Langley Research Center

Turbulence Modeling Resource

Exp: CFDVAL2004 Case 1 Details and Submission Guidelines

Return to: CFDVAL2004 Case 1 - Intro Page

Return to: CFDVAL2004 - Intro Page

Return to: Data from Experiments - Intro Page

Return to: Turbulence Modeling Resource Home Page

Relevant details for Case 1 are as follows:

Flow issues into an enclosed box (24 inches on each side).

M_freestream = 0

The atmospheric conditions varied, but were essentially standard atmospheric conditions at sea level, in a temperature-controlled room. These conditions can be given as approximately:

p_ambient = approx 101325 kg/(m-s^2)

T_ambient inside box = approx 75 deg F (approx 297 K)

T inside cavity = approx 83 deg F (approx 301 K)

Some derived relevant conditions are:

density_ambient = approx 1.185 kg/m^3

viscosity_ambient = approx 18.4e-6 kg/(m-s)

The diaphragm frequency = 444.7 Hz. The 2.0 inch-diameter diaphragm is circular, and is held in place with an o-ring seal 1.5 inches in diameter. The displacement (D) at the center of the diaphragm as a function of phase is given in the following file:

BCs from PIV experiment, updated 22 August 2003

Note that the displacement at the center of the diaphragm is offset - it displaces inward LESS and outward MORE (i.e., the reference position is not zero) Also in the file, the pressure (P) and temperature (T) as a function of phase are given. The location where these quantities are measured are as follows. The pressure is measured inside the cavity on the wall opposite the center of the diaphragm (the pressure is given with respect to an unspecified reference pressure, which may be presumed to be approximately atmospheric). The temperature is measured at the bottom (floor) of the cavity. (The voltage input to the diaphragm, prior to being amplified, is also included in the file, although this is not relevant information for CFD. It has been offset to positive values for LDV purposes.)

The following image shows a view of the cavity assembly from the underside. The diaphragm is on the right side (next to the displacement gauge). The image shows the location of the pressure gauge (mounted onto the side wall of the chamber opposite the center of the diaphragm), as well as the location of the thermocouple (at the bottom wall of the chamber).

photo of syntheic jet cavity assembly

The following figure shows the measured phase-averaged vertical velocity over the center of the slot as a function of phase, using three different measurement techniques.

plot of centerline velocity as function of phase

These data are given in the following files:

Submission Guidelines:

(Last updated: 16 October 2003)

Numerical predictions of this type of statistically unsteady flow are relatively new. The purpose here is to determine the state-of-the-art in modeling these types of unsteady synthetic-jet-type flows, so we want to explore which CFD methods work and which do not.

You can either model the internal cavity or specify an unsteady boundary condition at the slot exit.
You may choose to model the moving diaphragm, or you may employ an unsteady boundary condition that approximates its effect within the cavity.
Even though the geometry (and flowfield) is inherently three-dimensional, you may choose to model it two-dimensionally (across the center of the slot).

There is the requirement that you detail specifically how you choose to model the case, including all boundary conditions and approximations made. As we assess the methodologies used at the workshop, it will be important to know as many details as possible about the calculations/simulations.

Detailed requirements include:

1. The case must be run time-accurately, in order to simulate the unsteady nature of the case.

1a. RANS codes run in time-accurate mode (e.g., URANS) solve directly for phase-averaged variables, i.e. <u_i> and <u'_iu'_j>. Click on link to PDF write-up for more details. Therefore, RANS simulations should result in repeating or very-nearly-repeating periodic solutions. When periodicity is achieved, averages over one or more periods of oscillation yields the long-time-averaged (time-independent mean) values for these quantities.
1b. DNS, LES, or DES simulations will need to be post-processed to obtain both the phase-averaged and long-time-averaged (time-independent mean) values.

2. GRID STUDY: If you choose to model this case two-dimensionally, then it is strongly suggested that you perform the computation using at least two different grid sizes in order to assess the effect of spatial discretization error on the solution. If you model it three-dimensionally, then solutions using more than one grid size are encouraged, but not required. If you use more than one grid, submit each set of results separately.

3. TIME STEP STUDY: If you choose to model this case two-dimensionally, then it is strongly suggested that you perform the computation using at least two different time step sizes in order to assess the effect of temporal discretization error on the solution. If you model it three-dimensionally, then solutions using more than one time step are encouraged, but not required. If you use more than one time step, submit each set of results separately.

Specific quantities that result from your computations at particular locations will be required for submission. Note that for all the following, we adopt the coordinate system with x across the slot (across the 0.05 inch-wide gap) and y up, with the (x,y)=(0,0) origin at the bottom wall of the "box" into which the jet issues, at the center of the slot opening. All results from 3-D computations are to be given at the z=0 location (across the center of the slot). The requirements follow (if you are unable to provide a particular quantity, simply leave it out of the "variable" list, and reduce the number of columns of data submitted):

a. Long-time-averaged jet width (width is in the x-direction)
as a function of vertical height above
the lower wall, from the wall up to a height of at least 20 mm.
Define jet width as the horizontal distance
between which the vertical velocity exceeds its (maximum+minimum)/2
value (at that vertical height).
   Name this file: case1.avgjetwidth.ANYTHING.dat
    -where "ANYTHING" can be any descriptor you choose (should
     be different for each file if you are submitting multiple
     runs)
    -the file should be in 2-column format:
      1st line: #your name (pound sign needed)
      2nd line: #your affiliation (pound sign needed)
      3rd line: #your contact info (pound sign needed)
      4th line: #brief description of grid (pound sign needed)
      5th line: #number of time steps per cycle (pound sign needed)
      6th line: #brief description of code/method (pound sign needed)
      7th line: #other info about the case, such as spatial accuracy (pound sign needed)
      8th line: #other info about the case, such as turb model (pound sign needed)
      9th line: #other info about the case (pound sign needed)
     10th line: variables="y, mm","avg jet width, mm"
     11th line: zone t="jet width"
     subsequent lines:  y(mm)  avgjetw(mm)  <- this is the data

Download sample file case1.avgjetwidth.SAMPLE.dat

plot showing determination of jet width

b. Long-time-averaged horizontal velocity (u) and vertical velocity (v) profiles along
the center of the jet (at the center of the slot), from
the wall up to a height of at least 20 mm.  Also, submit
profiles along nine lines parallel to the wall,
at heights of 0.1 mm, 1 mm, 2 mm, 3 mm, 4 mm, 5 mm,
6 mm, 7 mm, and 8 mm above the lower
wall, from at least -10 mm to +10 mm to either side
of the jet centerline.
   Name this file: case1.avgvel.ANYTHING.dat
    -where "ANYTHING" can be any descriptor you choose (should
     be different for each file if you are submitting multiple
     runs)
    -the file should be in 4-column format:
      1st line: #your name (pound sign needed)
      2nd line: #your affiliation (pound sign needed)
      3rd line: #your contact info (pound sign needed)
      4th line: #brief description of grid (pound sign needed)
      5th line: #number of time steps per cycle (pound sign needed)
      6th line: #brief description of code/method (pound sign needed)
      7th line: #other info about the case, such as spatial accuracy (pound sign needed)
      8th line: #other info about the case, such as turb model (pound sign needed)
      9th line: #other info about the case (pound sign needed)
     10th line: variables="x, mm","y, mm","u, m/s","v, m/s"
     11th line: zone t="centerline"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)  <- this is the data along x=centerline
     next line: zone t="y=0.1 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)  <- this is the data along y=0.1mm
     next line: zone t="y=1 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)  <- this is the data along y=1mm
     next line: zone t="y=2 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)  <- this is the data along y=2mm
     next line: zone t="y=3 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)  <- this is the data along y=3mm
     next line: zone t="y=4 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)  <- this is the data along y=4mm
     next line: zone t="y=5 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)  <- this is the data along y=5mm
     next line: zone t="y=6 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)  <- this is the data along y=6mm
     next line: zone t="y=7 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)  <- this is the data along y=7mm
     next line: zone t="y=8 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)  <- this is the data along y=8mm

Download sample file case1.avgvel.SAMPLE.dat

c. Phase-averaged quantities at 8 different phases during
the cycle:  0 deg, 45 deg, 90 deg, 135 deg, 180 deg,
225 deg, 270 deg, 315 deg.; where you should align the phases of
your computation as described below.  Submit the following
phase-averaged <> quantities:  u (m/s), v (m/s), u'u'bar (m^2/s^2),
v'v'bar (m^2/s^2), and u'v'bar (m^2/s^2), where:
   u = phase-averaged horizontal velocity component (parallel to lower wall, across the slot)
   v = phase-averaged vertical velocity component (up, away from lower wall)
   u'u'bar = phase-averaged turbulent normal stress in horizontal direction (optional)
   v'v'bar = phase-averaged turbulent normal stress in vertical direction (optional)
   u'v'bar = phase-averaged turbulent shear stress in x-y plane
The locations for these data are the same as for the long-time-averaged
quantities.
   Name these files: case1.phase000.ANYTHING.dat
                     case1.phase045.ANYTHING.dat
                     case1.phase090.ANYTHING.dat
                     case1.phase135.ANYTHING.dat
                     case1.phase180.ANYTHING.dat
                     case1.phase225.ANYTHING.dat
                     case1.phase270.ANYTHING.dat
                     case1.phase315.ANYTHING.dat
    -where "ANYTHING" can be any descriptor you choose (should
     be different for each file if you are submitting multiple
     runs)
    -the file should be in 7-column format:
      1st line: #your name (pound sign needed)
      2nd line: #your affiliation (pound sign needed)
      3rd line: #your contact info (pound sign needed)
      4th line: #brief description of grid (pound sign needed)
      5th line: #number of time steps per cycle (pound sign needed)
      6th line: #brief description of code/method (pound sign needed)
      7th line: #other info about the case, such as spatial accuracy (pound sign needed)
      8th line: #other info about the case, such as turb model (pound sign needed)
      9th line: #other info about the case (pound sign needed)
     10th line: variables="x, mm","y, mm","u, m/s","v, m/s",
        "uu, m^2/s^2","vv, m^2/s^2","uv, m^2/s^2"
     11th line: zone t="centerline"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)
        uu(m^2/s^2)  vv(m^2/s^2)  uv(m^2/s^2)  <- this is the data along x=centerline
     next line: zone t="y=0.1 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)
        uu(m^2/s^2)  vv(m^2/s^2)  uv(m^2/s^2)  <- this is the data along y=0.1mm
     next line: zone t="y=1 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)
        uu(m^2/s^2)  vv(m^2/s^2)  uv(m^2/s^2)  <- this is the data along y=1mm
     next line: zone t="y=2 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)
        uu(m^2/s^2)  vv(m^2/s^2)  uv(m^2/s^2)  <- this is the data along y=2mm
     next line: zone t="y=3 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)
        uu(m^2/s^2)  vv(m^2/s^2)  uv(m^2/s^2)  <- this is the data along y=3mm
     next line: zone t="y=4 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)
        uu(m^2/s^2)  vv(m^2/s^2)  uv(m^2/s^2)  <- this is the data along y=4mm
     next line: zone t="y=5 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)
        uu(m^2/s^2)  vv(m^2/s^2)  uv(m^2/s^2)  <- this is the data along y=5mm
     next line: zone t="y=6 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)
        uu(m^2/s^2)  vv(m^2/s^2)  uv(m^2/s^2)  <- this is the data along y=6mm
     next line: zone t="y=7 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)
        uu(m^2/s^2)  vv(m^2/s^2)  uv(m^2/s^2)  <- this is the data along y=7mm
     next line: zone t="y=8 mm"
     subsequent lines:  x(mm)  y(mm)  u(m/s)  v(m/s)
        uu(m^2/s^2)  vv(m^2/s^2)  uv(m^2/s^2)  <- this is the data along y=8mm

Download sample file case1.phase000.SAMPLE.dat

d.  Phase-averaged time-history values of <u> and <v> as a
function of phase (deg) at three approximate point locations:
(x,y)=(0mm,0.1mm), (0mm,2mm), and (1mm,2mm).  Give the data at
every time step taken.  In other words, if your time step yields
100 steps per cycle, then give 100 phases between 0 deg and 360 deg.
You should align the phases of your computation as described below.
   Name this file: case1.phasehist.ANYTHING.dat
    -where "ANYTHING" can be any descriptor you choose (should
     be different for each file if you are submitting multiple
     runs)
    -the file should be in 5-column format:
      1st line: #your name (pound sign needed)
      2nd line: #your affiliation (pound sign needed)
      3rd line: #your contact info (pound sign needed)
      4th line: #brief description of grid (pound sign needed)
      5th line: #number of time steps per cycle (pound sign needed)
      6th line: #brief description of code/method (pound sign needed)
      7th line: #other info about the case, such as spatial accuracy (pound sign needed)
      8th line: #other info about the case, such as turb model (pound sign needed)
      9th line: #other info about the case (pound sign needed)
     10th line: variables="phase, deg","x, mm","y, mm","u, m/s","v, m/s"
     11th line: zone t="x=0 mm, y=0.1 mm"
     subsequent lines:  phase  x(mm)  y(mm)  u(m/s)  v(m/s)  <- this is the data at x=0mm, y=0.1mm
     next line: zone t="x=0 mm, y=2 mm"
     subsequent lines:  phase  x(mm)  y(mm)  u(m/s)  v(m/s)  <- this is the data at x=0mm, y=2mm
     next line: zone t="x=1 mm, y=2 mm"
     subsequent lines:  phase  x(mm)  y(mm)  u(m/s)  v(m/s)  <- this is the data at x=1mm, y=2mm

Download sample file case1.phasehist.SAMPLE.dat

e.  Field line-contour-plots (in one of the following formats: ps, eps, or jpg)
of phase-averaged <u>-velocity and <v>-velocity at the following phases:
45 deg, 90 deg, 135 deg, and 225 deg.; where you should align the phases of
your computation as described below.  These
plots should be black-and-white line plots only.
The plots should go from approx x=-3mm to 3 mm, and y=0mm
to y=8mm.  The x-to-y ratio of the plot should be 1.0.  For
all plot files, plot line contours of -20 m/s through 30 m/s in increments of
1 m/s.  Either (a) plot the lines of negative velocity as dashed lines
and the lines of positive velocity as solid lines, or (b) label the
contour lines, or (c) do both.  The purpose of submitting these plots is to get
a qualitative picture of the phase-averaged flowfield at
particular selected times of interest.  Altogether, submit 8 plots files.
   Name these files: case1.phase045.u.ANYTHING.eps
                     case1.phase090.u.ANYTHING.eps
                     case1.phase135.u.ANYTHING.eps
                     case1.phase225.u.ANYTHING.eps
                     case1.phase045.v.ANYTHING.eps
                     case1.phase090.v.ANYTHING.eps
                     case1.phase135.v.ANYTHING.eps
                     case1.phase225.v.ANYTHING.eps
(where the "eps" in this case means encapsulated postscript - use ps,
or jpg instead if appropriate).

Definition of Phase for the Computations, updated 20 October 2003

Matching the same phases with the experiment is not necessarily straightforward. One way to do it is to try to align a quantity from experiment (such as vertical velocity near the jet exit), but this can be imprecise because the CFD and experimental data are not necessarily well-behaved sine-waves. The best criteria for determining phase may in fact be a different measure altogether.

On the other hand, in this workshop we want to be able to compare CFD results with each other, so it is important to try to achieve the same phase definitions in order that all computations are similarly aligned. We have tried to choose criteria for determining phase that approximates experiment AND is specific enough so that different CFD solutions can be meaningfully compared.

Therefore, although participants are given some latitude to determine phase as appropriate, we encourage everyone to use the following steps to define phase in a uniform fashion for Case 1:

Step 1. Output vertical phase-averaged velocity (v) at the following point in space (over the slot) as a function of your iteration or time step number: (x,y,z)=(0, 0.1, 0)mm. Find vmax and vmin over the course of one phase-averaged period.
Step 2. Compute the mid-value vavg=(vmax+vmin)/2
Step 3. Define Phase=340 deg as the time when your velocity at this location approximately equals vavg (INCREASING). All other phases can be referenced from this, via the following relation:

Phase=(iter-it340)*360/nstep+340

where:

iter =iteration (or time step) number

nstep=no. of time steps per cycle

it340=iteration number when Phase=340 according to the above criteria

For example, if you are running 360 steps per cycle and you match the above criteria at time step number 5575, then

Phase=iter-5235

Thus, when iter=5235, then Phase=0; when iter=5415, then Phase=180; when iter=5595, then Phase=360. Note that Phase=360 also corresponds with Phase=0 (it repeats every 360 deg). This is illustrated in the following figure:

plot showing how to determine phase

As another example, say you are running 1080 steps per cycle and you match the above criteria at time step number 10002, then

Phase=(iter-10002)/3+340

Thus, when iter=8982, then Phase=0; when iter=9522, then Phase=180; when iter=10062, then Phase=360.

Return to: CFDVAL2004 Case 1 - Intro Page

Return to: CFDVAL2004 - Intro Page

Return to: Data from Experiments - Intro Page

Return to: Turbulence Modeling Resource Home Page

Privacy Act Statement

Accessibility Statement

Responsible NASA Official: Ethan Vogel
Page Curator: Clark Pederson
Last Updated: 05/15/2021