THE UNIVERSITY OF MI CHI GAN COLLEGE OF ENGINEERING Department of Meteorology and Oceanography ERRORS IN MEASUREMENTS OF WIND SPEED AND DIRECTION MADE WITH TOWER- OR STACK-MOUNTED INSTRUMENTS Gerald C. GlL^J^,. Lars E. 01'aSon' ",. Motozo uda ORA Project 0697.3' sponsored by: U. S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE PUBLIC HEALTH SERVICE NIH GRANT NO. AP-00233-03 WASHINGTON, D.C. administered through: OFFICE OF RESEARCH ADMINISTRATION ANN ARBOR June 1966

6/Ym ^^l /^^)

TABLE OF CONTENTS Page LIST OF TABLES v LIST OF FIGURES vii ABSTRACT ix CHAPTER 1 INTRODUCTION 1 2. EXPERIMENTAL STUDY 5 2.1 Smoke Tests 6 2.2 Flag (Wind Direction) Tests 6 2.3 Anemometer (Wind Speed) Tests 7 3. CALIBRATION OF WIND TUNNEL 10 3.1 Wind Direction Tests 10 3.2 Wind Speed Tests 10 4. RESULTS 12 4.1 Smoke Tests 12 4.2 Wind Direction Tests 12 4.2.1 Tower 12 4,2.2 Stack 14 4.3 Wind Speed Tests 14 4.3o1 Tower 16 4.3,2 Stack 17 5. SUMMARY 19 6. RECOMMENDATIONS CONCERNING MOUNTING OF WIND SENSORS 20 6.1 Relatively Open Towers 20 6.2 Triangular Solid Towers 21 6.3 Circular Stacks 22 BIBLIOGRAPHY 24 iii

LIST OF TABLES Table Page 1. Wind speed distribution in open tunnel. 25 2. Comparison of wind speed in open tunnel at two levels. 26 3-9. Wind speed distribution around open tower. 27-40 10,11. Wind speed distribution around solid tower. 41-44 12. Wind speed distribution around cylindrical stack. 45,46 13. Wind speed distribution in a vertical plane behind the center of a 47 cylindrical stack. 14. Summary of findings from an analysis of Figures 19 through 21. 48 v

LIST OF FIGURES Figure Page 1o Working section of low speed wind 49 tunnel. 2. Scale model of WJBK TV tower, 50 Detroit, Michigan. 3 Arrangement for smoke tests. 51 4. Flag arrangement on bar. 51 5 Boom and sensor locations. 52 6. Photographs of tower, stack, 53 anemometers and flags in tunnel. 7-9. Photographs of smoke tests. 54-56 10. Flag tests (wind direction tests) 57 around the open tower. 11o Flag tests (wind direction tests) 58 around the stack. 12. Wind speed distribution in open 59 tunnel 13. Wind direction profile around open 60 tower, section I. 14-15. Wind direction profiles around 61,62 open tower, section II. 16. Wind direction profile around solid 63 tower o 17. Wind direction profile around 64 cylindrical stack. 180 Coordinate-systems for anemometer 65 tests (wind speed tests). 19. Wind speed profiles at sensor5 Q, 66-79 located a distance, R, from open tower for winds from 3600 of arco vii

LIST OF FIGURES (concluded) Figure Page 20. Same as Figure 19, but for solid 80-83 tower, 21. Same as Figure 19, but for 84-86 cylindrical stack. 22. Wind speed profiles in a vertical 87 plane behind the center of a cylindrical stack. 23. Recommendations concerning mount- 88 ing of wind sensors on a relatively open tower. 24. Recommendations concerning mount- 89 ing of wind sensors on circular stacks. viii

ABSTRACT Towers and smokestacks affect wind flow so that wind sensors mounted on them do not indicate the true free-air flow. To determine the probable errors in measurements of wind speed and direction around such structures, quarter-scale models were tested in a large wind tunnel. Data on changes in wind speed and direction were obtained by using smoke, tiny wind wanes, and a scale model propeller anemometer. Most emphasis has been placed on a relatively open lattice-type tower, but a solid tower and a stack were also studied. The results show that in the wake of latticetype towers disturbance is moderate to severe, and that in the wake of solid towers and stacks there is extreme turbulence, with reversal of flow. Recommendations for locating wind sensors in the wind field relative to the supporting structure are given for each of the three structures studied. Guidelines are given regarding probable errors in measurements of wind speed and direction around different supporting structures. ix

1. INTRODUCTION The "usual method of measuring the speed and direction of the wind at several levels above the ground is to' erect a tower and mount the sensors out from it on retractable booms. A rule of thumb often used by one of the authors-.is, "Locate the instruments into the prevailing winde, and out from the tower at a distance equal to, or greater than, the tower width at that height.''" This is expected to give fairly accurate observations of wind speed and direction for about 270 degrees of arc. Other meteorologists use different guidelines. A modest search of the literature in June 1965 revealed only one study that documented errors in wind speed and direction measurements caused by the tower itself: Moses and Daubek determined the errors in speed and direction caused by an open tower when the wind sensor (a Bendix Aerovane) was mounted on a boom whose length was about half' the width o'f the tower. They found that: T fowonhe e lee side of the tower may'be reduced to nearly one-half its true value for light winds and nearly 25 per cent for speeds of 10 to 14 mph. An increase in -1

-2measured wind speed exceeding 30 per cent occurred when the wind blowing toward the anemometer made an angle of 20 to 40 deg with respect to the sides of the tower adjacent to the anemometer.... The effect on direction appeared to be relatively smaller, with the greatest mean deviation of 11 deg.... In view of such large errors in indicated wind speed about one tower, and of the...paucity of information on errors caused by others, we decided to make a wind tunnel study of the wind flow pattern around one or more typical tower models and to attempt, from critical analysis of the results, to develop guidelines for the distance and direction at which sensors should be located to achieve a specified accuracy. A quarter-scale model of a typical guyed tower of uniform cross-section was chosen for the first tests. In air pollution studies it is often suggested to the consulting meteorologist that he use an already built smokestack to support his wind instruments. To -obtain some information on the pros and cons of such ~~~~~~......

-3an installation, we decided to make some wind tunnel measurements of the flow around a quarter-scale model of a typical stack. The results of these tests are reported here. After beginning this study, we learned that a similar wind tunnel study on a quarter-scale tower.model, conducted by Hsi and Cermak, was nearing completion at Colorado State University. This study 5 has been reported in details, and a shorter technical article has been submitted f publication The article has been submitted for publication o The findings reported for an open tower are very similar to ours. Since completing our wind-tunnel tests with a simulated smokestack, we have learned of a report, "Estimation of the Effect of the 300-Meter Meteorological Mast Structure on the Wind-Gauge Readings," by a group of Russian scientists This report presents first a theoretical approach and then an experimental study of correction factors to be applied to wind observations made 7.5 meters out from a 2.4-meter diameter pipe mast -(boom length ~ 3D). This was not a model study but an in-place study,

-4made at 225 meters on a 300-meter mast. They used six wind-speed sensors-and two wind-direction sensors mounted on booms extending as much as 22 meters (D9D) from the mast. Since our data are limited to one wind speed and one Reynolds number, a direct comparison between the results of our smokestack study and their study is difficult. However, there is agreement in the general conclusions on arranging wind sensors on cylindrical masts or smokestacks.

2. EXPERIMENTAL STUDY The study was performed in a low-speed returnduct tunnel at The University of Michigan, having a working section 5 1/2 feet high, 8 feet wide., and 30 feet long (Figure 1). The quarter-scale tower model measured 30 inches across each face and 62 inches high (Figure 2); it was an accurate reproduction of two sections of the 1050-foot WJBK-TV tower in Detroit, which measures approximately 10 feet across each face. The model -shown in Figure 2, incorporates the elevator guide rails, radio frequency ducting, and other details of the structure. Speed and direction tests were conducted at two heights above the tunnel floor; see sections I and II, Figure 2. To simulate the greatest -possible density of elements in a tower of triangular cross-section, speed and direction tests were conducted with a solid model, each of the three faces being covered with cardboard (Figures 5b and 6c). Glazed sewer pipe, 14 inches in diameter and extending from floor to ceiling, was used as a quarter-scale model of a stack 56 inches in diameter (Figures 5a and 6d). -5

-6The wind speed in the tunnel was kept essentially constant for all tests, the speed being recorded by a Gill propeller anemometer4 (Figure 6, parts a through d), connected to an Esterline Angus milliampere recorder. Three different types of tests were performed. 2.1. Smoke Tests. Oil fog smoke was injected from one or more quarter-inch diameter outlets into the undisturbed air stream in front of the horizontally supported tower, and instantaneous photos of the smoke trail were made (Figures 3, 7, 8, and 9). The photographs were made with a 35-mm camera (Honeywell Pentax with 55-mm Takumar lens). The air stream was viewed through the plate glass windows which make up one wall of the tunnel. 2.2. Flag Tests. Flags mounted on a bar (Figures 4, 6b, and 6c) were moved to different locations in front of and behind the vertical tower or stack. Fivesecond time-exposure photographs were taken through windows in the ceiling of the tunnel to determine the flow and turbulence pattern (Figures 10 and 11) By using the film negatives and a projector, the pictures

-7were transferred directly to the graphs of Figures 13 through 17. The flags were an adaptation of Sakagami's 71 wind-direction sensors and The University of Michigan 3 flag sampler. The top edge of each flag had a narrow strip of opaque white Scotch tape (Figures 4 and 6c) which contrasted well with the black Bristol board on the tunnel floor (Figures 10 and 11). 2.3. Anemometer Tests. A second propeller anemometer, wired to a second Esterline Angus recorder but with its propeller cut to 2 3/4 inches in diameter, was employed to simulate a quarter-scale model of the Aerovane propeller (Figure 6a). This was moved to selected locations about the tower and its readings compared with readings taken at the same locations when the tower was not in the tunnel. These data yielded the velocity ratios given in Figures 19 through 22. The locations were selected to correspond to those at which an Aerovane would be mounted on a retractable boom extending 1/2, 1, or 2 tower widths (1/2D, 1D, 2D) out from the tower. One limitation in the setup used was that the anemometer was fixed in a forward direction and therefore measured only the flow

-8component parallel to the axis of the tunnel. For other wind directions the instrument response closely approximates the total speed multiplied by the cosine of the angle between the wind direction and the tunnel 4 o axis. For an angle of 10 the response is 98% of true; for 20 it is 94% of true. In our recommendations we have considered errors in wind direction greater than 10 degrees to be unacceptable. Since the directional error of the propeller anemometer is only 2% or less for such angles the limitation mentioned above is of no real consequence. Since the density of the tower structures is nonsymmetrical (Figure 2) five different boom arrangements could be considered for the open tower. These are shown in Figure 5. Furthermore, the difference in tower density with height has been considered. Sections I and II (see Figure 2) each have a total frontal area 2 A = 10 x 31 = 310 in; the shadow area of Section I is 2 2 A1 ^ 80 in, and that of Section II is A 140 in Hence the "shadow density" of Section I, A1/A - 80/310 x 100% ~ 26%, and that of Section II, A2/A 140/310 x 100% <. 45%.

-9To assure kinematic and dynamic similarities as well as possible, we applied criteria set forth by 5 Hsi and Cermak. The tower they used is very similar, in both structure and size, to the one used for this study. To determine whether difference in Reynolds number influences the wind distribution, Hsi and Cermak made velocity measurements for 10, 30, and 60 fps, and concluded that "the model wind data are not influenced by Reynolds number variation and are similar to the prototype wind-field." Hence our data for the open tower model should be valid at least for the wind-speed range of 10 to 60 fps (7 to 40 mph). For the solid tower and the smokestack, the effect of scaling upon the wind-field depends upon the Reynolds number. For the quarter-scale model, the Reynolds number is 1/4 that of a full-sized tower. This fact makes generalization of the tests with the solid tower and smokestack more difficult. It should be mentioned, however,.. that the solid tower was chosen only as an extreme case of the open tower, where the shadow density = 100%.

3o CALIBRATION OF WINED TUNNEL 3.1. Wind-Direction Tests. The open tunnel flow was tested for laminar flow by means of flagso A pipe with the flags mounted on it was placed in traverses at distances of 0, 5, 10, and 15 feet (measured from the forward end of the working section of the tunnel), and at heights of primary concern (.Sections I and II, Figures 1 and 2). Time-exposure photographs were taken through the windows in the ceiling of the tunnel. Analysis of the photographs indicates that the flow in the tunnel is undisturbed throughout; that is, there was no indication of convergence or divergence of flow within the accuracy of the technique (probably +3 degrees of axial flow) 3.2. Wind-Speed Tests. The wind-speed distribution in the "open tunnel" was tested for a given constant flow in the opening of the tunnel by use of the second propeller anemometer. The results are given in Table 1 and Figure 12. Figure 12 shows, the speed is slightly higher near the walls of the tunnel than in the center; moreover, the flow becomes somewhat more uniform farther downwind. in the tunnel, in the -10

-11area where most of tests were conducted. No significant difference of wind speed with height was found, at least not at the intermediate heights at which the experiments were run; see Table 2. The wind distribution in the open tunnel has been applied in evaluating the relative wind speed around each obstacle,

4. RESULTS 4.1. Smoke Tests. The pictures of the smoke flowing by the tower structure show that there are considerable disturbances in the flow pattern; see Figures 7 through 9. They indicate that cKarman vortex streets are formed behind the pipes in the tower structure. Although some qualitative data could be obtained from the smoke pictures, no quantitative data for distances of 1/2D, ID, and 2D from the tower could be obtained; therefore the smoke tests were quickly abandoned in favor of the flag and anemometer tests. 4.2. Wind-Direction Tests. 4.2.1. Tower. The wind-direction profiles analyzed from photographs taken in front of and behind the open tower are presented in Figures 13 through 15, and those for the solid tower in Figure 16. The general findings for the open tower are as follows: (1) Flow in front of the tower is laminar; flow in its wake is turbulent. (2) The width of the turbulent wake behind the -12

-13tower is fairly constant, ID to 1 1/2D, for distances of 1/2D to 6D downwind from the tower, and is about the same for Sections I and II and for the different orientations of the tower. See Figures 14 and 15. (3) For distances up to ID, the wake behind Section II is considerably more turbulent than that behind Section I; but at downwind distances greater than 2D there is little difference in the intensity of the turbulence. (4) The turbulence is rather strong behind the center of the tower, even at distances of more than 4D. The general findings for the solid tower are as follows: (1) Upstream from the tower the flow is not turbulent, but measurements of the direction of flow are in error at least 1D ahead of the tower (2) Downstream from the tower the flow is extremely turbulent at all distances up to 6D, and for the full width of our observations (2D). At downstream distances up to 1/2D, the flow direction near the center of the wake is reversed. At downstream distances up to 2D, and for a width of at

-14least 1D, the indicated wind direction may be from any azimuth. (3) All measurements of wind direction downstream from the tower were completely unreliable. 4o2o2. Stack. For a cylindrical obstacle the wind direction profile is dependent upon the Reynolds number, and the picture shown in Figure 17 cannot be generalized as well as could the results for the open tower. The general findings for the stack are as follows: (1) The flow in the wake of the stack is very turbulent (2) The turbulence reaches many stack diameters downwind o (3) One sector of the wake has a negative direction. (4) There are Karman vortex streets. 4.3. Wind-Speed Tests. The data obtained are summarized in Tables 3 through 13, and graphically represented in polar diagrams, Figures 19 through 21. Figure 18 shows the coordinate-systems used in the figures o

-15In the tables, Column 1 "Ref. No." is a reference number. Column 2 "Coordinates" give the relative location of the sensor; see Figure 18. Column 3 "U" is measured wind speed (mph) corrected for change in basic flow.* Column 4 "Flow Type" is a subjective evaluation of the flow stability made from the charts and/or observed during the runs. LL stands for "very laminar" L stands for "laminar" T stands for "turbulent" (%+0.3 mph) TT stands for "very turbulent" (>+0.5 mph). Column 5 "Uo" is undisturbed flow (mph), read from wind speed distribution in the open tunnel; see Figure 12, Column 6 "U/Uo" is relative wind speed, in percentages. On the graphs U/Uo is plotted against wind direction for different arrangements * The wind-tunnel propeller was driven by a d-c motor whose speed changed by as much as 5% owing to fluctuating line voltage o By using a control anemometer near the forward end of the working section of the tunnel (see Figure 1) the indicated wind speeds were corrected.

-16of tower and boom, and for different lengths of the boom. In general, the results for flow types agree with those from the flag tests. Thus, e.g., in cases where TT or T is indicated, the flags are either rotating or fluctuating more than 20, and for LL the flags indicate steady flow. 4.3.1, Tower. A series of runs for all five cases, A-E (see Figure 18), was made at Section I (the less dense section) for 1/2D, 1D, 2D, 3D, and 4D; see Tables 3 through 7 and Figure 19, parts a through j At Section II, runs were made for cases A and E only; see Tables 8 and 9 and Figure 19, parts k through n. For the solid tower, cases A, B, and C are covered by one run, Table 10, Figure 20a; and cases D and E by one run, Table 11, Figure 20b. The general findings for the open tower are as follows (1) In general, the density distribution in the tower structure influences the location of the wake,. the angle occupied by the wake, the effect of the shading, and the location of maximum readingso

-17(2) The mentioned effects are more pronounced when shorter booms are used. (3) There is no significant difference in the disturbing effect at Sections I and II. This might be because the height of the horizontal tower member of Section II (see Figure 2) is small compared to the length of the boom; hence the difference in structure density as seen by the sensor is less than the two density numbers would tend to indicate, (4) At a boom length of ID, the minimum wind speed varies significantly for the five cases. The general findings for the solid tower are as follows: (1) The distribution features for the solid tower are similar to those for the open tower, but the values of the ratio U/Uo are greatly accentuated. (2) At distances up to 2D in the center of the wake, wind flow is reversed (U/Uo<0) 4.3.2. Stack. A series of runs for boom lengths of 1/2D, ID, 2D, 3D, and 4D,as well as some for 6D, 8D, and 1OD, were made for the cylindrical stack; see Tables 12 and 13, and Figure 21. These data are

- 18valid for the Reynolds number corresponding to a quarter-scale model exposed to a wind speed of about 8 mph, but no generalization for other Reynolds numbers is claimed. The general findings for the stack are as followss (1). The sector occupied by the wake increases with decreasing boom length. (2) The deviation of the velocity ratio U/Uo from 100% increases with decreasing boom length. To determine the effect of mounting wind sensors near the top of a stack, observations were made in the wake of a stack not extending to the ceiling. Table 13 and Figure 22 present the velocity ratios U/Uo in a vertical plane through the center line of the stack. The general findings for the area near the top of a stack are as follows (1) The influence of the top of the stack becomes negligible at levels -3D below the top of the stack, at least for distances downwind up to 6Do (2) The velocity ratio U/Uo improves rapidly (appreaching 100%) as the observation level approaches and exceeds the top of the stack,

5o SUMMARY The results of this study show that: (1) There is a significant disturbance in the air flow pattern around and behind a structure, which is a complex function of tower structure and of sensor position. (2) There is a sector in the wake of the tower in which the measured value of the component of the wind speed parallel to the undisturbed flow is lower than the true value, and the measurement of flow direction is in error. (3) -There are other sectors, most of them close to the wake-sector, in which the measured speed is larger than the true speed but the measured direction is not in serious error. (4) There is a general correlation between the magnitude of fluctuations in wind direction and the decrease in wind speed. Table 14 summarizes the limits of reasonable accuracy of observations as given by Figures 19 through 21o The recommendations given in the next section are based on this tableo

6. RECOMMENDATIONS CONCERNING MOUNTING OF WIND SENSORS 6.1. Relatively Open Towers. (1) For measurements of wind speed that are accurate within -10% of the true value, we recommend that the sensor be placed not less than 1D out from the tower, and located out from the corner into the wind of primary concern. Installation types A, B, and C are preferable to types D and E. Preferably locate sensors at heights of minimum tower density, above or below horizontal cross members. For this extension and location, measurements of speed are true within -10% for a 310 sector of arc. If the boom is extended to 2D, the wind speed is accurate within -10% for a 3300 sector of arc. For these two arcs, measurements of wind direction are accurate within at least -10, and + o probably within -5 (2) With a boom extension of 1D, observations of wind speeds accurate within -5% of true can be obtained within only a 180 sector of arc. If the boom is extended to 2D, measurements of wind speed are accurate within -5% for a 240 to 270 sector of arc. *For'"Shadow densities" not exceeding 40 percent. -20

-21For these two arcs, measurements of wind direction are probably accurate within -5 (3) If wind measurements accurate within -5% in speed and -5 in direction are needed for the complete 360 of azimuth, we recommend that two sets of speed and direction sensors be used, placed 180 apart, located as in (1) above, and with a boom extension of not less than 1 1/2D. (4) When a single wind sensor is to be used, a straight line passing from the sensor through the center of the tower should point in the direction of minimum concern. (5) If circumstances preclude the use of the settings specified above, representative results can be calculated from the figures given for other boom lengths, tower structures, and/or tower orientations. Most of these recommendations are diagrammatically shown in Figure 23. 6.2. Triangular Solid Towers. (1) Preferably, solid towers should not be used for meteorological platforms.

-22(2) If only a solid tower is available, this study indicates* that, if a boom extending 2D is used, wind-speed measurements accurate within -10% of true can be expected for an arc of about 240. For this same arc measurements of wind direction are probably true within -5 6.3. Circular Stacks. (1) Preferably, stacks should not be used as meteorological platforms. (2) If a stack must be used, this study indicates* that by using a boom extension of 3D one can obtain wind speed measurements accurate within -10% of the true value through an arc of about 180, and wind direction measurements true within -5 for an arc of about 300; see Figure 24. (3) If winds accurate within -10% in speed and -5 in direction are needed for the complete 360 *This study was performed with a Reynolds number of approximately 2.7 x 105 for the solid tower (L = side of tower); and 1.3 x 105 for the stack (L - diameter), corresponding to an average tunnel speed of 8 to 9 mph and quarter-scale models of supporting structures. We do not claim that the results are directly applicable to other wind speeds or other values of the Reynolds number.

-23of azimuth, we recommend using two sets of speed and direction sensors, 180 apart and located not less than 2D out from the stack. (4) The accuracy of speed and direction measurements can be markedly improved by locating the top set of wind sensors 1/2D or higher above the stack.

BI LIOGRAPH.. 1o Borevenko, E. V., et al., "Estimation of the Effect of the 300-Meter Meteorological Mast Structure on the WindGauge Reading, " in Investiqation of-the Bottom 300-Meter Layer of the Atmosphere, N. L. Byzova, 1963, translated into> English by Israel Program for Scientific Translations, 1965, pp, 82-92. 2. Cermak, J. E., and J. D. Horn, Tower Shadow Effect, paper submitted for publication to Journal of Geophysical Research, 1966. 3. Harrington, J. B., G. C. Gill, and B. R. Warr, "'High Efficiency Pollen Samplers for Use in Clinical Allergy," J. of Allergy, Vol 30, No. 4, 1959, ppo 357-3755 4. Holmes, R. M., G. C. Gill, and M. W. Carson, "A Propeller-Type Vertical Anemometer," J. Appl. Meteor., Vol. 3, No, 6, Dec. 1964, pp. 802-804. 5, Hsi, G., and J. E. Cermak, Meteorological-Tower Induced Wind-Field Perturbations, Tech. Report, College of Engineering, Colorado State University, October 1965. 6. Moses, M, and H. Go Daubek, "Errors in Wind Measurements Associated with Tower-Mounted Anemometers," Bulletin of the American Meteorological Society, Vol. 42, No. 3, March 1961, ppo 190-194. 7 o Sakagami, J., On the Structure of Atmsperic Turbulence near the Ground, Nat. Sci. Reports, 1951, Ochanomizu University, Tokyo, Vol. 1, pp. 40-50....-2~~~4~

TABLE 1 Wind Speed Distribution in Open Tunnel. (z = 15 inches)* Refo Coordinate*'Uo Ref. Coordinate* Uo No,. _ (inches) (mph) No. (inches) (mph) 1 -60 -36 8.8 41 60 -36 8.85 2 -24 8.35 42 " -24 8o6 3 -12 8.3 43 " -12 8o5 4 0 8.2 44 " 0 8.55 5 12 8.3 45 i 12 8.6 6 " 24 8.5 46 " 24 8.85 7 " 36 9.0 47 36 9.15 11 -30 -36 8.8 51 90 -36 8o9 12 " -24 8.4 52 -" 24 8.75 13 " -12 8.3 53 " -12 8 55 14 0 8.3 54 0 8.6 15 " 12 8.4 55 12 8.65 16 " 24 8.65 56 24 8.85 17 36 9.1 57 36 9.15 21 0 -36 8.8 61 120 -36 9.0 22 -24 8.55 62 " -24 8.65 23 " -12 8.3 63 " -12 8.6 24 0 8.35 64 " 0 8.6 25 12 8.5 65 " 12 8 65 26 24 8.8 66 " 24 8.95 27 36 9.1 67 36 9.15 31 30 -36 8.8 81 180 -36 8.85 32 " -24 8.6 82 -24 8.65 33 " -12 8.45 83 " -12 34 " 0 8.5 84 " 0 8.7 35 " 12 8.55 85 " 12 9o0 36 " 24 8.8 86 1 24 37 " 36 9.25 87 " 36 9.45 x = along tunnel axis y = across tunnel z = height -25

TABLE 2 Comparison of Wind Speed in Open Tunnel at Two Levels. Coordinates* Uo (mP ph) o, l U 15-Uo, 31 x y z = 15 z =31 (inches ) ( inches) (inches) ~ (mph) A -30 -36 8.8 8.5 0.05 0 0 8.35 8.4 0.05 60 0- 8.55 8.5 0.05 180 0. 9.45 9.65 0.20 *x along tunnel axis y = across tunnel z = height -26

TABLE 3 Wind Speed Distribution around Open Tower At Section I, Case A. Ref. Coordinates U Flow* Uo U/Uo No. r e (mph) Type (mph) (%) 1 1/2D 360 8.4 LL 8.7 97 2 8 3454 894.100 3 " 330 8.3 " 8.3 100 4 315 8.7 L 8.35 104 5 300 8.7 LL 8.5 102 6 " 285 9,0 8.45 107 7 " 270 9.0 " 8.4 107 8 " 255 9.1' 8.45 108 9 " 240 9.1 1 8.45 108 10 " 225 9. 1 " 8.45 108 11 " 210 9.4 L 8.4 112 12 " 195 9.15 LL 8.45 108 13' 180 5.8 T 8.8 66 14 " 165 6.5 T 8.6 76 15 " 150 9.0 LL 8.5 106 16 " 135 9.15 " 8.4 108 17 120 9.2 " 8.45 109 18 " 105 9.1 9 8.45 108 19 " 090 9,1 l 8,45 108 20 " 075 9.0 " 8.45 107 21 " 060 8.7 " 8.4 104 22 " 045 8.7 L 8.35 104 23 " 030 8.6 LL 8.3 103 24 " 015 8.4 " 8.4 100 25 " 000 8.4 " 8.7 97 1 1D 360 8.5 LL 8.7 98 2 " 345 8.4 T 8.35 101 3 " 330 8.3 LL 8.25 101 4 " 315 8.75 " 8.5 103 5 " 300 8.8 " 8,7 101 6 "285 9,1 " 8.8 104 7 " 270 9.2 " 8.75 105 8 255 9.5.5 8.8 108 9 " 240 9.1 " 8.75 104 10 " 225 9.3 " 8.65 107 -27

TABLE 3 (concluded) Wind Speed Distribution around Open Tower At Section I, Case A, Ref. Coordinates U Flow* Uo U/uo No. r - (mph) Type (mph) (%) — 11 ID. 210 9.3 LL 8.55 108 12 195 9.2 " 8.65 106 13 180 6.1 T 8.8 69 14 165 6.4- TT 8.6 76 15 - " 150 8.3 T 8.5 98 16 " 135 9.1 LL 8.5 107 17 120 9.3 " 8.75 107 18 105 9.5 " 8.8 108 19 " 090 9.2 " 8.75 105 20 075 9.1 " 8.8 104 21 " 06-0 8.8 " 8.7 101 22 "045 8.75 8.5 103 23 -" 030 8.6 " 8.3 103 -24 " 015 8.4 8.35 101 25." 000 8.5 T 8,7 98 1?2D 225 9.8 LL 9.3 106 2 " 210 9. 6 L 8.8 109 3 " 195 9.7 L 9.3 10:4 4 " 180 6.4 T 8.8 73 5 " 165 8.7 T 8.55 102 6 " 150. 9.4 LL 8.6 109 7 135 9.3 LL. 8.9 104 1 4D 2 " 180 6.6 T 9.0 73 3 " 165.9.5 LL 8.5 112 4 " 150 10.1 LL 9.2 110 *LL = very laminar L = laminar T = turbulent (% -+0.3mph) TT = very turbulent (> +0.5mph) -.-28

TABLE 4 Wind Speed Distribution Around Open Tower At Section I, Case B. Ref. Coordinates — U Flow UoQ -U/Uo No'. r 9e _ (mph) Type (mph) 1. 1/2D 360 -.8.4 LL 8.7- 97 2 345 8.4" 8.4 100 3 " 330 8.3 " 8.3 100 4 " 315.8.7 L 8.35 104 5 " 300 8.75 " 8.5 103 6 " 285 9.0 LL 8.45 107 7 " 270 8,9 " 8.4 106 8 " 255 8.9 " 8.45 105 9 " 240 9.2 " 8.45 109 10 225 9.2 " 8.45 109 11 " 210 9.25 L 8,4 107 12 " 195 9.2 LL 8.45 108 13 " 180 5.8 T 8.8 66 14 " 165 6.3 L 8.6 73 15 150 8.4 T 8.5 99 16 " 135 8.7 T 8.4 103 17 120 9.2 LL 8.45 109 18 " 105 8.9 " 8.45 105 19 090 9.1 8.45'107 20 " 075 9.0 " 8.45 107 21 " 060 8.75 " 8.4 104 22 " 045 8.7 L 8.35 104 23 " 030 8.6 LL 8.3 103 24 " 015 8.4 " 8.4 100 25 " 000 8.4 8.7 97 1 1D 360 8.5 LL 8.7 98 2 " 345 8.4 T 8.35 101 3 330 8.3 LL 8.25 101 4 " 315 8.75 " 8.45 103 5 " 300 8.8 " 8.7 101 6 " 285 9.1 " 8.8 104 7 " 270 9.2 " 8.75 105 8 " 255 9.5 " 8.8 108 9 " 240 9.3 " 8.75 106 10 " 225 9.3 " 8.65 107 "-29

TABLE 4 (concluded) Wind Speed Distribution Around Open Tower At Section I, Case B, Ref., Coordinates,U Flow Uo U/U No.- - r - (mph) Type (mph) (%) 11 1D.210 9.15 LL 8.55 107 12 i 195 9.2 " 8.65 106 13 " 180 6.1 T 8.8 69 14 165 5.7 T 8.6 66 15 " 150 8,7 T 8.5 102 -16 135 9.2 LL 8.5 108 17 120 9.3 " 8.75 106 18 105 9.5 " 8.8 108 19 090 9.1 8.75 104 20 075,,9.1 " 8.8 104 21 " 060 8.8 " 8.7 101 22 045 -8.75 " 8.5 103 23 " 030 8.6 " 8.3 103 24 " 015 8.4 T 8.35 101 25 000 8.5 LL 8.7 98 1 2D 225'9.8 LL 9.3 105 2' " 210 9.6 " 8,8 109 3 "" 195 9.7 " 9.3 104 4 " 180 6.4 T 8.8 73 5 " 165 8.6 T 8.55 100 6 " 150 9.35 T 8.55 109 7 " 135 9.5 LL 8.9 107 1 4D 195 2 1 180 6.6 T 9.0 73 3 " 165 8-.7 T 8.7 100 4 " 150 9.4 LL 9.3 101 -30

TABLE 5 Wind -Speed. Distribsution Around Open Tower At Section I, Case C, Ref. Coordinates U Flow Uo U/Uo No.....,. (mph) Type (mph) (%)' 1: 1/2D 360 8.4 LL 8.7 97 2 345 8.4 8.4 100 3 330 8.2 " 8 3 99 4 " 315 8.7 L 8035 104 5 300 8.8 7 LL 8.5 103 6 " 285 9.0 " 8 45 107 7 270 9.0 " 8.4 107 8 " 255 9 1 " 8.45 108 9 ": 240 9, 1 9 8,45 108 10 "1 225 91 8.45 108 11 " 210 9 35 8.4 111 12 " 195 9,1 T 8.45 108 13 " 180 6: 6 8.8 75 14- " 165 4.5 " 8.6 52 15 150 6..7 " 8.5 79 16 " 135 8.0 TT 8.4 95 17 —,,"l 120 9.25 L 8.4 110 18: ". 105 9. 0: LL 8.45 107 1-9: l'" 090 9.1 8.45 108 20 " 075 9.0 8.45 106 21. - - 060 8.o7 o 8.4 104 22 " 045 8~.7 " 8 35 104 23 " 030 8-.6 1- 8.3 103 24 " 015 8.4 " 8,4 100 25 000 8 8.4 1: 870 97 1 1D 3-60 8-5 LL 8.7 98 2 " 345 8.4 8.35 101 3 " 330 803 " 8.25 101 4 315 8.75 L 8.5 103 5 " 300 8.8 LL 8.7 101 6 " 285 9 1 L 8.8 104 7 " 270 9 2 LL 8.75 105 8 " 255 9o5 a 8.8 108 9 " 240 9.1 " 8.75 104 10 " 2-25 9.15 L 8.65 106 -31

TABLE 5 (concluded) Wind Speed Distribution Around Open Tower At Section I, Case C. Ref.7: "Coordinates U Flow UO U/U0 No. r - Oe (mph) Type (mph) (%) 11 1D 210 9.2 LL 8.55 107 12 7'" 195 9.1 " 8.65 105 13 " 180 7.1 T 88 8 81 14 " 165 7.6 " 8.6 89 15 " 150 8-.4 4" 8.5 99 16 " 135 9.3 L 8.5 109 17 " 120 9.25 LL 8 6 107 18 -" 105 9.1 " 8 7 104 19 -' - 090 9.2 " 8.75 105 20 " 075 9.1 " 8.8 104 21 - " 060 8.8 " 8.7 101" 22;- 045 8 75 8.5 103 -23 " 030 8.6 " 8.3 103 2-4 " 015 8.4 " 8.35 10-1 25 " 000 8.5 8.7 98!1- 2D: 225 9.55 LL 9.3 103 2 2 210 9.35 8.8 106:3 "- - 195 9.;.7 " 9.3 104 4." 180 7.4 T 8.8 84 5 " 165 9.3 L 8.55 109 -6 "- 150 9 3 LL 8.6 108 7 " 135 9.5 ", 8.9 107i 2 4D 180 7.7 T 9.0 86 3 " 165 9.0 " 8.5 106 4 " 150 9.0 LL 9.15 108 -32

' TABLE 6 Wind Speed- Distr ibution Around Open Tower -At Section I, Case D. Ref. Coordinates U Flow Uo U/Uo No. r 0.p (mph) Type,(mph) (%) 1 1/2D 360 8.4 LL 8' 4 100 2 " 345 8. 1 " 8,3 98 -3'" 330 8.2 8.3 99 4 " 315 8.2 " 83 99 5 " 300 8.4 - 8.3 101 6 " 285 8.7 LL 8.35 104 7 270 8.85 8.35 106 8 - 255 9.1 " 8.35 109 9 " 240 9.2 L 8.35 110 10 225 7.8 T 8.35 93 11 " 210 8.4 8.4 100 12 195 5.95 " 8.5 69 13 -i - 180 6.0 8,6 70 14 " 165 8.6 LL 8.5 101 5 " 150 9.4 L 8.4 112 16 " 135 7 5 T 8.35 90 17 " 120 9.3 LL 8035 111 18 "105 9.1 " 8.35 109 19 " 090 8 85 " 8,35 106 20 075 8.7 " 8.35 104 21' 060 8.4 " 8 3 101 22 045 8.2 L 8.3 99 23 " 030 8.2 LL 8.3 99 24 " 015 8.1 8.3 98 25 " 000 8.4 " 8.4 100 i 1D 360 8.6 LL 8.4 102 2 " 345 8.7 - 8.5 102 3 330 8,2 LL 8.3 99 4 " 315 8.4 - 8035 101 5 " 300 8.5 LL 8.4 101 6 " 285 8.8 " 8.55 103 7 " 270 8.95 " 8.6 104 8 " 255 9.1 " 8.6 106 9 "i 240 9 3 " 8.55 109 10 -225 9 o2 L 8.5 110 -33

TABLE 6 (concluded) Wind Speed Distribution Around Open Tower At Section I, Case D. RefO Coordinates U Flow Uo U/Uo No. r e0 (mph) Tvye (mph) (%) 11.. 1D.210 7.6 TT 845 90 12 " 195 6.5 85 72 13 180 6.2 8.6 72 14 D 165 9.0 L 8.5 106 15 " 150 -7.7:TT 8.45 91 16 135 9.o4 LL 8.5 111 17 120 9.4 " 8.55 110 18 " 1 0 5 0O — 18 - 8 105 9 o l1 8.35 108 19 " 090 8o.95 8.6 104 20 0. "75'88'" 8.55 103 21 060 8.5 8.4 101 22 " -045 8.4 8.35 101 23 030 8.3 - 8.25 101 24 " 015 8.,75 8.5 103 25 - 000 8.6 - 84 102.1 I -2D 240 2 ~' 225 9,6 LL 8.9 107 3 210 9,3 " 8.6 108 4 " 185 7.9 TT 8.5 93 5 " 170 7.1 " 8.7 92 6 " 165 7.6, 8.5 90 7, 150.9o4 LL 806 109 8 "1.135 9. 6. 0 8.9 107 1 4D 195 9.3 LL 8.7 108 -2 180 7.6 TT 8o9 86.3 165 9.2 LL 8.7 106 -34

TABLE 7 Wind Speed Distribution Around Open Tower At Section I, Case E, Ref. Coordinates U Flow Uo U/U0 No. r.. (Ph) T0pe (mph) yph) (%)1 1/2D 360 8.4 LL 8 4 100 2 " 345 8.1 8,3 98 3 330 8.7 8 4 104 4 " 315 8.2 L 8.3 99 5 " 300 8.4 8.3 101 6 " 285 8.7 LL 8.35 104 7 " 270 8,95 " 8 35 107 8 " 255 9.05 " 8 35 108 9 240 9,3 " 8.35 111 10 " 225 7.5 T 8.35 90 11 210 9.4 L 8,4 112 12 0 195 8.6 LL 8 5 101 13 " 180 3.5 T 8, 6 41 14 " 165 8.6 LL 8.5 101 15 " 150 9.4 L 8.4 112 16 " 135 7.5 T 8-.35 90 17 " 120 9 3 LL 8.35 11I 18 l 105 9.05 " 8,35 108 19 090 8 95 8.35 107 20 " 075 8.7 8.35 104 21 - 060 8.4 L 8.3 101 22 " 045 8.2 " 8.3 99 23 " 030 8.7 LL 8.4 104 24 "015 8.1 " 8.3 98 25 " 000 8,4 8.4 100 1 1D 360 8 6 LL 8 4 102 2 345 8,75 " 8.5 103 3 " 330 8.3 - 8,25 101 4 " 315 8.4 LL 8.35 101 5 " 300 8 5 " 8.4 101 6 " 285 8.75 " 8.55 102 7 " 270 9,0 " 8.6 104 8 " 255 9.1 " 8.6 106 9 240 9 4 " 8.55 110 10 " 225 9.4 " 8.5 111 -35

TABLE 7 (concluded) Wind Speed Distribution Around Open Tower -;. At Section I, Case E. Ref. Coordinates U. Flow U0 U/Uo No.) _ r eo 0 (mph) Type (mph) (%) 11 1D. 2-10 7.7 TT 8.45 91 12 1 195 9 0 L 8.5 106 13 " 180 3.6 TT 8.6 42 14? 165 9.0 L 8.5 106 15 - 150 7.7 TT 8.45 91 16. "135 9.4 LL 8.5 111 17 " 120 9.4 " 8.55 110 18 " 105 9.. 8.6 106 19 " 090 9.0 " - 8.6 104 20;o 075 8.75 8,55 102 21 i 060 8.5 8.4 101 22 045 8.4 8.35 101 23 " 030 8.3 - 8.25 101 24 " 015 8.75 - 8.5 103 25 " 000 8. 6 8.4 102 1- 2D 240 2 - 225 9.6 LL 8.9 107 3- 210 94 " 8.6 109 4 -" 195 7.6 T 8.5 90 5 " 180 4.5 T 8.7 52 6 ". 165 7.6 T 8.5 90 T7;- et - 150 9.4 LL 8.6 109 8 " 135 9.6 " 8.9 107 1 4D 195 9.2 LL 8.7 106 2 " 180 6.0 T 8.9 68 3 165 9.2 LL 8.7 106 -36

TABLE 8 Wind Speed Distribution Around Open Tower At Section II, Case A. Ref. Coordinates U Flow Uo U/Uo NOo. _r e9 (mph h) Type (mph) (%) 1 1/2D 360 8 o4 LL 8.7 97 2 " 345 8 4 - 8.4 100 3 330 8.35 LL 8.3 101 4 " 315 8.3 - 8o35 100 5 300 8-7 8.5 102 6 " 285 8.9 8.45 105 7 " 270 9.05 LL 8.4 108 8 11 255 9 ol 8.45 107 9 ~ 240 9 25' 8.45 109 10. 225 9.45 8.45 111 11. 210 9.3 L 8o4 111 12 " 195 8,6 TT 8.45 102 13 " 180 5.2 - 8.8 59 14 8" 165 5o0 " 8.6 58 15 1 150 8.8 8o5 103 16 " 135 8o3 LL 8.4'99 17 " 120 8.35 11 8o45 99 18 105 9.1 8 45 108 19 m" 090 9.1 1 8.45 108 20 " 075 9o0 " 8.45 107 21 " 060 8.7 " 8o4 104 22 " 045 8.7 L 8.35 104 23 m 030 8.6 LL 8.3 103 24 " 015 8.4 - 8o4 100 25 " 000 8.4 - 8o7 97 1 1D 360 8.5 LL 8o7 98 2 " 345 8o4 T 8o35 101 3 " 330 8.35 LL 8.25 101 4 " 315 8.4 " 8.3 101 5 " 300 8.85 m~ 8.4 105 6 m1 285 9.2 1 8,6 107 7 270 9.15 " 8.75 104 8 mm 255 9.3 "8 8.8 105 9 " 240 9 45 8" 8 75 108 10 mm 225 9 o5 m 8.65 110 -37

TABLE 8 (concluded) Wind- Speed Distribution' Around Open Tower At Section II, Case A.Re'f o Coordinates!: U Flow Uo U/U0o NO.. -r eo~ (mph). Type (mph) (%) 11 1D- 210 9.4 LL 8.o55 110 12;, 195 9.2 T 8.45 109 13 Ir 180 6.05 TT 8.5 71 14 " 165 8.15 " 8.6 95 15 - 150 8.4 -" 8.5 99 16 — 135 8.4 LL 8. 3 101 17 "@ 120 8.35 8025 101 18. 105 9.5 8 8 108. 19. 090 9.2 8.75 105 20 -. 075 9.1. 8.8 104 21 " 060 8.8 - 8.7 101 22: - 045 8.75 - 8.5 103 23 " 030 8.6 8.3 103 24. 015 8.4 - 8 35 101 25 000 8.5 - 8.7 98 2 — 2.D2. 225 10 0 LL 9.3 107 3 210 9.6 " 8.8 109. 4... 195 9.5 " 8.55 111 5 I. 180 7.15. TT 8.6 83' 6. Il 165 8.4 8.55 98 7 1 150 9..45 LL- 8.6 110

TABLE 9 Wind Speed Distribution Around Open Tower At Section II, Case E. Ref. Coordinates U Flow U U/U No. r e (mph) Type (mph) (%) 1 1/2D 360 8o4 LL 8,4 100 -2 " 345 8.1 " 8.3 98 3 330 8 7 " 8,4 104 4 315 8.2 L 8.3 99 5 " 300 8-.55 LL 8.3 103 6 " 285 8.8 " 8.35 105 7 " 270 9.0 " 8.35 106 8 " 255 9 25 8.35 11I 9 " 240 9.1 8.35 109 10 " 225 7o0 TT 8.35 84 11 " 210 800' 8.4 95 12 " 195 6.15 T 8.5 72 13 " 180 3 85 TT 8o6 45 14 " 165 6.15 T 8,5 72 15 " 150 8.0 " 8.4 95 16 " 135 7,0 TT 8.35 84 17 " 120 9.1 LL 8.35 109 18 " 105 9o25 " 8.35 111 19 " 090 9 0 " 8,35 106 20' 075 8.8 " 8.35 105 21 " 060 8.55 " 8.3 103 22 " 045 8.2 L 8.3 99 23 " 030 8.7 LL 8.4 104 24 " 015 8.1 " 8.3 98 25' 000 8.4 " 8.4 100 1 1D 360 8.6 LL 8.4 102 2 " 345 8.7 - 8.5 102 3 " 330 8.2 - 8.3 99 4 " 315 8.4 - 8.35 101 5 " 300 8.7 - 8.4 104 6 " 285 8.8 - 8.55 103 7 " 270 9.1 - 8,6 106 8 " 255 9.3 - 8.6 108 9 11 240 9.35 - 8.55 109 10 l" 225 9.4 - 8.5 111 -39

TABLE 9 (concluded) Wind Speed Distribution Around Open Tower At Section II, Case E-, Ref. - -Coordinates U Flow U U/U. No. r 0 (mph) Type (mph) (%) 11 1D 210 7.5 TT.: 8,45 89 12 195 7.9 " 8.5 91 13 " 180 4.7 " 8 7 55 14 " 165 7o9 " 8 5 91 15 150 7.5 " 8.45 89 16 " 135 9.4 - 8.5 111 17 " 120 9.35 - 8.55 109 18 " 105 9.3 - 8 6 108 19. 090 9.1 - 8.6 106 20 " 075 8.8 - 8.55 103.21 060 8.7 8.4 104 22' 045 8.4 - 8.35 101 23 " 030 8.2 - 8.3 99 24 " 015 8.7 - 8.5 102 25 " 000 86 - 8.4 102 1I 2D 225 9.6 LL 8.9 108 2'. 210 9.45 " 8.6 111 3. I" 195 7.7 TT 8.5 91 4 " 180 - 6.2 8,7 71 5 ",6 165 7.7 8.5 91 6 " 150 9.45 LL 8.6 111 1 4D 195 9.65 LL 8.65 111 2 1 180 7.1 TT 8,9 80 3 " 165 9.65 " 8.65 111 -40

TABLE 10 Wind Speed Distribution Around Solid Tower Case A (B,C). Ref. Coordinates U Flow UO U/U No. r e (mph) Type (mph) (%) 1 1/2D 000 - 2 " 015 3 030 7.1 LL 8.3 85 4 " 045 7.6 L 8.35 91 5 " 060 9.2 LL 8.4 109 6 " 075 9.5 T 8.35 114 7 " 090 9.65 " 8.4 115.8 II 105 10.15 L 8.45 120 9 " 120 11.7 " 8.45 138 10 " 135 13.1 T 8.45 155 11 " 150 13.2 " 8.45 157 12 165 8.9 TT 8.4 106 13 11 180 6.7 8.8 76 14 " 195 - -- 8.6 15 " 210 - - 8.5 16 225 -- 8.4 - 17 " 240 11.65 TT 8.4 136 18 I 255 12.3 L 8.45 146 19 270 11.3 LL 8.45 134 20 " 285 10.0 L 8.45 118 21 " 300 9.2 LL 8.4 109 22 " 315 7.6 L 8.35 91 23 330 7.1 LL 8.3 85 24 345 25 360 1 1D 360 8.45 LL 8.7 97 2 345 7.75 " 8.35 93 3 " 330 8.1 " 8.25 98 4 315 8.8 " 8.45 104 5 300 9.75 " 8.7 112 6 285 10.5 8.6 122 7 " 270 11.2 L 8.75 128 8 I 255 11.65 " 8.8 132 9 " 240 12.2 1 8.75 140 10 " 225 13.2 TT 8.65 155 -41

TABLE 10 (concluded) Wind Speed Distribution Around Solid Tower Case A (B,C). Ref. Coordinates U Flow UO U/Uo No, _ r eO (mph) Type (mph) (%) 11 1D 210 13. 0 TT 8.55 152 12 " 195 13.15 T 8.45 156 13 " 180 7.15 TT 8 8 80 14 " 165 -- L 8,6 15 " 150 - " 8.5 - 16 0 135 22.7 TT 8.5 32 17 " 120 13,15 T 8.6 153 18 " 105 11.8 L 8*7 136 19 090 11.2 " 9.3 121 20 " 075 10.0' 8.8 113 21 060 9.75 " 8.7 112 22 045 8.8 LL 8.45 104 23 " 030 8.10 " 8.25 98 24 015 7.75 8.35 93 25 " 000 8,45 " 8.7 97 1 2D 210 11.8 T 8.8 134 2 " 195 11.7 -- 8.55 136 3 1 180 8.6 TT 8.8 98 4 165 1.5 5 8.55 17 5 " 150 10.9 " 8.6 127 6 " 135 13 9 T 8.9 156 7 120 12.6 L 8.4 150 -42

TABLE 11 Wind Speed Distribution Around Solid Tower Case D (E). Refo Coordinates U Flow Uo U/Uo No. r e (mph) Type (mph) (%) 1 1/2D 360 4.9 L 8.4 58 2 " 345 5,1 LL 8.3 62 3 " 330 9.2 " 8.4 109 4 " 315 6.1 8.3 73 5 300 -8.3 8.35 99 6 " 285 11.2 8.35 134 7 " 270 13 1 T 8.35 157 8 " 255 8.45 TT 8.2 103 9 " 240 13.9 8.55 162 10 225 -. - 8.5 11 210 8.4 12 " 195 -- 8.5 13 180 ~ — 8.6 14 165 -- 8.5 15 150 -- 8.4 16 " 135 -~ 8.5 - 17 120 13.9 TT 8.55 162 18 " 105 8.45 " 8.2 103 19 090 13.1 L 8.35 157 20 1 075 11.2 8.35 134 21 1 060 8.3 8.35 99 22 " 045 6.1 " 8,3 73 23 " 030 9.2 " 8 4 109 24 " 015 5 1 8.3 62 25 1 000 4.9 " 8.4 58 1 1D 360 6.9 LL. 8.4 82:*2 " 345 7.35 3 8.3 89 3 330 9.75 " 8.7 112 4 " 315 803 " 8.35 99 5 1 300 9.2 8.4 109 6 285 11.0 L 8.55 128 7 " 270 12.0 " 8.6 140 8 " 255 13.8 T 8.6 161 9 "t 240 14.4 1 9.25 156 10 " 225 0.8 TT 8.5 9 -43

TABLE 11 (concluded) Wind Speed Distribution Around Solid Tower Case D (E), Ref. Coordinates U Flow U/U No. r e o (mph) Type (mph) (%) 11 1D 210 -' - 8.45 - 12 " 195 -- 8.5 13 " 180 8.6 14 165 - -- 8.5 --- 15 " 150 8.- 8.45 16 " 135 0.8 TT 8.5 10 17 " 120 14.4 T 9.25 156 18 105 13.8 " 8.6 161 19 " 090 12.0 L 8.6 140 20 " 075 11.0 " 8.55 128 21 " 060 9.2 LL 8.4 109 22 " 045 8,3 " 8.35 99 23 " 030 9.75 " 8.7 112 24 " 015 7.5. " 8.3 89 25 " 000 6.9 " 8.4 82 I 2D 240 9.65 T 8.4 115.2 -" 225 13.5 " 8.9 159 3 " 210 10.2 TT 8.6 116 4 " 195 4.2 " 8.5 49 5 " 180 0.9 " 8,7 10 6 " 165 4,2 " 8.5 49 7 " 150 10,2 " 8.6 116 8 " 135 13.5 T 8.9 159 9 " 120 9.65 " 8,4 115 1 4D 195 9,7 TT 8.7 112 2 " 180 5.9 " 8.9 66 3 " 165 9.7 " 8.7 112 -44 —

TABLE 12 Wind Speed Distribution Around Cylindrical Stack. (Horizontal Distribution at height 15") Ref. Coordinates U Flow UO U/UO No. r -. mo (mph) TYRe (mph) (%) 1 1/2D 180 --- TT 8.6 2 -" 165 - " 8.6 3 1 150 0.0 " 8.4 0 4 " 135 7.6 " 8.4 90 5 " 120 11.7 L 8.35 140 6 " 105 11.1 8.3 134 7 " 090 10.2 LL 8.3 123 8 " 075 9.5 " 8.3 115 9 " 060 8.5 " 8.3 102 10 " 045 7.5 " 8.3 90 11 " 030 6.7 " 8.3 81 12 " 015 7.4 " 8.35 89 21 1D -180 TT 8.5 22 " 165 -- " 8.5 23 " 150 8.5' 8,4 101 24 " 135 11.3 L 8.35 135 25 " 120 11,1 LL 8.3 134 26 " 105 10.7 " 8.3 129 27 " 090 9.8 " 8.5 118 28 075 9.1 L 8.3 110 29 " 060 8.5 " 8.3 102 30 " 045 7.9 " 8.3 95 31 " 030 7.5 " 8.3 90 32 " 015 7.4 " 8.35 89 33 " 000 7.5 " 8.4 89 41 2D 180 0,5 TT 8.6 85 42 " 165 5.4 " 8.5 64 43 " 150 9.8 T 8.45 116 44 135 10,1 L 8.4 120 45 " 120 10 0 LL 8.45 118 46 " 105 10.2 8.5 120 47 " 090 9.5 " 85 112 48 " 075 9.1 " 8.4 108 49 " 060 8.7 " 8.35 104 50 " 045 8.5 " 8.3 102 51 " 030 8.1 " 8.3 98 52 " 015 8.0 " 8.35 96 53 " 000 8.0 " 8,4 95 -45

TABLE 12 (concluded) Wind Speed Distribution Around Cylindrical Stack. (Horizontal Distribution at height 15") Ref. Coordinates U Flow U U /Uo No. r e0 (mh) Type ((mph) (ph) (%) 61 3D 180 4.0 TT 8.65 46 62 " 165 7.2 T 8. 5 85 63 150 9.4 L 8.5 111 64 Il 135 9.7 LL 8.55 113 65 120 -— 9.8 " 8.7 113 66 " 105 8.8 " 8.8 116 67 1 090 9.7 1 8.8 110 68 " 075 9.3 " 8.7 107 69 " 060 8.9 " 8.55 104 7 0 045 8.7 " 8.35 104 71 " 030 8.4 " 8.25 102 72 " 015 8.1 " 8.3 98.73 1 000 8.3 " 8.35 99 74 4D 180 5.0 T 8.7 57 75 " 165 8.0 " 8.5 94 76 " 150 9.4 L 8.55 110 77 " 135 9.7 " 8.75 111 78 " 120 9.8 LL 9.0 109 79 1 105 10.4 " 9.1 114 80 " 090 9.8 1 9.1 108 81 " 075 9.6 " 9.05 106 82 " 060 9.0 " 8.9 101 83 " 045 8.9 " 8.5 105 84 " 030 8.5 " 8.2 104 85 " 015 8.1 " 8.2 99 86 " 000 8.3 " 8.3 100 88 6D 180 6.4 T 8.7 74 89 " 165 9.0 " 8.55 105 90 " 150 9.4 LL 8.8 107 91 8D 180 6.5 T 8.7 75 92 " 165 9.5 L 8.6 110 93 l" 1150 9.9 LL 9.2 108 94 1OD 180 6.9 T 8.8 78 -46

TABLE 13 Wind Speed Distribution in a Vertical Plane Behind the Center of a Cylindrical Stack. Refo Coordinates* U/Uo Ref. Coordinates* U/Uo No. x z (%) No. x z (%) 1 1/4D -1/2D 112 29 1/4D 2D 2 1/2D " 107 30 1/2D " 3 1D " 99 31 D " - 4 2D " 97 32 2D " 0 5 4D " 101- 33 3D " 53 34 4D " 73 6 1/4D OD 50 35 5D " 83 7 1/2D " 55 36 6D " 93 8 1D' 76 9 2D " 88 37 1/4D 21/2D --- 10 4D " 95 38 1/2D 39 D " --- 11 1/4D 1/2D 40 2D " 3 12 1/2D " -- 41 3D " 35 13.1D " 32 42 4D " 57 14 2D " 71 43 5D " 76 15 4D " 93 44 6D " 88 16 1/4D 1D -- 45 1/2D 17 1/2D -- 46 1D 18 1D " 3 47 2D " 0 19 2D " 45 48 3D " 35 20 4D: 88 49 4D " 57 50 5D " 76 21 1/4D 11/2D - 51 6D " 76 22 1/2D " 23 1D 24 2D " 33!24:2D33 *x2D " *X boom length" or 25 3D 70 distance downwind 26 4D,' 76 I26 4D " 7611:from stack. 27 5D " 99 z distance down from 28 6D " 10028 6D "I 100 top of stack, in units of D (= diameter of stack) -47

TABLE 14 Summary of Findings from an Analysis of Figures 19 through 21. Boom Operation angle (deg) Extremes of Case Exten- for wind speed Velocity Ratio sion accuracy of: U/U, (%) ___.________ _____ 10%, ~. 5% Maximum Minimum (a) Open Tower, Section I A 1D 325 180 110 68 2D 340 270 109 73 B 1D 325 180 109 63 2D 340 270 109 71 C 1D 340 200 109 81 2D 345 270 109 84 D 1D 240 190 111 70 2D 330 250 109 72 E ID 230 200 112 42 2D 330 260 109 52 Section II A I1D 340 200 109 71 2D 345 240 112 82 E 1D 220 180 111 55 2D 280 240 112 71 (b) Solid Tower A 1D 110 000 159 00 E 1D 000 000 160 00 2D 250' 210 159 10 (c) Stack 1D 000 000 146 00 2D 160 130 120 N5 3D 180 130 116 46 4D 190 160 113 57 -48

OBSERVATION WINDOWS Figure 1. Working section of low speed wind tunnel. 20'. 1 I~~~~~~~~~~~~~~~~~~~n ~~ ~SP~~ljD OUT T ~s.Pr OEDSENS:TOWER` -PLATE..GLASS WALL Figure 1. Working section of low speed wind tunnel,

-502 61 / -s3/4-y41,X\ 6 -i 10 Section 11 31 10 Section I' 1 1/8 _^____________ I I626 10 31 4 ~-~f~ ~ 9 #.:, 1 /4 Figure 2. Scale model of WJBK-TV tower, Detroit, Michigan. (Dimensions are in inches.)

-51CEILING TWIND WIND 15 SECTION A-A A A -5T'4 Nozzle No. I- C \ WIND 2 *4> ^^^. L; ~FLOOR ~ /////x,,./// FROM THE SMOKE GENERATOR Figure 3. Arrangement for smoke tests. (Dimensions are in inches.) WHITE TAPE EDGE v'. — - - 1 3/4 / // A BRASS PIPE 3/4 _ 4 I- 1 -t- 1 -t- 1 - 72 -- Figure 4. Flag arrangement on bar. (Dimensions are in -inches. )

CASE A CASE B CASE C CASE D CASE E (a)Open Tower Sensor Locat ions CASEABC\ CASEA DE (b) Solid Tower (c) Stock Figure 5. Boom and sensor locations.

-53(a) (b) (c) (d) Figure 6. Photographs of tower, stack, anemometers and flags in tunnel. (a) Open tower arranged for wind speed tests. (b) Open tower arranged for wind direction tests; flags at Section II, above floor. (c) Solid tower arranged for wind direction tests. (Note reversal of flags in wake of tower.) (d) Stack (shortened) arranged for wind direction tests. (Note oscillation of flags in wake of stack.)

-545: E1~~~~~~~~~~~~~~~~~~~i i Figure 7. Photograph of Smoke Tests.

-55Figure 8. Photograph of Smoke Tests.

-56Figure 9. Photograph of Smoke Tests. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~id"| _s f w~~~~~ m ^01~~~~~~~~~ D.<: —-~_ ~~-sir miur 9.Poorp fSoeTs

j~~~~~~~~~~~~~~~~~~~~~~~~~~~ji~ Figure 10. Flag tests (wind direction tests) around the open tower.

-58(a) (b) Figure 11. Flag tests (wind direction tests) around the stack. Flags located 21" ahead of zero line in (a), or 7" = -0.5D ahead of stack; 15" ahead of zero line in (b), or 1" = 0.1D ahead of stack; 14" back (c) of zero line = 1D at (c); 28" = 2D at (d); 56" = 4D at (e). (d) A_~~

-591075U1 1075 I05 105. 100 100 105 -60"- 102.5 \Uq/ 102.5 0' 30"60"- W f 90"I j ~~~~~~I Figure 12. Wind speed distribution in open tunnel. (Units of U/Uo are in percent.)

1/11.111' I'1'!1 /0 Iai H N _ l _ 0 I i i I o I I I " 1 I I I I/I II I II I I III I ~ I J~~Ii ~ ~ HD 0 I /.. I/ \ S, I' I I I I UI III II I lHAAlA 1 11111111 I II I 1-I I / \ / cx I 111111111 II A AAAAA II I H 15 111 1 1 / I,-09/ 1rr1 Tr~l111111 A'AA^AA/AAAAII Illl l111 / 1 11111 111 Ill l lA i A/\A A'A I A 1I I I I I I o 1111 II1'11' i'IAAAAAA^I/ I 11111i1l

,/ / / /, i / / / / / / / / /. / / /// / / / / d. *I.~~~~~~~~~~~~ I,, I I I ol _^^^^ \I' t -.ID 0 0.1.2 4D ID 2D 4D 6D \\\\\ \ \ \ \ \ \ \ \ \ \ \ \ \ \' \ \ X ~ \ \ \~ \ Figure 14. Wind direction profile around open tower, Section II.

*//7// /*,-',////////7/7////77/// U~58mph )- -"- +o -- - __ _ - -.ID 0 01 AD 4D ID 2D 4D \\\\ \\ \\\\\~N~~N\N\N\~N~ NN % Nxx \~ \\~ Figure 15. Wind direction profile around open tower, Section II.

- i<.: ~ <~~~~< ~~~- - 0' —^ ^ <~~~~~~ - ~ ~ <Q 0 - 2 1 I 2 Figure 16~~~~~~~~~~~~~. Win diecio prfl rudsldtwr Uc~~ 8mp -;'....... ~0 < X - - ^ ^ >Sb^ ^0 0 0 ^ ^ >~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^ - 1^^ ^ ^ ^1~~~~~~~~~ ~ ^ / ^^^^ ^ ^ ^ ^ ^ ^ $~~~~~~~~~C " 88^^^^ ^ ^^ ^ ^ ^~~~~~~~~ *"* **^ * ~ ^^ ~*^ ^. \ \ \^ Sm — ~~~~~~~~~~~~~~~~~~~~~^ - - 4' )~ p', - ^ >^" -- 4') 0' /{5 ~ ^ ^ ^%^^^^ ^^ <) 0 0 ^ ^ (^~~~~~ r^^~- I -04 -O I D 00.I.2.4 ID 2D41 A ANGLE OF FLUCTUATION 0 ROTATING VANE Figure 16. Wind direction profile around solid tower.

/ / / / / / / / / / / /' / / / // /// /'/ / / / / / / / / / /I, U m h - D - -. _ - _ - _~ - _ ~~~~~~~stack.~~.. UokSmph..... D - -' C.< A ANGLE OF FLUCTUATION ~ ROTATING VANE Figure 17. Wind direction profile around cylindrical stack.

-65Q D R R R I_ I (a) Stack (b) Cases D aE (c) Cases A,B C Figure 18. Coordinate systems for anemometer (wind speed) tests.

-66U/Uo(%) 1500 120 210 -9~0^ Q ^22700 Sectionl 00 0~2 \0o ~ RID (a, a distance R from open tower structure, for winds from 360 of arc.7 Dirn Section I 0~ R= ID (a) Figure 19, Wind speed profile at sensor, Q, located a distance R from open tower structure, for winds from 3600 of arc.

-67U/Uo(%) 1500 120 2100 210 \1200 2400 B^ ^-60 ~/ OR~ R 90~0 Q 2700 -R4D /3 0o \ Section I (b) Figure 19. (Continued)

-684: 150~ 120 2100 6~100\ / 120" 24 4 / Case B 30 30 -"- R=/2D,Section I 00 \ —-R=ID (C) Figure 19. (Continued) Figure 19. (Continued)

-69U/Uo(%) 1500 120 2100 1002400 120~ 240 60 900 Q1 270~\ Wind 760r'\ d^J /A00 Case B o R=2D Section I / 0 \ —- R=4D (d) Figure 19. (Continued)

-70U/Uo(%) 1500 120 210~ 10 ~120o ___ / <, 240~-/' ~60^'n Dr-60 Figure 19. (Continued) Section, 0O R I (e) Figure 19. (Continued)

-71U/Uo(%) 150 120- 2100 1200' 240~/ - R 90~ Q0 1 270 Wind / o00\ Dirtm 00^ Case C —- R=2D Section I / \ -02D (f) Figure 19. (Continued)

-72U/Uo(%) 150 ~ 120 2100 120 ~240 900<ts 60 R ~900 270o-, Wind 60o 0o14 Case D 300 ---- -/R=1/2D Section I / - \Io R=ID (g) Figure 19. (Continued)

-73U/Uo(%) / )C~Z~-rC- 1.00 156 0 120 210r Figure 19. (Continued) 60 I /A^ \ R 90" Q0 270 Windll 60\ 0, Case D ~30 R=2D Section I / o \ — o —R=4D (h) Figure 19. (Continued)

-74U/Uo(%) 1 50 120 210~ \1200 \\ I 80 1200 / T _, 0 / Win!(i) igure 19. Continued) 6. D\ ir/n.8-..., Case E -R= 1/2D 306 0330 Section I / R= ID (i) Figure 19. (Continued)

-75U/Uo(%) 150 120 210 10~ 80 1200 / - \ 240~ 90" Q 1 2700\ Wind / 600" Dir'n A00"Case E 30300~ R=2 Section I O/ " \ -o-R=4D (j) Figure 19. Continued

-7 6~~~8150~ \^ U/Uo(%) / ~ 50 120- 210 "-120" "\ 240"^^ 1200~300 Figure 19. (Continued) - Figure 19.w (Continued)

-77U/Uo(%) 150~ 120 2100 \120~ / 240~ 602.90 Q 270Wind 60\ jC00 Case A /30330\ -o-R=2D Section / D \ (1) Figure 19. (Continued)

-78U/Uo(%) 150, o120 2100 10o.0 9Q0 120 / / 240/ Wind 600 Dirn Case E --- R=1/2D SectionIT / 0 \ R=ID (m) Figure 19. (Continued)

-79U/Uo(%) 1500 120 2100 100 —-- 1200 / \ 2400 / \ 60 / A \ -90~ Q - -I 270 Wind 600 Dir 00 Case E 30o3o - R=2[ Section / 0 \ — R=4D (n) Figure 19. (Concluded)

-80U/uo(%)'~~~ / 2 10^ \ 150~ T V^~~~ 20- / ^1,.^^.. r.-I0 \,, (_)' -t I~~~~~~~~~p.~~~~~~~~~~.0 R 2 90,I \ W.ind1 ^^ ^ ^^'.. No Obser.,-'-.a- - R: 1/2 O Case A 0 ~ -.-o — R= I D (a) Figure 20. Same as Figure 19, but for solid tower.

-81U/Uo(%) 1500 1^__20 100 2 51200 / 80= \ 2400 90 l Q1 270 Wind 600\ Dir /-000. Case A /3030\ R 2D / 1 \- (b) Figure 20 (Continued) Figure 20. (Continued)

-82U/Uo(%) 1500 12210'100 N,. 6 0/, -r -H ) (c) Figure 20. (Continued) Figure 20. (Continued)

-83U/Uo(%) 150 120 _ 2100 9210 C90ae QE 0 270 i\ / \\ I I1! Wind 960~ /'rQJ j20o %3 0c:0 o R=2D Case E / \ —o — R =4D (d) Figure 20. (Concluded)

-84\U/Uo(%) 150~ 120 210,,_1 00'^\ \20~ \\8 0 42R0~40 /'~ Wind d 9,00 ^^300^^^ s'-oc, o~ *!-'f 30 -- ~330~\ — a R-1/20 Stack 0/ 0" \o R=ID (a) Figure 21. Same as Figure 19, but for stack.

-85\ U/Uo( %)/ r/d 90bV' Q10 - 6Fi 1.0 ( 21300C 120s-30 ~o~R=2D Stacko / 0 ^.^-R 3D -':~"~ -- -/ -~(b ) Figure 21. (Continued)

-86U/Uo(%) / 150~ 210~ ~3 120\120' o\ 2400o ~900 Q 70-27 Wind 60\ Dir/.n 00 300 3300 R=4D Stack /\ — e- R-R=6D (c) Figure 21. (Concluded)

ID UX. 8mph 2O 3D " "0 -25 50 75 100 U/U (/) 025 5N ID NN 3N 4N 5N 6DN N 025.5D ID 2D 3D 4D 5D 6D" Figure 22. Wind speed profiles in a vertical plane behind the center of a cylindrical stack.

-88/ A As 3300 (b') ~,~ ~_~ ~~~~~0(C) - t 1, 240 I^ /^^^^h^^~^r ^l /2 D 2200 A: Speed true within + 10 % B: Speed true within 5 % Direction true within + 100 Direction true within + 50 C: Arrow points: 1. toward wind direction of minimum concern, or 2. toward wind direction of lowest frequency. Figure 23. Recommendations concerning mounting of wind sensors on a relatively open tower ("shadow densities" not exceeding 40 percent).

-89A e 1600 (0a) -2R D.( A 1 80\ or B ~~1300 R=$0 O3D A: Speed true within-+ I0% B' Speed true within + 5% Direction true within ~ 5~ Direction true within +_ 5~ C' Arrow points' I. toward wind direction of minimum concern, 2. toward wind direction of lowest frequency. Figure 24. Recommendations concerning mounting of wina sensors on circular stacks.

901 NV12I THE UNIVERSITY OF MICHIGAN GRADUATE LIBRARY DATE DUE