\documentclass{article} \usepackage{amssymb, amsmath, amsfonts, amsthm, graphics, enumerate, mathrsfs, mathtools, tikz-cd} \usepackage{geometry, soul, hyperref} \usetikzlibrary{positioning,arrows,scopes} \usepackage{fancyhdr} \pagestyle{fancy} % \setlength{\headheight}{26pt} % \setlength{\oddsidemargin}{-0.2in} % \setlength{\evensidemargin}{-0.2in} % \setlength{\topmargin}{0in} % \setlength{\textwidth}{6.5in} %\setlength{\headwidth}{6.5in} \setlength{\textheight}{8.5in} \theoremstyle{definition} \newtheorem*{thm}{Theorem} \newtheorem*{pthm}{Pre-Theorem} \newtheorem{prob}{Problem} \newtheorem{ex}{Exercise} \newtheorem*{dfn}{Definition} \newtheorem{eg}{Example} \newtheorem*{prop}{Proposition} \newtheorem*{rmk}{Remark} \newtheorem*{conj}{Conjecture} \newcommand{\CC}{\mathbb{C}} \newcommand{\FF}{\mathbb{F}} \newcommand{\RR}{\mathbb{R}} \newcommand{\ZZ}{\mathbb{Z}} \newcommand{\QQ}{\mathbb{Q}} \newcommand{\PP}{\mathbb{P}} \newcommand{\cO}{\mathcal{O}} \newcommand{\cU}{\mathcal{U}} \newcommand{\cM}{\mathcal{M}} \DeclareMathOperator{\Sym}{Sym} \DeclareMathOperator{\Div}{Div} \DeclareMathOperator{\Supp}{Supp} \DeclareMathOperator{\indeg}{indeg} \DeclareMathOperator{\Pic}{Pic} \DeclareMathOperator{\Hom}{Hom} \DeclareMathOperator{\im}{im} \newcommand{\dsp}{\displaystyle} \newcommand{\floor}[1]{\lfloor{{#1}}\rfloor} \newcommand{\ceil}[1]{\lceil{{#1}}\rceil} \title{log(m) meeting 6} %\date{\today} \lhead{ Log(M): Origami on lattices \\} \rhead{ Meeting 6 \\ 22 February 2019} \begin{document} %\maketitle %\begin{center} %Notes, Fall 2017, assorted %\end{center} \section{Logistics} \ul{Next meetings}: Wednesday February 27, 9-9:45am and Friday March 1, 12:15 - 1pm \medskip \noindent \ul{Midterm presentations}: Tuesday February 26, 5 - 6:30pm \bigskip {Expectations} for next meetings: \begin{itemize} \item Problem 7: If I am given a hexagon configuration in terms of the $6$ crease vectors, what are the $6$ dihedral angles in this configuration? \item Problem 8: mountain / valley labellings %Once we understand the neighborhood of each flat configuration, we can ask: % How do these neighborhoods fit together? % One approach: divide configuration space into regions labelled by whether each crease is a ``mountain'' or ``valley'' fold (or neither) \item \textbf{[Writing]} Continue writing up relevant discussion from this week in draft of final report \item \textbf{[Visualization]} Share code for updated visualization program \end{itemize} \section{Hexagon configuration space} \subsection{Changing coordinates} \setcounter{prob}{6} In the hexagon configuration space, we discussed two choices for how to put ``coordinates'' on this space: crease vectors and dihedral fold angles. How do we change coordinates between these, from crease vectors to dihedral angles? For example, if we have consecutive crease vectors \[ v_1 = (1,0,0),\, v_2 = (0,1,0),\, v_3 = (-1,0,0)\] then the corresponding dihedral angle is $\theta_2 = 0$. If we have consective crease vectors \[ v_2 = (0,1,0),\, v_3 = (-1,0,0),\, v_4 = (0,0,1)\] then the corresponding dihedral angle is $\theta_3 = \frac{\pi}{2}$. \begin{prob} Given three nonzero vectors $v_0,\,v_1,\,v_2 \in \RR^3$, what is a formula to express the dihedral angle between the planes $p_{01} = \RR v_0+\RR v_1$ and $p_{12} = \RR v_1 + \RR v_2$? \noindent (Possible hint: \url{https://en.wikipedia.org/wiki/Dihedral_angle#Mathematical_background}) \end{prob} Our goal will be to use this expression for the angles to express the {\em energy} \[ E(\Phi) = \left( \sum_i |\theta_i - \pi|^2 \right)^{1/2}\] of a given fold configuration in terms of the crease vectors, and to study configurations of constant energy. \subsection{Montain / valley diagrams} To understand a topological space it often helps to cut it up into smaller, more manageable pieces. For a fold configuration which is ``near flat,'' we say a crease is a {\em mountain fold} if it is higher than a secant line between its two adjacent flat regions, and a {\em valley fold} if it is lower. (We take the nearby flat configuration as reference for what ``higher'' and ``lower'' mean.) Suppose we label each crease in a configuration with ``mountain'' or ``valley'' or neither. Visually we can indicate these respectively by a solid line, a dotted line, or no line. \begin{eg} For two perpendicular creases (so there are $4$ total crease vectors), the following shows possible mountain/valley labellings: \begin{center} \includegraphics[scale=0.3]{mv_diagram_small} \end{center} The following mountain/valley labelling is not possible: %\begin{center} \qquad \includegraphics[scale=0.1]{mv_impossible} %\end{center} \end{eg} \begin{prob} In the following diagrams with $5$ creases coming from a central vertex, which mountain/valley labellings are possible? How do these fit together in configuration space near the unfolded state? \begin{enumerate} \item creases with equal spacings \begin{center} \includegraphics[scale=0.2]{5fold_equal} \end{center} \item creases perpendicular plus one at $45^\circ$ angle \begin{center} \includegraphics[scale=0.2]{5fold_square} \end{center} \end{enumerate} \end{prob} \subsection{Neighborhoods of flat configurations} The paper "Hodge Theory and the Art of Paper Folding" claims that for every point in the configuration space that does not lie completely in a plane (i.e. corresponds to a non-degenerate linkage), a small neighborhood around this point looks like an open ball in $\RR^3$. The $11$ flat configurations of a folded hexagon are discussed in the ``Discrete Folding'' paper, and shown in Figure 9: \begin{center} \includegraphics[scale=0.7]{flat_hexagon_configs} \end{center} \setcounter{prob}{4} \begin{prob} \,[see ``Hodge Theory'' paper] For each of the $11$ flat configurations of the hexagon, \begin{enumerate}[(a)] \item What are the parameters $f$, $b$, and $w$ as used in Theorem 1.1? \item What is the signature of the null cone describing this neighborhood of configuration space? \end{enumerate} \end{prob} \ul{Discussion}: For the unfolded configuration, we have $f=6$, $b=0$, and $w=1$ (OR $f=0$, $b=6$, $w=-1$.) The corresponding null cone has signature $(3,1)$. A small neighborhood in configuration space is a cone over two disjoint $2$-spheres $S^2 \times S^0$. For the other $10$ configurations, $f=3$, $b=3$ and $w=0$. The corresponding null cone has signature $(2,2)$. A small neighborhood in configuration space is a cone over the $2$-torus $S^1 \times S^1$. \section{Quadratic forms} Given an $n\times n$ (real) symmetric matrix $Q$ we can define a {\em quadratic form} on $\RR^n$ by \begin{align*} % \RR^n \times \RR^n &\to \RR \\ (u, v)\quad &\mapsto u^T Q v . \end{align*} We denote this map $f_Q: \RR^n \times \RR^n \to \RR$. The {\em null cone} of a quadratic form is the set \[ Z(Q) = \{x \in \RR^n \text{ such that }f_Q(x,x) = x^TQx = 0 \}. \] % If $x$ is in the null cone, then for any $\lambda \in \RR$ % \[ (\lambda x)^T Q (\lambda x) = \lambda^2 (x^T Q x) = 0\] % so $\{ \lambda x : \lambda \in \RR\}$ is contained in the null cone. % Why is this called a ``cone''? See % \url{https://en.wikipedia.org/wiki/Cone_(topology)}; % this makes $Z(Q)$ (essentially) the cone over the ``unit distance'' null vectors The {\em unit null cone} (nonstandard terminology) is the set $$\hat{Z}(Q) = \{ x \in\RR^n : x^TQx = 0 \text{ and } |x|^2=1\}.$$ % A quadratic form $f_Q$ is {\em nondegenerate} if the matrix $Q$ is invertible (i.e. all eigenvalues are nonzero). % The {\em nullity} of $f_Q$ is the dimension of the $0$-eigenspace of $Q$. % The {\em spectral theorem} says that up to a change of coordinates, % a nondegenerate quadratic form $f_Q$ is equivalent to % a form associated to a matrix % \[ Q = \begin{pmatrix} % 1 & 0 & \cdots & 0 & 0 & 0 \\ % 0 & \ddots & 0 & 0 & 0 & 0\\ % \vdots & 0 & 1 & 0 & 0 & 0\\ % 0 & 0 & 0 & -1 & 0 & \vdots \\ % 0 & 0 & 0 & 0 & \ddots & 0 \\ % 0 & 0 & 0 & \cdots & 0 & -1 % \end{pmatrix}.\] % The quadratic form has % {\em signature} $(p, q)$ % if there are $p$-many $+1$'s and $q$-many $-1$'s in this diagonal form. % We described the null cone of a quadratic form of signature $(1,1)$, $(2,1)$, and $(3,1)$. \begin{prob} For each case below, the null cone $Z(Q)$ is the cone over the ``unit null cone'' $\hat{Z}(Q)$; we want a nice geometric description of $\hat{Z}(Q)$. % (Hint: in the general case your answer should involve $n$-spheres; % see \url{https://en.wikipedia.org/wiki/N-sphere}) \begin{enumerate}[(a)] \item Describe the null cone of a quadratic form with signature $(1,3)$. \item Describe the null cone of a quadratic form with signature $(2,2)$. \item \,[Challenge(?)] Describe the null cone of a general quadratic form with signature $(p,q)$. \end{enumerate} \end{prob} \ul{Discussion} (b) When the signature is $(2,2)$, the unit null cone is described by \[ \hat{Z}(Q) = \{ (a,b,c,d) : a^2 + b^2 - c^2 - d^2 = 0,\, a^2 + b^2 + c^2 + d^2 = 1\} .\] By adding and subtracting the two constraints, this is equivalent to \[ \hat{Z}(Q) = \{ (a,b,c,d) : a^2 + b^2 = \frac12,\, c^2 + d^2 = \frac12\} .\] This shows that $\hat{Z}(Q)$ is a {\em product} of the spaces \[ \{ (a,b) : a^2 + b^2 = \frac12\} \times \{ (c,d) : c^2 + d^2 = \frac12\}.\] Each of these factors is a circle of radius $\sqrt{1/2}$, so topologically $\hat{Z}(Q) \cong S^1 \times S^1$. (c) In general, when the signature is $(p,q)$ the null cone \[ \hat{Z}(Q) = \{a_1^2 + \cdots + a_p^2 - b_1^2 - \cdots - b_q^2 = 0,\, a_1^2 + \cdots + b_q^2 = 1\}\] can be expressed equivalently as \[ \{a_1^2 + \cdots + a_p^2 = \frac12,\, b_1^2 + \cdots + b_q^2 = \frac12 \} \] so $\hat{Z}(Q) \cong S^p \times S^q$ is topologically a product of two spheres. \begin{eg} The null cone of of a quadratic form of signature $(1,1)$ is two lines passing through the origin. The unit null cone $\hat{Z}(Q)$ is a set of $4$ discrete points, which is equal to the product of spheres $S^0 \times S^0$. \end{eg} \begin{eg} The null cone of a quadratic form of signature $(2,1)$ is a ``doubled cone''. % \begin{center} % \includegraphics[scale=0.5]{double_cone} % % \tiny{Source: cliffnotes.com} % \end{center} The unit null cone $\hat{Z}(Q)$ is two disjoint circles, which is $S^1\times S^0$. \end{eg} \end{document}