• Version
• 246 Downloads
• 1.48 MB File Size
• 1 File Count
• August 31, 2016 Create Date
• August 31, 2016 Last Updated

National Advisory Committee for Aeronautics, Report - An Experimental Investigation of Rectangular Exhaust Gas Ejectors Applicable for Engine Cooling

An experimental investigation of rectangular exhaust-gas
ejector pumps was conducted to preside data that would serce
as a guide to the design of ejector applications for aircraft
engines with marginal cooling. The pumping characteristics
of rectangular ejectors- actuated by the exhaust of a single—
cyl-inder aircraft engine were determined for a range of ejector
mixing-section area from 20 to 50 square inches, orer-all length
from 12 to 42 inches, aspect ratio from I to 5, diflusing exit
area from 20 to 81 square inches, and exhaust-nozzle aspect
ratio from I to 42. A few tests were conducted with a multi—
stage ejector, a divided ejector, and an ejector incorporating
bends along its length.

With a decrease in the quantity of air pumped and an in-
crease i-n the length of ejector, the ejector pressure rise increases
to optimum values. Optimum ralues of ejector area were
found to depend upon mass-flow ratio of air to exhaust gas for
giren engine operating conditions. Difiuser-ezit sections con—
siderably improved the performance at the ejectors. An
arrangement of a straight mixing section with a difiusing exit
and a flattened exhaust nozzle provided the most farm-able
ejector performance. An ejector composed of a straight mixing
section of 24-inch length and 25-sguare-inch area with a difius-
ing exit of 12-inch length and 1.87 exit-area—entrance-area ratio
presided a pressure rise of 6‘ inches of water for a mass flow
of air re prese-ntatire of cooling requirements (sis: times the mass
flow of erhaust gas) for the engine when operated at a cruise
power of 85 indicated horsepower.

A simplified analysis, which considers the effect of pertinent
ejector rariables and indicates the perfomance in terms of
known engine quantities, was made. The agreement between
theory and experiment was fair orer the range of ejector con—-
figurations tested, except that a serious discrepancy existed in
that the optimum ejector areas prescribed by theory were smaller
and the ralues of pealc pressure rise predicted at the small
optimum areas were higher than indicated hy the tests.