
batteries, including the Douglas A-26 (P&W R-2800, same basic engine as
the DC-6) and B-29 (Wright R-3350, same basic engine as the DC7 and
Connie). Once one engine is running, its generator can begin recharging
the batteries and supplying power for the next engine start.  I have
noticed that if an engine fails to start readily, another engine is
often tried next and then the stubborn engine started once another
engine or two is running. A long crank before getting the first engine
running is very often followed by a longer interval between getting the
first one running and starting the next one than is normally seen,
presumably to allow for some battery charging.

A funny thing about the demands on the starter of a piston engine-
having more cylinders can actually reduce the peak load on the starter
because the compression (energy required from the starter) and
downstroke (a portion of the energy delivered back to the starter)
cycles overlap. A small engine (e.g. an automotive 4-stroke 4-cylinder
with zero overlap) actually puts a higher *peak* torque demand on a
starter than an 8-cylinder with the same size cylinders. The *average*
demand is a bit higher with more cylinders though, and since aircraft
radials have MANY cylinders and each is very large, the average demand
is still large. My point is that its not as large as you might think
since overlap from 14, 18, or more cylinders smooths out  the peaks, and
also aircraft radials typically have very low static compression ratios
on the order of 7:1 with turbo- or super-charging or a combination of
the two. (Torque demand from the starter goes up with compression ratio,
which is why Diesel engines need huge starters compared to
spark-ignition engines). Normally aspirated automobiles, in contrast,
average more than 8:1 compression, and 10:1 is fairly common with
between 4 and 12 cylinders. So you can't just look at the starter from a
400 cubic inch auto V8 and scale it up linearly to estimate the starter
needed for a 2800 cubic inch 18-cylinder radial. The radial does not
need a starter 7 times as large or drawing 7 times the power that the
automotive starter draws, but perhaps (just a wild guess) only 2-3 times
as big.

Stephen Lacker
Applied Research Laboratories, The University of Texas at Austin
PO Box 8029, Austin TX 78713-8029