Академический Документы
Профессиональный Документы
Культура Документы
The visualization of flow patterns has played a flow. The smoke was faint and difficult to make
singularly important role in advancing our under- out, but studying it was a beginning.
standing of the mechanics of fluids. Visualization In France, about 1899, E.J. Marey, who was fa-
hasTed to the discovery of flow phenomena, has mous for his photographic studies of animal
helped in the development of mathematical models locomotion, turned his attention to photographing
for cornplex flow problems and in the verification air in motion. Marey, cognizant of the work of
of principles, and has been an important tool in the Mach, used a vertical wind tunnel with a 200 x 300-
development of complicated engineering systems. mm crosS section. The front and sides of the test
Visual observation of flow phenomena was the first section were glass and the. back was covered with
and, for a long time, the only experimental tech- black velvet. Air was drawn into the tunnel by a
nique available. Dust or smoke particles in air and small suction fan after passing through fine silk
dirt or other debris in water provide the necessary gauze at the inlet. F-l shows a sketch of this wind
contamination for flow visualization. Interest in tunnel. Smoke from burning wood shavings entered
flight led experimenters to visualize air flows in or- the wind tunnel upstream of the gauze straighteners
der to understand the mechanics of objects moving through a row of fine tubes. Marey made excellent
through the air. smoke-flow photos using a magnesium flash.
Flow visualization in wind tunnels began with the The interest in smoke visualization seems to have
work of Ludwig Mach of Vienna in 1893. Ma.ch's waned somewhat in the first quarter of the century.
indraft wind tunnel had a cross-section of 180 x Smoke experiments began in 1911 at the National
250 mm and was driven by a centrifugal fan that Physical Laboratory in England. Around 1921 at
could produce a speed of 10 m/sec. Wire mesh over the U.S. Naval Gun FactoryA.F. Zahm may have
the inlet straightened the flow. One side of the test used smoke to analyze the air flow over naval
section was glass and the others were black on the vessels. Ludwig Prandtl and his a.ssociates used
inside. The flow was observed a.nd photographed smoke in their Gottingen wind tunnelin 1923 to
using silk threads, cigarette smoke, and glowing study flow over airfoils and bluff bodies.
iron particles. Mach obtained only smoke-flow Interest in this technique developed rnore rapidly
photos for the flow past a plate perpendicular to the after 1925. Around 1930, smoke visualization was
used in England by L.F.G.Simrncms, N.S. Dewey,
and T. Tanner to study the flow about circular disks
THOMAS J. MUELLER (AF), a profes-
and symmetrical airfoilS. The srnoke was produced
sorof aerospace and mechanical engi- by titanium tetrachloride. In 1933, K.W. Clark used
neering, joined the faculty at the Univ. stannic chloride smoke in conjunction with other
of Notre Dame in 1965. SinCe then he
has been engaged in teaching and re- techniques to study the flow over several wings.
search in the gas dynamics of sepa- In 1932, for lectures at Cambridge Oniv., W.S.
rated flows, propulsion aerodynamics,
computational fluid mechanics, bid- Farren used a modified version of Simmons' smoke
fluid mechanics, and flow visualiza- tunnel at the National Physical Laboratory. The
tion. Before coming to Notre Dame he
was on the faculty of the Univ. of Illi- NACA built its fitst smoke tunnel about 1933 and a
nois at Champaign-Urbana, where he second one in 1938. These tunnels seem to have
received a Ph.D. in 1961, and he was a
senior research scientist at the United been used mostly for demonstrations. The tunnel
Aircraft Research Laboratories from built by R.W. Griswold, II, at Old Lyme, Conn.,
1963 to 1965. In 1973-74 he was a visiting professor at the Von Kar-
man Institute for Fluid Dynamics in Belgium. He was an AGARD about 1940, used smoldering rotten wood to gener-
(NATO) lecturer in fluid mechanics in 1974 and 1976. Mueller reo ate smoke.
ceived the 1980 AIAAASEE Educational Achievement Award and
was an AIAA Distinguished Lecturer for 1980-81. Although the interest in the period frOID 1925 to
50 \111\\\11\1\\1\\U\1\\111\\111\11\\\1I\il\\1\\111\
TRP00668
the early 1940s was in using smoke visualization to
help solve practical aerodynamic problems, the
smoke photographs were not as good as Marey's.
Many of the visual results were useful, but few if
any significant quantitative results were obtained.
All these tunnels were two-dimensional and various
types of "smoke" were used, the most popular be-
ing obtained by burning rotten wood or usingtita-
nium tetracholoride.
The. best of these. early. two-dimensional tunnels
was developed by A.M. Lippisch in Germany. He
obtained a large number. of good smoke photos of
the flow around plates, cylinders, and airfoils, in-
cluding the Lippischrotor wing. He also began de- operational in 1937, was developed primarily for
veloping an intermittent smoke-delivery system. class-room demonstrations. It had a 38:1 con-
Lippisch's evaluation of the potential of the smoke traction section and produced speeds of up to 3
tunnel is clear when he writes: "The value of the m/sec. Brown made photos using titanium tetra-
smoke tunnel consists chiefly in demonstrating the chloride for smoke. In 1940 he began operating a
effects of a given body on the flow and to show the three-dimensional indraft smoke tunnel. This tun-
true course of flow in the case of special devices. nel had a single screen followed bya 9: 1 contraction
Without carrying out tedious force measurements in area. Its 610 X 61O-mm test section was about
we are, therefore, in a position to develop aerody- 914 mm long, and speeds of up to 12.2 m/sec could
namically good designs or to observe a special flow be attained using a 1 hp d.c. motor. Smoke was pro-
phenomenon which enable us, for instance, to de- duced by coking wheat straw and was introduced
termine the true causes of stalling effects." upstream of the anti-turbulence screen through a
There was, evidently, no serious consideration row of tubes (a rake). Brown made another new
given then to obtaining quantitative data from such three-dimensional smoke tunnel but made little pro-
experiments. gress with it during World War II. (In 1943, the first
About this same time, F.N.M. Brown at the really successful oil-smoke generator was developed
Univ. of Notre Dame began his research in flow by Preston and Sweeting in England.)
visualization. His first smoke tunnel, which became In 1947 work began at the Univ. of Notre Dame
January 1983 51
on a research three-dimensional smoke tunnel. As a
result of the lessons learned earlier, this smoke
tunnel had five bronze screens at the inlet of the
AIRIN 12:1 Smith-and-Wang contraction section and the
same size test section. With it useful speeds of 10.7
m/sec could be attained using a squirrel-cage fan
driven by a 5-hp motor. Coked wheat straw pro-
duced the smoke. A total of 12,000 watts of steady
lighting was used to photograph the flow.
BLACK VELVET This three-dimensional smoke tunnel evolved
slowly into the one used today. F-2 shows the 1958
version of this tunnel, and F-3, the latest version of
Brown's smoke tunnel. In conjunction with his
smoke-tunnel development, Brown developed a
movable smoke rake and the first easy-to-use
AIRTO keros~ne smoke generator that could produce large
FAN
l1~r;'IoH::> .
"q!Je~tloH:s of smoke. he was also the fIrst to take 3-D
F1 The low-speed smoke tunnel of E.J. Marey in 1899. and stereo photographs of smoke flows.
F2 Brown's threedi
mensional smoke tun SMOKE PIPE
nel in 1958.
- LAYERS OF SCREENING
PREVENT TURBULENCE
SMOKE GENERATOR
2169
~
JDARKROOM
smoke-flow experiments. The most notable smoke
. ~..-.--J
MODEL DRIVE
January 1983 53
'\ SMOKE STREAM INJECTOR questions that the force and pressure distribution
\ YAWING MECHANISM
investigations had been unable to."
\\ \" DOWNSTREAM
\ DIFFUSER
Useful results were obtained in this tunnel and in
the larger 51 x 914-mm tunnel built by Lippisch for
\ \ Princeton. The two major improvements in the
latter tunnel over the pilot one were its size and
speed range. It also had a large contraction ratio of
19.5:1. Test velocities usually ranged between 9 and
12 m/sec, and, with a great deal of adjustment to
the pitch and yaw of the injector tubes, smokelines
F5 Threedimensional lowspeed smoke tunnel at Princeton could be maintained to a speed of 20 m/sec. Com-
Univ. was later scaled up to almost twice the size. bined with the use of 406-to-61O-mm-chord models,
F6 e su ersonic
smoke tunnel eveloped
by V.P. Goddard at
Notre Dame Univ. With
it he made photos like
ALL DIMENSIONS IN MM.
~ TEST SECTION
M o =1.38
t
1 290
SUPPORT
CONCRETE FLOOR
. ~... . . ~. ' .... "~ .. - ' .. ',.,.,'.', "";' . '" ....... : ,'. ' ....
chanical vibration. According to Hazen and this permitted test Reynolds numbers of up to
Lehnert: "This tunnel is notable for the fact that, 830,000.
although it contains a large number of incorrect de- These two-dimensional smoke tunnel studies led
sign features, it proved the value of such flow to the design and construction of a 305 x 406-mm
visualization techniques and answered many of the three-dimensional smoke tunnel at Princeton.
(Continueclon page 62)
54 Astronautics & Aeronautics
Smoke Visualization
(Continued from page 54)
CAREER OPPORTUNITIES. Shown in F-5, it used a %-hp d.c. motor to deliver
rtbw speeds from 0 to 9 m/sec, the Preston-and-
Sweeting smoke generator, and a number of differ-
BeechAireraft Is Hiring ent types of smoke delivery. Smoke was supplied to
Ai.f.
E
the test section in single or multiple heavy streams,
similar to the method used by Brown, by injecting
from tubes in the honeycomb or in much finer
streams from the injector rake in the first con-
traction section. Smoke was bled directlY"Through
January 1983 61