ENGINEERING RESEARCH INSTITUTE UNIVERSITY. OF MICHIGAN ANN ARBOR Final Report OPTICAL TECHNIQUES FOR STUDY OF BLOWERS J. LOUIS YORK Associate Professor of Chemical alfd Metallurgical Engineering ALEXANDER WEIR., JR. Associate Research Engineer,'Engineering Research Institute SAMUEL D. PHILIPPS Owens-Corning Fiberglas Corporation, Newark, Ohio Project 2287 OWENS-CORNING FIBERGLAS CORPORATION NEWARK, OHIO December, 1954

- NGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN ABSTRACT The possibility of applying aerodynamic optical techniques to the study of blower and spout operation became of interest in early 1954. The facilities and experienced personnel at the Aircraft Propulsion Laboratory of the University of Michigan made it possible to run preliminary tests there, and this report covers such tests. The primary purpose of this series of tests was the gathering of information and experience on the applicability of optical techniques to blower investigationso A secondary purpose was possible comparison of gas flows through blowers of different design. 1 Optical techniques, both shadow and schlieren spark photographyo are informative and could be designed for study of blowers in actual service on wood-forming, either on research or factory machines. Systematic studies with optical techniques would assist in. understandinrg gas flows,. turbulence, and energy exchange during the wool-forming process. Increasing primary air flow through the blowers increases induced air flow sufficiently to deflect the primary air jets and preverntc direct impingement. Some other observations are given in the detailed comments on the negatives, although many of them cannot be explained with present knowledge of the flows. ii

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN OPTICAL TECHNIQUES FOR STUDY OF BLOWERS EXPERIMENTAL WORK Shadowgraph, schlieren and interferometric techniques are commonly used for examing density gradients in gas flow. The simplest system optically is the shadow system, in which light from the source passes through the region to be studied and falls on a photographic plate. A density gradient in the test region deflects the light rays passing through the test region, the amount of deflection being directly proportional to the magnitude of the density gradient. If a uniform density gradient exists throughout the region studied, there will be uniform illumination of the photographic plate. If the density gradient is not uniform, however, there will be a change in illumination on the photographic plate, this change being proportional to the derivative of the amount of deflection, or proportional to the second derivative of the density. In the schlieren technique, the change in illumination of a photographic plate is directly proportional to the density gradient,- or to the first derivative of the density in a direction perpendicular to the schlieren knife edge. A diagram of the "balanced" mirror system employed in these tests is shown in Fig. 1. Light from the source is reflected from the first parabolic mirrQr so that parallel light passes through the region to be studied, is focused by a second parabolic mirror> and illuminates a photographic plate. A knife edge, located at the focal point of the second parabolic mirror, ine tercepts those light rays which are deviated by density changes in the region to be studied. This interception is responsible for the change in illumination at the photographic plate. The sensitivity of the schlieren system is inversely proportional to the size of the light source in the direction perpendicular to the knife edge, directly proportional to the amount of light cut off at the knife edge, and directly proportional to the focal length of the second parabolic mirror. An interferometer requires a more complex and delicate optical sys tem, but provides directly a measure of density. The initial parallel light beam is split into two beams, one of which passes through the region under J - 1

ENGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN study, the other serving as a reference. The two beams are then recombined and interference fringes show the density in the test region. A knowledge of gas compositions and pressure distribution permits determination of temperature distribution. The University of Michigan interferometer is of the MachZehnder type with 8*inch mirrors. Such an instrument appears at the moment to be impractical for blower studies and no tests were made with it. Blowers were installed in the schlieren system (Fig. l) at the location marked "nozzle." Photographs were made while light passed through the space between the opposing rows of slots and parallel to the rows, and while light passed perpendicular to the rows and to the central plane of gas flow. With the light parallel to the slot rows/ the interaction of the primary jets could be seen, but the primary jets themselves could not be distinguished. Light perpendicular to the slot rows clearly showed each pair of opposing primary jets, but the body of the blower blocked off the zone of interaction. All schlieren photographs after negative 3 were made with the knife edge horizontal at about 50 percent cutoff. Shadowgraphs were made by removing the knife edge from the schlieren system. All photographs are by the'open-flashshut" method. The room is darkened, the film uncovered, the one-microsecond light flashed, and the film covered. Most of the work was performed on four factory-type blowers; the 119- 94 — 59-, and 49X-slot. These were photographed in both positions at three primary air pressures: 50, 100, and 200 psi. Both shadow and schlieren photographs were made in most of these combinations. A 59-slot blower with an 8-degree skirt and a 5-slot experimental blower were briefly tested. Two 94-slot blowers with 3/32-inch radius curvature on the top and with 5/32-inch radius were compared. RESULTS Ninety photographs were made on July 20 and 21, 1954. Ten failed to produce good negatives and one was spoiled in developing. In most cases duplicates were made at each set of conditions. All good negatives are appended to copy number 1 of this report. The complete data sheet and detailed comments on all negatives are in the Appendix. DISCUSSION Figures 2 and 3 (negatives 1 and 2, respectively) show the effect of orientation of the knife edge in schlieren photographs. The density 2

ENGINEERING RESEARCH INSTITUTE E UNIVERSITY OF MICHIGAN - A BB 4% -4 -.4,ny rrM 16 li cm yx I LU ur x B n u 9'-.4g.4-.I C86 ~~Bi: ~~~~~~~~~~~~~.. ~~~~~Be ~ ~ ~ ~ ~.. ~~gie ~ 44 4-.. 4-.'..4-.. 1 —. -- F I At 6 — DEF~, SPARK-G- (AP LIGHT SO0URCE FIRST PA~dRABOLIC MlDRRORR NOZZLEb SEZCOND PARABOLIC MIRRORt KNIFE EDGE CAMERAA Fig. I. Parabolic Mirror Schlieren System *&: Fig. 2. 119-Slot Blower, 100 ps1I, Knif: Edoge Vertical (Negative 1) 5 ~~~~~~~~ Y ~ ~ ~ ~ ~ ~ - -----

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - gradients appear only perpendicular to the knife edge -- vertical in negative 1 and horizontal in negative 2. The background atmosphere is relatively uniform except for dust flecks and the concentric rings which are slight pressure fluctuations corresponding to sound waves originating at the blower discharge. Negative 1 shows the angle of expansion of the primary jets and the apparently complete mixing of opposing primary jets within a short distance below the blower. Negative 2 shows the same angle of expansion, and more clearly shows the primary jets before they appear to lose identity. A sharply defined jet boundary persists almost to the bottom of the blower, although pressure gradients have begun to spread to the surrounding air. Both pictures are of a 119-slot blower at a primary air pressure of 100 psi. Figure 4 (negative 7) shows the same blower at a primary air pressure of 200 psi. The doubled air flow is inducing more air from the top of the blower down through the space between the slot rows, but the velocity through the slots is not changing. The result is greater deflection of the primary air jets - so much that they do not impinge directly but become essentially parallel before disintegrating. The overall angle of expansion remains the same, however. Figure 5 (negative 11) is a shadowgraph at the same conditions as Figs. 2 and 3. Much more detail is visible in the primary air jets, but the expansion portion is difficult to see. Figures 6 and 7 (negatives 27 and 28) are schlieren photographs of a 49X-slot blower at 100 and 200 psi. Turbulence is greater than for the 119-slot blower at the same pressures, and the primary air jets disintegrate more quickly. At 200 psi the initial jets from the 49X appear broken and irregular, which is probably the result of delivery of air by the center slots down along the skirt (similar to the deflection at 200 psi in the 119slot blower) and by the end slots toward the center. The overall angle of expansion is approximately the same for both blowers. Figure 8 (negative 39) is a shadowgraph of a 59-slot blower at 100 psi. Comparison with Fig. 5 (negative 11) shows sharp impingement and very little deflection of the primary jets in the 59. Apparently, less air is induced through the top of the blower. Figures 9 and 10 (negatives 51 and 52) are schlieren photographs of a 94-slot blower at 100 psi and 200 psi respectively. Comparison with Figs. 3 and 4 (negatives 2 and 7) shows much the same flow relation. The ll9-slot blower gives less of-an angle of impingement, but at 200 psi neither blower shows significant direct impingement of the jets. Figure 11 (negative 57) is a schlieren photograph perpendicular to the slot rows of a 94-slot blower at 100 psi. The identity of the individual jets is sharply defined for some distance below the bottom of the blower. The angle of expansion at the end is about 6 degrees.,______,.____,.__.-.. )1 Ill I, I, 1_____1:___i_' -r

I ENGINEERING RESEARCH INSTITUTE UNIIVERSITY OF MICHIGHAN -- F ig 3. ]3.9-Slot Blower, 100 psi.,) Knife Edge Hor~izontal (Negative 2) Fig. I. 1.9Slot B9lowerl, 200 psi, Sehileren (Negative 7) 5

ENGINEERING RESEARCH INSTITUTE E UNIVERSITY OF MICHIGAN Fig. 54. 119-Slot Blower, 100 psi, Shadow (Negative 11) I v1 Fig. 6. 4t9X$Slot Blower,.i100 psi, Sch.ierern (Negative 27) 6

J.- ENGINEERING RESEARCH INSTITUTE 8 UNIVERSITY OF MICHIGAN - ig. -7. )igX-iot Blower, 200 psi, Schlieren (Negative 28) JJ Fig. 8. 5j9-1ot Blower, 100 psi, Shadow (Negative 39) f7

i L, ENGINEERING RESEARCH INSTITUTE 0'U"NIVERSITY OF MICH~IGAN I Fi1g. 9. 94.Sloot Blowrr, 100 psi", Sch'ler n (Negative 51) Figl..0. 94-Slot Blower, 200 psi3 Scheltieren (Negative 52) 8

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - Figure 12 (negative 67) is a schlieren photograph perpendicular to the slot rows of a 49X-slot blower at 100 psi. This blower is designed to deliver the air jets at an angle in this view, but the angle is not apparent although the jets can be identified. Except at the ends of the slot rows, the jets are moving straight down. All blowers show about the same overall angle of expansion from observation along the slot rowss but views perpendicular to the slot rows show variations from contraction to about 10-degree expansion. The variation is not consistent enough for explanation from these photographs. All views along the slot rows show a definite change in flow characteristics of the induced air at the change in cross section where the slots discharge. Apparently, a sudden acceleration of the induced air and a corresponding decrease in pressure occur. In examining all the photographs made by these techniques, the lack of focusing must be kept in mind. Everything in the parallel portion of the light beam is in focus and no depth is apparent. ACKNOWLEDGEMENT The assistance of Robert E. Cullen and James A chollsA in. ad.apt ing the schlieren system for this study is appreciatedo.1 9

: __ ____~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:H 0 0 0.~.,~ O t~ ~: O~: H3:v e-' XI f s wO i a: t-,wC i tr f?,:;:* 0 04 U. T —4. z -,,( m rr" z m m Z: m "r

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN APPENDIX DATA SHEET (see page 13 for explanation of terms) Air Negative Blower Position Photo. Remarks Pressure 1 119 Parallel 100 Schl Knife-edge vertical, 50% cut-off 2 (4847- Knife-edge horizontal, 50% cut4046) Parallel 100 Schl off 3 (4847- Knife-edge horizontal, 75% cut4046) Parallel 100 Schl off 4 (48474046) Parallel 100 Schl Plate added on top of blower to 5 (4847- simulate bushing. No apparent 4046) Parallel 100 Schl effect, but plate left on. 6 (48474046) Parallel 100 Schl 7 (48474046) Parallel 200 Schl Higher pressure shows jets longer 8 (4847- but deflected more. Same angle 4046) Parallel 200 Schl of expansion. 9 (48474046) Parallel 200 Schl 10 (48474046) Parallel 100 Shad 11 (4847- Better definition of jets, less 4046) Parallel 100, Shad definition of expansion zone. 12 (48474046) Parallel 200 Shad 13 (48474046) Parallel 200 Shad 14 (48474046) Parallel 50 Shad 15 (4847- 50 psi gives lower turbulence, 4046) Parallel 50 Shad more direct impingement of jets 16 (4847- (15 spoiled in developing). 4046) Parallel 50 Schl 17 (48474046) Parallel 50 Schl;....,,,......'...',,,.'. 11

7- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - Air Negative Blower Position e re Photo. Rfemarks Pressure 18 49X Parallel 50 Shad19 (7915) Parallel 50 Shad Much turbulence. Primary jet 20 (7915) Parallel 100 Shad appears irregular at 200 psi. 21 (7915) Parallel 100 Shad Jets disintegrate sooner than 22 (7915) Parallel 200 Shad in 119, mixing is complete 23 (7915) Parallel 200 Shad sooner. 24 (7915) Parallel 0 Shad 25 (7915) Parallel 50 Schl 26 (7915) Parallel 50 Schl 27 (7915) Parallel 100 Schl Angle of expansion about 20~. 28 (7915) Parallel 200 Schl 29 (7915) Parallel 200 Schl 30 59 Parallel 50 Schl 31 (N8737) Parallel 50 Schl 52 (N8737) Parallel 100 Schl 33 (N8737) Parallel 100 Schl Underexposed. 34 (N8737) Parallel 200 Schl 35 (N8737) Parallel 200 Schl 36 (N8737) Parallel 50 Shad 37 (N8737) Parallel 50 Shad Sharp impingement of jets, much 38 (N8737) Parallel 100 Shad turbulence at 200 psi. Angle 9 (N8737) Parallel 100 Shad of expansion about 240. 40 (N8737) Parallel 200 Shad 41 (N8737) Parallel 200 Shad 42 94 Parallel 50 Shad 43 (N6299) Parallel 50 Shad Sharp impingement of jets, ex44 (N6299) Parallel 100 Shad cept at 200 psi, where jets de45 (N6299) Parallel 100 Shad flect and are parallel for long 46 (N6299) Parallel 200 Shad distance. 47 (N6299) Parallel 200 Shad 48 (N6299) Parallel 50 Schl 49 (N6299) Parallel 50 Schl 50 (N6299) Parallel 100 Schl Angle of expansion about 24". 51 (N6299) Parallel 100 Schl 52 (N6299) Parallel 200 Schl 53 (N6299) Parallel 200 Schl 54 (N6299) Perp. 50 Schl 55 (N6299) Perp. 50 Schl 56 (N6299) Perp. 100 Schl Slight identity of jets at 50 57 (N6299) Perp. 100 Schl psi, sharp identity at 100 and 58 (N6299) Perp. 200 Schl 200. End expansion about 6~. 59 (N6299) Perp. 200 Schl,,9'~.- i,',,, 12

-- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Air Negative Blower -Position PressurePhoto, Remarks 60 (N6299) Perp. 50 Shad 61 (N6299) Perp. 50 Shad 62 (N6299) Perp. 100 Shad 63 (N6299) Perp. 100 Shad 64 (N6299) Perp. 200 Shad 65 (N6299) Perp. 200 Shad 66 49X Perp. 100 Schl Sharp identity of jets at 100 67 (N7915) Perp. 100 Schl psi but not at 200 psi. Ends 68 (N7915) Perp. 200 Schl show angle, but center jets 69 (N7915) Perp. 200 Schl move straight down. 70 59 Perp. 100 Schl 71 (N8737) Perp. 100 Schl 72 (N8737) Perp. 200 Schl Negatives useless; see 74-77. 73 (N8737) Perp. 200 Schl 74 N8737) Perp. 100 Schl Sharp identity of jets. 50 75 (N8737) Perp. 100 Schl contraction at end at 100 psi, 76 (N8737) Perp. 200 Schl 0~ at 200 psi. 77 (N8737) Perp. 200 Schl 78 119 Perp. 100 Schl 79 (N4847) Perp. 100 Schl Slight identity of jets. Slight 80 (N4847) Perp. 200 Schl contraction at end. 81 s(N4847) Perp. 200 Schl - -, 82 59 Perp. 100 Schl Slight identity of jets. 100 83 (8~skirt)) Perp. 200 Schl expansion at end. 84 5 Perp. 100 Schl Sharp identity of jets. 85 5 Perp. 200 Schl Turbulence between jets. 86 5 Perp. 200 Schl 87 Not Taken 88 94(5445) Parallel 200 Schl No apparent change. 89 3/32 rad Parallel 200 Schl 90 94(5307) Parallel 200 Schl No apparent change. 91 5/32 rad. Parallel 200 Schl ] NOTE: Blowers are indicated by slots, with serial number given at first -usage. Positions are parallel-w-light beam parallel to slot rows.a-nd perpendicular (perp) —"light beam perpendicular to slot rows. Air pressures are indicated as pounds per square inch gauge. "Photot indicates the type of photography employed —shadow (shad) or schlieren (schl).