Difference between revisions of "Tutorial Week 5"

Latest revision as of 23:02, 30 December 2020

You can print the PDF.

Salt water with concentration of $K$ in g/L flows into Tank A at a rate $a$ in L/min. The mixed solution in Tank A flows into Tank B at a rate $b$ L/min. Another pipe takes the solution in Tank B back into Tank A at a rate $c$ in L/min. Finally, solution drains out of Tank B at a rate $a$ in L/min. Note: the initial volume $V$ in each tank is the same and $a + c = b$ (to ensure that the volumes in the tanks are constant).
1. Write down the system in matrix form that gives the amount of salt in each tank (call these $A(t)$ and $B(t)$) at any given time.
2. Show that the eigenvalues of the matrix for this system are real, and therefore it is impossible for this system to have oscillations (damped or otherwise). \textit{Hint}: express the matrix in terms of $b$ and $c$ only. Optional: show that both eigenvalues are negative.

Consider the matrix $A = \left(\begin{array}{cc} 3\alpha - 8 & 7 \\ -4 & 3\alpha + 8 \end{array}\right).$
1. Find the eigenvalues and eigenvectors of $A$.
2. For what values of $\alpha$ (i) are both eigenvalues postive, (ii) are both eigenvalues negative, (iii) do the eigenvalues have opposite sign, (iv) is one eigenvalue zero?
3. Sketch the eigenvectors and some solution curves of the system $\mathbf{x'} = A\mathbf{x}$ in the case when $\alpha = 0$.

@@ Line 15: / Line 15: @@
 ## Sketch the eigenvectors and some solution curves of the system $\mathbf{x'} = A\mathbf{x}$ in the case when $\alpha = 0$.
-===Solutions:===
+[[Tutorial Week 5 Solutions]]
-<ol>
-<li> '''8pts total'''
-<ol>
-<li>'''2 pts''' for form (constant coefficient linear system with inhomogeneous terms), '''2 pts''' for the details of the terms (A, B) so total '''6 pts'''.
-\begin{equation}
-\begin{pmatrix}
-A'(t) \\ B'(t)
-\end{pmatrix}=
-\begin{pmatrix}
--b/V & c/V \\
-b/V & -(a + c)/V
-\end{pmatrix}
-\begin{pmatrix}
-A(t) \\ B(t)
-\end{pmatrix}
-+
-\begin{pmatrix}
-K a \\ 0
-\end{pmatrix}
-\end{equation} </li>
-<li> '''2pts''' By using $b = a + c$, eigenvalues are ('''1pt''') $\frac{-b \pm \sqrt{bc}}{V}.$ Since $b \ge 0$ and $c \ge 0$ these are real ('''1pt''' - must be explained). </li>
-</ol>
-<li>'''14pts total'''
-<ol>
-<li> '''6pts''' Eigenvalues are $3\alpha \pm 6$ ('''2pts''') with corresponding eigenvectors ('''4pts''') $e_1 = \begin{pmatrix}1\\2\end{pmatrix}$ and $e_2 = \begin{pmatrix}7\\2\end{pmatrix}.$
-</li>
-<li> '''4pts - one for each category''' There is a zero eigenvalue for $\alpha = \pm 2$, the eigenvalues are both negative for $\alpha < -2$, the eigenvalues are both positive for $\alpha > 2$, the eigenvalues have mixed signs for $-2 < \alpha < 2$.</li>
-<li> '''4pts''' Sketch of eigenvectors from (a) worth '''2pts''' and sketch of solution curves as below '''2pts'''.
-[[File:52cPhasePlotSketch.png]]
-</li>
-</ol>
-</li>
-</ol>