Changeset 498 for trunk

Timestamp:

11/28/2009 12:34:09 AM (2 years ago)

Author:

powell

Message:

Major updates to the documentation.

Files:

• trunk/matml/src/ternary/ternary.tex.in (modified) (5 diffs)

trunk/matml/src/ternary/ternary.tex.in

@@ -39 +39 @@
   procedure at the edges.
 \item Added spinodal region boundary calculation.
+\item Added a graphical interface for setting the free energy parameters.
 \end{itemize}
 
 {\tt Ternary} is nearly complete, with the one un-implemented feature being to
 build and neatly visualize how free energy functions create a full
-temperature-composition ternary phase diagram.  It would also be nice to have
-these functions in a library, so other programs could call them.
+temperature-composition ternary phase diagram.
 
 \section{Compiling {\tt Ternary}}
@@ -101 +101 @@
 this procedure is pretty much the same.
 
+The {\tt Ternary} GUI does all of these things as well each time the user
+changes the free energy function parameters or otherwise updates the display.
+
 \section{Bugs}
 
-Version 0.5.0 is releasing with an incomplete implementation of
-{\tt hullRefine}
-which is fine for interior points, but does not attempt to refine points on the
-edges, i.e. pseudobinaries.
-
-Also, {\tt hullReturnPhaseBoundaries} returns more phases than appropriate in
-some cases, with some of them empty.
+There are several bugs in the Ternary algorithms, an incomplete implementation,
+and a problem in the GUI, as follows:
+
+\begin{itemize}
+\item Finding some phases depends on the initial discretization of the space.
+  For example, if there is a small miscibitily gap which none of the initial
+  points happens to be in, it will not show up in the phase diagram isotherm.
+
+\item The refining algorithm does not work well near a critical point.
+
+\item The edge-values test often produces two critical points and a three-phase
+  region where there should be just one critical point.
+
+\item Version 0.5.0 is releasing with an incomplete implementation of {\tt
+    hullRefine} which is fine for interior points, but does not attempt to
+  refine points on the edges, i.e. pseudobinaries.
+
+\item The function {\tt hullReturnPhaseBoundaries} returns more phases than
+  appropriate in some cases, with some of them empty.
+
+\item The {\tt Ternary} GUI does not destroy phase boundaries which are
+  eliminated by changes to the system, nor does it remove a spinodal which the
+  user asks to remove.
+
+\item The {\tt Ternary} GUI is incomplete: it is missing the polynomial and
+  Gaussian free energy terms, and cannot add or remove phases.
+\end{itemize}
 
 \section{Future Goals}
@@ -115 +138 @@
 The 1.0 release should have the following features:
 \begin{itemize}
-\item Fix for the above bugs.
+\item Fixes for as many of the above bugs as possible.
 \item An XML format for the free energy parameters, and ability to load and
   save in this format in {\tt libfreenergy}.
 \item Some ability to import and export ThermoCalc\textregistered database
   files.
-\item A graphical interface exposing all of the free energy parameters for
-  multiple free energy functions.
 \item Ability to click the isotherm view lower convex hull triangles to return
   the triangle number and its corners.
@@ -131 +152 @@
 \end{itemize}
 
+Other wishlist items include:
+\begin{itemize}
+\item ``Line compounds'' which consist of a single point, this is already
+  possible to do manually but is not in the GUI or either front end.
+\item Phases which do not cover the whole ternary space, also possible to do
+  manually but it would be nice to have some support in {\tt book.c}.
+\end{itemize}
+
 \section{Copyright}
 
@@ -194 +223 @@
 diagrams.  The APIs are relatively immature and bound to change.
 
+\subsection{{\tt Ternary} 0.5.0}
+
+The {\tt Ternary} GUI has landed, but is buggy and not yet complete.
+
+This version fixed major bugs in, and improved the accuracy of, the spinodal
+calculation.
+
+There are also new arbitrary polynomial terms in the free energy functions.
+And there is a new {\tt adjust\_gaussians} function to aid in the use of
+Gaussians for a single-function free energy surface with known free energies of
+certain phases.
+
 \bibliographystyle{unsrturl}
 \bibliography{ternary}