DocumentaionDocumentation for fieldview can be found under /usr/local/fv/doc in our own filesystem and on the web.
Firing it up...First of all you want the flieldview binary directory in your $PATH, so add s.th. like this in you shell configuration file (if you use cshell):
set path=($path /usr/local/fv/bin)
Last time I checked the system configuration file is not set up to include this directory in the path variable, so you have to do it in you own config file. Other than that you should not have to set any variables for fielview.
If you're logged into Ape just type fv at the shell prompt to start fieldview. If you're logged in from a different machine, you might have to say fv -x instead, if you don't have OpenGL support. Check the documentation for details. In any case you need a minimum of 24bit color depth to run fieldview. Running it remotely also makes it a lot slower (and you will not be able to use some of the mouse gestures that make fieldview so convenient. Or at least I haven't figured out how to use them, see below).
Example: Multiblock-Structured Meshes and Flow Data Postprocessing using Fieldview and the Plot3D FormatMy own experience is pretty much limited to this, so that's really all I can talk about. Many things however should apply generally.
I found mesh visualization with flieldview very convenient, because it is very fast and allows you to easily create slices of a structured mesh and sweep them through the domain, which is very helpful for checking the connectivity of a block-structured mesh.
If you use FLO107-MB, the mesh and flow solution files are written out in PLOT3D format. Filedview needs this data in single precision for some reason, so you might have to convert your files, or make sure you write them out in single precision. This little Fortran90 example converts a PLOT3D mesh from double precision to single precision. It is mainly intended to show how the PLOT3D format works, in case you happen to have the mesh in a different format.
To read the mesh into Filedview select "PLOT3D Data Input" from the pull-down menu "Data Files" in the Fieldview Toolbar menu. In the dialog window that pops up, you have to specify the format of your mesh. Most likely it will be unformatted (if you use FLO107) or binary. Choose "XYZ" as data type and, if you're reading in a Multiblock mesh, choose "Multi-Grid" as data format (They use the term Multi-Grid in the menu, but this has nothing to do with multigrid methods, it refers to multiblock meshes). Then choose your mesh file and click on "OK". After the mesh has been read, exit the popup window and play around with the mouse buttons in the graphics window a little bit to see how to navigate. It's pretty straightforward. As mentioned before if you're running fieldview remotely, you most likely can't use the zoom, rotate, etc. mouse gestures. In that case select "Option" in the "Transform Controls" window, and select the "running" radio button under "Mouse Interpretation".
To see parts of the mesh, you have to create an active surface in the "Computational Surface" window. You can sweep a slice of a mesh that you've created with the "Sweep Control" function. Just check the "sweep" button. Normally you would create a surface on a solid body boundary to have a look at the mesh there. You can also create an active surface on a cut anywhere in the computational domain for visualization of the flow solution there. To visualize a flow solution you have to have read in the mesh first. Then you go to "PLOT3D Data Input" in the pull-down menu "Data Files" again. This time you select Q as the data type.
If you have already created a computational surface where you want the data to be displayed, you can for example choose "Scalar" in the "Computational Surface" panel to change from mesh display to a display of a scalar variable of the solution. In the lower left corner of the panel check "Display Type" for several different rendering options.
Which variable is being displayed, can be changed in the Function menu in Filedview's toolbar. You can create your own functions, if you click on Create in the Function menu.
Another useful feature is the "Surface Plot" button on the "Computational Surface" panel. For example, it is straightforward to create surface-cp plots and save them as postscript files.