     The data files on this tape contain reduced mean flow data and 
Reynolds averaged data from Amy Alving's experimental investigation of
boundary layer relaxation from convex curvature.  The investigation was
performed at Prceton University's Gasdynamics Laboratory.
     All measurements were made in a subsonic, open-return wind tunnel
driven by a suction fan located downstream of the test section.  The 
flow entered the tunnel through a bell-mouth, passed through a honeycomb
flow-straightener and into a settling chamber containing a series of five
screens placed perpendicular to the flow.  The flow exited the settling
chamber, passed through a two-dimensional 6:1 contraction and entered
the 0.15m x 1.22m test section.  At the contraction exit, a 1.0mm trip
wire provided a uniform transition site for the boundary layer on all 
four walls.  The boundary layer on the test wall developed on a flat
plate in zero pressure gradient, with a nominal freestream velocity
of 31m/s and a freestream turbulence intensity of 0.3%.  At a distance 
of 1.5m downstream of the trip wire, the boundary layer thickness was 
22.7mm, and the momentum thickness Reynolds number was approximately 6000.
     At this point, the test wall was subjected to 90 degrees of convex
curvature with a constant radius of 300mm.  In an attempt to isolate 
curvature effects from pressure gradient effects, the wall opposite the 
test wall was contoured to minimize the imposed pressure gradient on the 
test wall, and the boundary layer on this wall was removed by suction.
In addition, sidewall jets were used to minimize the strength of the 
secondary flow.
     After passing through the curved section, the boundary layer was
allowed to relax on a flat plate in zero pressure gradient.  This 
recovery region was 4.9m long.  At the end of the recovery region,
the flow passed through a diffuser, through the fan and a series of 
baffles, and exited into the room.  
     Several ports were located along the length of the test wall to 
provide access for measurement probes.  The port centers were located
as follows:


      Port #         s (distance in m from the end of curvature)

         1        -0.646     (Upstream of the entrance of the
                                curved section)
         2         0.098     (Just downstream of the exit from 
                                the curved section)
         3         0.238
         4         0.379
         5         0.529
         6         0.799
         7         1.099
         8         1.600
         9         2.100
        10         2.600
        11         3.200
        12         3.830
        13         4.404

     Only ports 1 through 9 were actually used, due to the inability 
to obtain a zero pressure gradient beyond port 9.  A series of static 
pressure taps along the centerline of the test wall was used to measure 
the pressure coefficient along the length of the test wall.  All static 
pressure measurements were referenced to a tap located on the test wall 
0.36m downstream of the trip wire.  A pitot-static tube located 0.86m
downstream of the trip wire was used to measure the reference freestream 
velocity.
     Using a pitot probe, mean velocity profiles were measured at ports 
1 through 9.  These profiles were used to calculate all integral 
parameters of the boundary layer.  The skin friction coefficient was
determined by fitting the velocity profiles to the law-of-the-wall.
     Measurements of the instantaneous velocity were made at ports 1 and 
3 through 9.  These mesurements were made using constant-temperature
hot-wire anemometry.  A single normal wire, placed perpendicular to the
flow and parallel to the test wall, was used to measure the instantaneous 
streamwise velocity (u).  A pair of crossed-wires oriented perpendicular 
to both the test wall and the mean flow was used to measure instantaneous 
streamwise (u) and normal (v) velocities simultaneously.  A pair of 
crossed-wires oriented parallel to the test wall and perpendicular to 
the mean flow was used to measure instantaneous streamwise (u) and 
spanwise (w) velocities simultaneously.  All wires were calibrated using
a dynamic calibration technique described in Alving (1988).
     For those who desire more detatiled information, the investigation
is thoroughly documented in the following publications:

     1)  Alving, A.E.  Ph.D. thesis, Princeton University, 1988.
     
     2)  Alving, A.E., Smits, A.J., & Watmuff, J.H.  1990,
             "Turbulent Boundary Layer Relaxation from Convex 
              Curvature,"  J. Fluid Mech. 211, 529-556.

     3)  Smits, A.J., et al.  1989,  "A Comparison of the Turbulence 
              Structure of Subsonic and Supersonic Boundary Layers,"
              Phys. Fluids A 1 (11), 1865-1875.         
     

     The contents of the files on this tape are as follows:
     
     Filename:        Description of contents:

     INFO.DAT         Information about where each survey was made,
                      the pressure coefficient and skin friction
                      coefficient, reference velocity and friction
                      velocity for each survey.           

     CP1.DAT          Pressure coefficient measured at static pressure 
                      taps along the centerline of the test wall.
     
     CP2.DAT          Pressure coefficient measured at the center of
                      each port.

     P1MEAN.DAT       Pitot survey at Port 1
     P2MEAN.DAT       Pitot survey at Port 2
     P3MEAN.DAT       Pitot survey at Port 3
     P4MEAN.DAT       Pitot survey at Port 4
     P5MEAN.DAT       Pitot survey at Port 5
     P6MEAN.DAT       Pitot survey at Port 6
     P7MEAN.DAT       Pitot survey at Port 7
     P8MEAN.DAT       Pitot survey at Port 8
     P9MEAN.DAT       Pitot survey at Port 9

     P1U.DAT          Single normal wire hotwire data at Port 1
     P1UV.DAT         u and v crossed wire data at Port 1
     P1UW.DAT         u and w crossed wire data at Port 1

     P3U.DAT          Single normal wire hotwire data at Port 3
     P3UV.DAT         u and v crossed wire data at Port 3
     P3UW.DAT         u and w crossed wire data at Port 3
    
     P4U.DAT          Single normal wire hotwire data at Port 4
     P4UV.DAT         u and v crossed wire data at Port 4
     P4UW.DAT         u and w crossed wire data at Port 4

     P5U.DAT          Single normal wire hotwire data at Port 5
     P5UV.DAT         u and v crossed wire data at Port 5
     P5UW.DAT         u and w crossed wire data at Port 5

     P6U.DAT          Single normal wire hotwire data at Port 6
     P6UV.DAT         u and v crossed wire data at Port 6
     P6UW.DAT         u and w crossed wire data at Port 6

     P7U.DAT          Single normal wire hotwire data at Port 7
     P7UV.DAT         u and v crossed wire data at Port 7
     P7UW.DAT         u and w crossed wire data at Port 7

     P8U.DAT          Single normal wire hotwire data at Port 8
     P8UV.DAT         u and v crossed wire data at Port 8
     P8UW.DAT         u and w crossed wire data at Port 8

     P9U.DAT          Single normal wire hotwire data at Port 9
     P9UV.DAT         u and v crossed wire data at Port 9
     P9UW.DAT         u and w crossed wire data at Port 9

     RDCP.FOR         
     RDMEAN.FOR       These four files are described below.
     RDHW.FOR
     RDINFO.FOR

     
     The file INFO.DAT is composed of six columns of data.  The first
column is a filename corresponding to one of the other data files 
listed above (excluding CP1.DAT and CP2.DAT).  The second column
is the streamwise position, s,  of the probe tip for each data survey,
measured in meters, relative to the end of curvature.  A negative value
is a position upstream from the end of curvature, a positive value is
downstream from the end of curvature.  The third column contains the
static pressure coefficient, Cp,  interpolated from the data in CP1.DAT 
and CP2.DAT.  To perform the interpolation, a linear variation in Cp
between adjacent points was assumed.  The fourth column contains
the skin friction coefficient, Cf, which was calculated by fitting the
velocity profile data from P1MEAN.DAT, P2MEAN.DAT, etc... to the
logarithmic portion of the law of the wall.  The values of s and Cp are
the same for each of the four data surveys at each port.  Likewise, the
value of Cf is assumed to be the same for each of the four surveys at 
each port.  (A blank entry in INFO.DAT means that the value is
the same as the first non-blank entry above it.)  The fifth column
contains the value of the reference velocity, Uref, as measured by the 
upstream pitot-static tube, for each survey.  The sixth column contains
the value of the friction velocity, Utau, for each survey.  The 
procedure used to calculate Utau is as follows.  For the pitot probe
surveys (P*MEAN.DAT files), the local freestream velocity , Ue, is 
calculated from the velocity profile data, as is the skin friction
coefficient (as described above).  The value of Utau for these surveys
was calculated by multiplying the freestream velocity by the squareroot
of one-half the skin friction coefficient.  For the three hot-wire
surveys at each port, the value of Utau was calculated by the value of
Utau from the pitot survey at that port multiplied by the reference 
velocity from the particular hot-wire survey and divided by the 
freestream velocity from the pitot survey.  Thus,

        (Utau)pitot = (Ue)pitot * sqrt(Cf/2)

        (Utau)hotwire = (Utau)pitot * (Uref)hotwire / (Uref)pitot



     The static pressure data files (CP1.DAT, CP2.DAT) each have a 
header at the beginning of the file, which contains brief descriptive 
information about the data, and the values of both the reference 
dynamic head and the average temperature during the survey.  The 
header is followed by four columns of numerical data.  The first 
column is simply an index.  In CP1.DAT this corresponds to the number 
of the static pressure tap.  In CP2.DAT the index corresponds to the 
port number.  The second column contains the streamwise position of 
the static tap, in meters, referenced to the end of curvature.  As
stated before, a positive position is downstream of the end of curvature, 
while a negative position corresponds to a location upstream of the end of 
curvature.  Note that no measurements were made in the curved section 
of the tunnel.  The third column contains the measured pressure 
difference , in Pa, between the static tap at the position indicated 
and the reference static tap.  The fourth column contains the calculated 
pressure coefficient.
     The mean data files from pitot surveys (P1MEAN.DAT, etc.) also 
have a header at the beginning of the file.  The header contains 
descriptive information about the data, the pitot probe outside 
diameter, and all pertinent measured and calculated parameters of 
the freestream flow and the boundary layer.  The header is followed 
by 7 columns of numerical data.  The first column is an index.  The 
second column is the height of the probe, in mm, above the test wall.  
This height has been corrected to account for the inside and outside 
diameters of the pitot probe.  The third column contains the measured 
value of U (average velocity in the freestream direction) normalized 
by the local freestream velocity.  The fourth column contains the 
nondimensional quantity formed by the product of the local freestream 
velocity and the height above the wall divided by the kinematic viscosity.  
The fifth column contains the nondimensional height from the wall in 
wall units, Y+.  The sixth column contains the measured velocity 
normalized by the friction velocity, U+.  The seventh column contains 
the calculated deviation of the measured value of U+ from the value 
predicted by the law of the wall.  As mentioned earlier, the value 
of the skin friction coefficient is determined by fitting the measured 
velocity profile to the law of the wall.  The friction velocity is 
computed from the skin friction coefficient and the local freestream
velocity, as described above.
     The hotwire data files (P1U.DAT, etc.) begin with a file header, 
which contains the friction velocity (determined from the formulas
given above) and boundary layer thickness (calculated from the pitot
survey data), the index of the first line of data in the file, the 
index of the last line of data in the file, the number of wires, and 
the number of probes.  (Thus the single normal wire data will indicate 
1 wire and 1 probe, while the crossed-wire data will indicate 2 wires 
and 1 probe.)  
     In the normal wire data files, the header is followed by 6 columns
of numerical data.  The first column is the index of the data point.  
The second column is the height, in mm, above the test wall, of the 
center of the active length of the wire.  The third column is the 
calculated average streamwise velocity (U), in m/s. The fourth, fifth, 
and sixth columns are the calculated variance, skewness and kurtosis, 
respectively, of the fluctuating component of the streamwise velocity (u').
     In the P*UV.DAT crossed-wire data files, the header is followed by 12 
columns of numerical data.  The first column is the index of the data 
point.  The second column is the height, in mm, above the test wall, of 
the center of the active lenghts of the two wires.  The third column is
the calculated average streamwise velocity (U), in m/s.  The fourth column 
is the calculated average velocity (V), in m/s, normal to the wall and 
perpendicular to the freestream velocity.  Columns 5, 8 and 12, contain, 
respectively, the variance, skewness and kurtosis of the fluctuating 
component of velocity in the freestream direction (u').  Columns 7 and 11 
contain, respectively, the variance and skewness of the fluctuating 
component of V (v').  The sixth column contains the average value of 
the product u'v'.  Columns 9 and 10 contain, respectively, the average
values of the products (u'**2)(v') and (u')(v'**2). 
      The structure of the P*UW.DAT files is exactly the same as that of
the P*UV.DAT files.  Simply replace V by W and v' by w'.  In this case,
W is the velocity parallel to the test wall and normal to the freestream
flow direction.  w' is the fluctuating component of W.
     The four files RDCP.FOR, RDMEAN.FOR, RDHW.FOR, and RDINFO.FOR contain 
FORTRAN program listings of code that can be compiled and linked and used 
to read in data from each of the three types of data files on this tape.  
Use RDCP.FOR to read in data from the files containing static pressure data.
Use RDMEAN.FOR to read in data from the files containing mean flow data from 
pitot surveys.  Use RDHW.FOR to read in data from the files containing hotwire
data.  Use RDINFO.FOR to read in the data in INFO.DAT.

     If there are any questions about the contents of this tape or the
experimental investigation described above, please contact:

                      Prof. A.J. Smits
                      Dept. of Mech. and Aerospace Engineering
                      D302 E-Quad
                      Olden Street
                      Princeton University
                      Princeton, NJ  08544

                      Phone:  (609) 258-5117
