CPLOT.DOC 13 May 1996 So, did the original version of CPLOT write out a TECPLOT file? If so, I should dig it out so I can try a TECPLOT data file... 06 May 1996 I added the ability to shade in the crystal, melt, or void. Still a slight defect in plots of the crystal region, because ICRYS2 and JCRYS2 (one of them, anyway), measures the rightmost left corner node, or the rightmost right corner node, I forget which. The mistake shows on plots. 02 May 1996 Replaced "REFLECT" by "DOUBLE" and "HALF". I don't like those "switch" commands so much anymore. Now working on the problem of the offset left hand side in the doubling scheme. FIXED. Now working on problem of color contours not matching up with control volume sides. There seems to be a confusion between the primary and corner nodes again, in this problem of color contours. As part of my effort to avoid this in the future, I am renaming "X" to "XP". Fixed a flaw which meant before that the crystal boundary was ALWAYS drawn. 17 April 1996 Two problems: 1) If you type the command "dat=file" and "file" doesn't exist, you can never get out of the error condition. 2) If you REFLECT the data, do a zoom, and then type FULL, you go back to the HALF region limits, in which only a bit of the reflected data shows up. I fixed those, and noted two new problems! 1) Unless you choose a color table, the first plot has an ugly color table, mostly black, and GRFINI whines that no color table has been set. 2) When you reflect data, the crystal/void boundary on the new, reflected left hand side is slightly further to the left than the colored data, suggesting a discrepancy. 12 April 1996 I modified the way the color bar is set up, making the default be ITABLE=9, and excluding the lowest and highest colors. I renamed the VSCALE command to SCALEV. 08 April 1996 Added SCALENB, SCALENC, and SCALENP commands for scale factors for bottom, corner, and primary nodes. Replaced ICRYS, JCRYS by ICRYS1, ICRYS2, JCRYS1, JCRYS2, so that I can list the left and right, top and bottom of the crystal, whether I reflect the region or not. 05 April 1996 I improved the method of displaying the control volumes, so that the blanking or display of corner nodes, control volumes, and primary nodes should be more consistent. 31 March 1996 Compiled a "dictionary" file CPLOT.DICT of all the variable names. I updated the definitions of KMELT and KVOID to add the changes I already made to KCRYS. 28 March 1996 I added "CVN", to display the control volume numbers, but I haven't tested it yet. I got the vorticity plots to work in DISPLAY, so now I've come back to CPLOT to see what I can make of them. 26 March 1996 I ran CRYSTAL to get some vorticity data, but CPLOT fails when it tries to read the file. I tracked down the problem, finally: In CRYSTAL, the HKSI and HETA variables are not set for the void volumes. So when I call GRADNT to get the derivatives of velocity, I was dividing by zero, but getting a NAN result instead of a warning, and the problem only showed up when CPLOT tried to read the data. I got my first contour plots of vorticity, and didn't like them much. I corrected the graphing of the crystal boundary by refining the definition of KCRYS, which I now interpret to apply to the control volume or cell, rather than to the primary node. I should similarly refine KMELT and KVOID. KCRYS(I,J)=0, cell is exterior to the crystal. 1, cell is outside the crystal, but touching boundary. 2, cell is inside the crystal, but touching boundary. 3, cell is interior to the crystal. 25 March 1996 I added the capability to draw vorticity contours and colors, given that CRYSTAL writes that stuff out to the plot file now. But I haven't tried it out yet. 23 February 1996 I changed LVOID(I,J) (and LCRYS and LMELT). Instead of being LOGICAL, it is an INTEGER, having the values: KVOID(I,J)=0, node is not in or on the void. 1, node is on void boundary. 2, node is in void interior. That way, I can blank out all the nodes in the void, or just the nodes in the interior, and so on, without the painful test IF(LVOID(I,J).and..not.(LMELT(i,j).or.LCRYS(i,j)))then replacing it by if(kvoid(i,j).eq.2)then... Also, I added CRYS, MELT and VOID commands, to switch display of data inside the crystal, melt and void regions. Unhappily, CRYS conflicts with the CRY command for showing the crystal side... I took a look at the resulting YCC plots, and it looked almost OK, although there were some funky lines around the crystal... I have to decide whether the boundary between melt and crystal is through primary or corner nodes... 26 January 1996 Added "stream colors" and velocity magnitudes. Added low white to high black color table. 22 January 1996 Mainly for debugging, I added commands to plot the X and Y coordinates as line or color contours. 18 December 1995 Added the REFLECT command, to reflect around the Y axis. It seems to be working. Added the usual "echo" file to keep a copy of the input. 15 December 1995 After a confusing error this morning, I decided to dimension all arrays MAXL, MAXM rather than L, M. 14 December 1995 Added the DIFF command to read two restart files and compute their difference, for display. Added the DXCDP command to display (dXCdP,dYCdP) vector plots. Unfortunately, my first DXCDP plot is not very believable... 13 December 1995 Replaced "GETDAT" by "RSREAD", which reads more data from CRYSTAL. Corrected a tiny error in TRILHK that caused occasional gaps in color contour plots, which I had already experienced and corrected in DISPLAY. 16 September 1995 Added corrected pressure, P+PC, but PC seems to be zero or very small. Try to "mask" out areas where plots are meaningless. In particular, turn off nodes in the interior of the crystal, or in the "void", for drawing pressure. 13 September 1995 Angular momenta are all zero. I fixed the computation of contour lines so that the "ragged" boundary line doesn't occur. I fixed the computation of color contours so that each patch of space is only drawn once, drastically cutting down on the size of the file. This is an improvement I should transfer to DISPLAY. I still need to * find out why the color tables don't correspond to their descriptions. * find out how to shift the image so I can put in the color bar. 12 September 1995 Max figured out that X and Y were interchanged in the plot. I figured out where the crystal side was, and the free surface. We are wondering if there is anything computed in the "void" area. I am trying to fix the color bars (which don't show up if the picture fills the page), and the color plots, and to get the control volumes to be plotted. Got the unit velocities to show up, and pressures, and control volumes. I tried to graph the turbulent K, turbulent Epsilon, and the viscosity, but they all seem to be zero. 11 September 1995 Trying to figure out what XIN, YIN are doing. They are the crucible boundary. Now need to show the flow surface and crystal interface. Today, the following showed up: Temperature contour lines. Streamlines. Velocity vectors. Are the velocities actually node centered? Have no idea how to draw the control volumes. Temperature colors exhibit an error. Pressures? 10 September 1995 Got the boundary to show up. Got the nodes to show up.