rearrange figures

Author: akohlmey

lammps-tutorials.tex

@@ -2327,18 +2327,6 @@ \subsubsection{Stretching the PEG molecule}
 fix mynvt all nvt temp 300 300 100
 fix myrct PEG recenter 0 0 0 shift all
 \end{lstlisting}
-
-\begin{figure}
-\centering
-\includegraphics[width=\linewidth]{PEG-distance}\\[-2ex]
-\caption{a) Evolution of
-the radius of gyration $R_\text{gyr}$ of the PEG molecule
-from \hyperref[all-atom-label]{Tutorial 3}, with the force
-applied starting at $t = 15\,\text{ps}$.  b) Histograms of the dihedral angles of type 1
-in the absence (orange) and in the presence (blue) of the applied force.}
-\label{fig:PEG-distance}
-\end{figure}
-
 To investigate the stretching of the PEG molecule, let us compute its radius of
 gyration~\cite{fixmanRadiusGyrationPolymer1962a} and the angles of its dihedral
 constraints using the following commands:
@@ -2356,6 +2344,17 @@ \subsubsection{Stretching the PEG molecule}
 $\phi$ (compute \lmpcmd{dphi}) is returned as a vector by the \lmpcmd{compute dihedral/local}
 command and must be written to a file using the \lmpcmd{dump local} command.
 
+\begin{figure}
+\centering
+\includegraphics[width=\linewidth]{PEG-distance}\\[-2ex]
+\caption{a) Evolution of
+the radius of gyration $R_\text{gyr}$ of the PEG molecule
+from \hyperref[all-atom-label]{Tutorial 3}, with the force
+applied starting at $t = 15\,\text{ps}$.  b) Histograms of the dihedral angles of type 1
+in the absence (orange) and in the presence (blue) of the applied force.}
+\label{fig:PEG-distance}
+\end{figure}
+
 Finally, let us simulate 15 picoseconds without any external force:
 \begin{lstlisting}
 run 15000
@@ -2854,17 +2853,6 @@ \subsubsection{System preparation}
 The \lmpcmd{undump} command is used to cancel the previous \lmpcmd{dump} command.
 Then, a new \lmpcmd{dump} command with a larger dumping period is used.
 
-\begin{figure}
-\centering
-\includegraphics[width=\linewidth]{NANOSHEAR-equilibration}\\[-2ex]
-\caption{a)~Pressure, $p$, of the nanosheared electrolyte system
-simulated in \hyperref[sheared-confined-label]{Tutorial 4} as a function of the
-time, $t$.  b)~Distance between the walls, $\Delta z$, as a function of $t$.
-\textcolor{blue}{The orange line
-shows the raw data, and the blue line represents a time-averaged curve.}}
-\label{fig:NANOSHEAR-equilibration}
-\end{figure}
-
 \begin{note}
 {\color{blue}
 Just like the \lmpcmd{undump} command can cancel an active \lmpcmd{dump}, other
@@ -2888,6 +2876,17 @@ \subsubsection{System preparation}
 variables \lmpcmd{walltopz} and \lmpcmd{wallbotz}, i.e.~the distance between the
 two centers of mass of the walls.
 
+\begin{figure}
+\centering
+\includegraphics[width=\linewidth]{NANOSHEAR-equilibration}\\[-2ex]
+\caption{a)~Pressure, $p$, of the nanosheared electrolyte system
+simulated in \hyperref[sheared-confined-label]{Tutorial 4} as a function of the
+time, $t$.  b)~Distance between the walls, $\Delta z$, as a function of $t$.
+\textcolor{blue}{The orange line
+shows the raw data, and the blue line represents a time-averaged curve.}}
+\label{fig:NANOSHEAR-equilibration}
+\end{figure}
+
 Finally, let us run the simulation for 30~ps by adding a \lmpcmd{run} command
 to \flecmd{equilibrate.lmp}:
 \begin{lstlisting}
@@ -3118,6 +3117,16 @@ \subsubsection{Prepare and relax}
 environment {\color{blue}through charge equilibration}.
 \end{note}
 
+\begin{figure}
+\includegraphics[width=\linewidth]{SIO-slice}
+\caption{A slice of the amorphous silica simulated during
+\hyperref[reactive-silicon-dioxide-label]{Tutorial 5}, where atoms are colored by their charges.
+Dangling oxygen groups appear in greenish, bulk Si atoms with a charge of about
+$1.8~\text{e}$  appear in red/orange, and bulk O atoms with a charge of about
+$-0.9~\text{e}$ appear in blue.}
+\label{fig:SIO-slice}
+\end{figure}
+
 Next, copy the following three crucial lines into the \flecmd{relax.lmp} file:
 \begin{lstlisting}
 pair_style reaxff NULL safezone 3.0 mincap 150
@@ -3147,7 +3156,6 @@ \subsubsection{Prepare and relax}
 If no control file is provided, as in this tutorial, LAMMPS uses its default values,
 which correspond to those in Adri van Duin's original stand-alone ReaxFF code~\cite{van2001reaxff}.}
 \end{note}
-
 Next, add the following commands to the \flecmd{relax.lmp} file to track the
 evolution of the charges during the simulation:
 \begin{lstlisting}
@@ -3190,16 +3198,6 @@ \subsubsection{Prepare and relax}
 the histogram is updated and averaged, \lmpcmd{-1.5 2.5} set the value range,
 \lmpcmd{1000} is the number of bins, and \lmpcmd{v\_vq} is the variable being histogrammed.}
 
-\begin{figure}
-\centering
-\includegraphics[width=\linewidth]{SIO-charge}\\[-2ex]
-\caption{a) Average charge per atom of the silicon, $q_\text{Si}$, atoms as
-a function of time, $t$, during equilibration of the $\text{SiO}_2$ system
-from \hyperref[reactive-silicon-dioxide-label]{Tutorial 5}.  b) Volume of the
-system, $V$, as a function of $t$.}
-\label{fig:SIO-charge}
-\end{figure}
-
 We can also use the \lmpcmd{fix reaxff/species} to evaluate what species are
 present within the simulation.  It will be useful later when the system is deformed,
 and bonds are broken:
@@ -3221,6 +3219,18 @@ \subsubsection{Prepare and relax}
 Run the \flecmd{relax.lmp} file using LAMMPS.  As seen from \flecmd{relax.species},
 only one species is detected, called \lmpcmd{O384Si192}, representing the entire system.
 
+
+\begin{figure}
+\centering
+\includegraphics[width=\linewidth]{SIO-charge}\\[-2ex]
+\caption{a) Average charge per atom of the silicon, $q_\text{Si}$, atoms as
+a function of time, $t$, during equilibration of the $\text{SiO}_2$ system
+from \hyperref[reactive-silicon-dioxide-label]{Tutorial 5}.  b) Volume of the
+system, $V$, as a function of $t$.}
+\label{fig:SIO-charge}
+\end{figure}
+
+
 \begin{note}
 {\color{blue}With the \lmpcmd{aniso} keyword, the three dimensions of the simulation
 box can change independently.  This is particularly relevant for solids and other
@@ -3239,13 +3249,12 @@ \subsubsection{Prepare and relax}
 plot the probability distributions, $P(q)$ (see Fig.~\ref{fig:SIO-distribution}\,a).
 
 \begin{figure}
-\includegraphics[width=\linewidth]{SIO-slice}
-\caption{A slice of the amorphous silica simulated during
-\hyperref[reactive-silicon-dioxide-label]{Tutorial 5}, where atoms are colored by their charges.
-Dangling oxygen groups appear in greenish, bulk Si atoms with a charge of about
-$1.8~\text{e}$  appear in red/orange, and bulk O atoms with a charge of about
-$-0.9~\text{e}$ appear in blue.}
-\label{fig:SIO-slice}
+\includegraphics[width=\linewidth]{SIO-distribution}\\[-4ex]
+\caption{a) Probability distributions of charge of silicon (positive, blue) and oxygen
+(negative, orange) atoms during the equilibration of the $\text{SiO}_2$ system
+from \hyperref[reactive-silicon-dioxide-label]{Tutorial 5}.  b) Same probability distributions
+as in panel (a) after the deformation.}
+\label{fig:SIO-distribution}
 \end{figure}
 
 \subsubsection{Deform the structure}
@@ -3285,15 +3294,6 @@ \subsubsection{Deform the structure}
 The only difference with the previous \flecmd{relax.lmp} file is the path to
 the \flecmd{relax.data} file.
 
-\begin{figure}
-\includegraphics[width=\linewidth]{SIO-distribution}\\[-4ex]
-\caption{a) Probability distributions of charge of silicon (positive, blue) and oxygen
-(negative, orange) atoms during the equilibration of the $\text{SiO}_2$ system
-from \hyperref[reactive-silicon-dioxide-label]{Tutorial 5}.  b) Same probability distributions
-as in panel (a) after the deformation.}
-\label{fig:SIO-distribution}
-\end{figure}
-
 Next, let us use \lmpcmd{fix nvt} instead of \lmpcmd{fix npt} to apply a
 Nosé-Hoover thermostat without a barostat:
 \begin{lstlisting}
@@ -3465,7 +3465,6 @@ \subsubsection{Decorate the surface}
 next a
 jump SELF loop
 \end{lstlisting}
-
 Run the simulation with LAMMPS.  When the simulation is over,
 it can be seen from the \flecmd{decorate.species} file that
 all the created hydrogen atoms reacted with the $\text{SiO}_{2}$ structure to