Normal modes of a rectangular elastic membrane

Click For Summary
SUMMARY

The discussion focuses on deriving the normal modes of a rectangular elastic membrane using the two-dimensional wave equation. The solution involves separating variables into functions of space and time, leading to expressions for the normal modes characterized by coefficients G and H. The participants clarify the use of trigonometric identities to express the solution in terms of amplitude A and phase angle φ. The final expressions for G and H are derived, confirming the relationship between the coefficients and the phase angle.

PREREQUISITES
  • Understanding of the two-dimensional wave equation
  • Familiarity with the method of separation of variables
  • Knowledge of trigonometric identities and their applications
  • Basic concepts of normal modes in wave mechanics
NEXT STEPS
  • Study the derivation of normal modes in different boundary conditions
  • Explore the application of Fourier series in wave equations
  • Learn about the physical interpretation of amplitude and phase in wave mechanics
  • Investigate the implications of standing waves versus traveling waves
USEFUL FOR

Students and researchers in physics, particularly those focusing on wave mechanics, elastic materials, and mathematical modeling of physical systems.

ContagiousKnowledge
Messages
17
Reaction score
2
Homework Statement
Consider a thin elastic sheet that spreads from x = 0 to x = x' and y = 0 to y = y' where the sheet's transverse displacement ψ is 0 x=0, y=0, x=x', and y=y'. at Demonstrate that the amplitudes and phase angles in the normal mode expansion are given by $$ A_{i,j} = \sqrt {G^{2}{}_{i,j} + H^{2}{}_{i,j}} $$
and
$$ tan^{-1}(\frac {H_{i,j}} {G_{i,j}}) $$
Confirm that
$$ G_{i,j} = \frac {4} {x'y'} \int_{0}^{y'} \int_{0}^{x'} \psi(x, y, 0) sin(\frac {\pi i} {x'}x)sin(\frac {\pi j} {y'}y) dx dy $$
$$ H_{i,j} = \frac {4} {x'y' \omega_{i,j}} \int_{0}^{y'} \int_{0}^{x'} \frac {\partial \psi(x, y, 0)} {\partial t} sin(\frac {\pi i} {x'}x)sin(\frac {\pi j} {y'}y) dx dy $$
Relevant Equations
$$ \frac {\partial^2 \psi} {\partial t^2} = v^2( \frac {\partial^2 \psi} {\partial x^2} + \frac {\partial^2 \psi} {\partial y^2} ) $$
Let's try inputting a solution of the following form into the two-dimensional wave equation: $$ \psi(x, y, t) = X(x)Y(y)T(t) $$
Solving using the method of separation of variables yields
$$ \frac {v^2} {X(x)} \frac {\partial^2 X(x)} {\partial x^2} + \frac {v^2} {Y(y)} \frac {\partial^2 Y(y)} {\partial y^2} =\frac {1} {T(t)} \frac {\partial^2 T(t)} {\partial t^2} =-\omega^2$$

$$ T(t) = A_{i,j}cos(\omega_{i,j} t) + B_{i,j}sin(\omega_{i,j} t)$$
$$ X(x) =Ccos(k_{x} t) + D_{i,j}sin(k_{x} t) $$
$$ Y(y) =Ecos(k_{y} t) + Fsin(k_{y} t) $$
where
$$ \frac {1} {X(x)} \frac {\partial^2 X(x)} {\partial x^2} = -k_{x}^2 $$
$$ \frac {1} {Y(y)} \frac {\partial^2 Y(y)} {\partial y^2} = -k_{y}^2 $$
The boundary conditions ψ(0, y, t) = ψ(x, 0, t) = 0 tells us that C=E=0.
the conditions ψ(x', y, t) = ψ(x, y', t) = 0 tell us that
$$ k_{x} = \frac {\pi i} {x'}, $$
$$ k_{y} = \frac {\pi j} {y'} $$
$$ \omega = \pi v \sqrt {\frac{i^{2}}{x'^{2}}+\frac{j^{2}}{y'^{2}}} $$
From this, we can ascertain the particular solution for given i and j
$$ \psi(x, y, t) = [G_{i,j}cos(\omega_{i,j} t) + H_{i,j}sin(\omega_{i,j} t)]sin(\frac {\pi i} {x'}x)sin(\frac {\pi j} {y'}y)$$
$$ G_{i,j}=A_{i,j}D_{i}F_{j}, H_{i,j}=B_{i,j}D_{i}F_{j} $$
Since the wave equation is linear, the general solution is a superposition of all normal modes.
$$ \psi(x, y, t) = \Sigma_{i=0,\infty} \Sigma_{j=0,\infty}[G_{i,j}cos(\omega_{i,j} t) + H_{i,j}sin(\omega_{i,j} t)]sin(\frac {\pi i} {x'}x)sin(\frac {\pi j} {y'}y) $$

$$ \psi(x, y, 0) = \Sigma_{i=0,\infty} \Sigma_{j=0,\infty}G_{i,j}sin(\frac {\pi i} {x'}x)sin(\frac {\pi j} {y'}y) $$
$$ G_{i,j} = \frac {4} {x'y'} \int_{0}^{y'} \int_{0}^{x'} \psi(x, y, 0) sin(\frac {\pi i} {x'}x)sin(\frac {\pi j} {y'}y) dx dy $$

$$ \frac {\partial \psi(x, y, 0)} {\partial t} = \Sigma_{i=0,\infty} \Sigma_{j=0,\infty} \omega_{i,j} H_{i,j}sin(\frac {\pi i} {x'}x)sin(\frac {\pi j} {y'}y) $$
$$ H_{i,j} = \frac {4} {x'y' \omega_{i,j}} \int_{0}^{y'} \int_{0}^{x'} \frac {\partial \psi(x, y, 0)} {\partial t} sin(\frac {\pi i} {x'}x)sin(\frac {\pi j} {y'}y) dx dy $$
I've omitted some math because entering it with latex was taking a long time, but I have arrived at the correct expressions for G and H. I don't know how to proceed from here.
 
Last edited:
Physics news on Phys.org
ContagiousKnowledge said:
$$ T(t) = A_{i,j}cos(\omega_{i,j} t) + B_{i,j}cos(\omega_{i,j} t)$$
Did you mean to use a cosine function in both terms?

I have arrived at the correct expressions for G and H. I don't know how to proceed from here.
An expression of the form ##G\cos \omega t + H\sin \omega t## may be expressed in terms of an amplitude ##A## and a phase angle ##\phi## as ##A \cos (\omega t + \phi)## or ##A \cos (\omega t - \phi)##. The choice of + or - in front of ##\phi## is a matter of convention. ( I don't know which convention you use in your course.) Can you derive expressions for ##A## and ##\phi## in terms of ##G## and ##H##?
 
No, I did not intend to use cosine for both terms.

I am accustomed to using -ϕ to indicate a wave moving in the positive direction and +ϕ for a wave moving in the negative direction.

As for deriving expressions for the amplitude and phase angle for the normal modes in terms of G and H, could you elaborate a little more on what you are asking of me? I know that it is possible to derive such expressions because it is stated in the problem; I assume your question was rhetorical. My ultimate goal is to figure out how one would derive these expressions.
 
ContagiousKnowledge said:
No, I did not intend to use cosine for both terms.
OK

As for deriving expressions for the amplitude and phase angle for the normal modes in terms of G and H, could you elaborate a little more on what you are asking of me? I know that it is possible to derive such expressions because it is stated in the problem; I assume your question was rhetorical. My ultimate goal is to figure out how one would derive these expressions.
Use the trig identity ##\cos(\alpha + \beta) = \cos(\alpha)\cos(\beta)-\sin (\alpha) \sin(\beta)## to help you write ##G\cos(\omega t) + H\sin(\omega t)## in the form ##A\cos(\omega t + \phi)##.
 
I've been playing with this algebra for a while and I'm afraid I don't see how to rewrite the expression in that form.
$$ Gcos( \omega t - \phi) + Hsin( \omega t - \phi) = G[cos( \omega t)cos( \phi ) + sin( \omega t)sin( \phi )] + H[sin( \omega t)cos( \phi ) - cos( \omega t )sin( \phi )] $$
Which can be rewritten as
$$ [Gcos( \phi ) - Hsin( \phi )]cos( \omega t) + [Gsin( \phi ) + Hcos( \phi)]sin( \omega t) $$
although this doesn't seem particularly useful. I have tried rearranging the expression in numerous ways. Presumably, we will have to square both sides of the equation eventually since H and G are both squared in the given expression for amplitude. I tried squaring the expression in its current form, but the result is fairly convoluted and doesn't appear to simplify to anything useful.
I apologize if I'm missing something simple.
 
Apply the trig identity to ##A\cos(\omega t + \phi)## and compare with ##G\cos \omega t + H\sin \omega t##.
 
@TSny Again, I've been rearranging trig functions for a long time, but it hasn't been very productive.
$$ Acos(\omega t + \phi) = A[cos(\omega t)cos(\phi) - sin(\omega t)sin(\phi)] = Gcos(\omega t) + Hsin(\omega t) $$
$$ A = \frac{Gcos(\omega t) + Hsin(\omega t)}{cos(\omega t)cos(\phi) - sin(\omega t)sin(\phi)} $$
We could try setting ϕ to different values to generate more expressions
$$ \phi = 0 $$ $$ A = G + Htan(\omega t) $$
$$ \phi = \frac{\pi}{2} $$ $$ -A = Gtan(\omega t) + H $$
Although, we could have derived this without using the cos(a + b) identity, and this doesn't seem to bring me closer to the answer regardless.
Among other attempts, I've also tried substituting the given expression for A directly hoping to find something that clearly resembles Gcos(ωt)+Hsin(ωt), but to no avail.
$$ Acos(\omega t + \phi) = \sqrt{G^{2} + H^{2}}[cos(\omega t)cos(\phi) - sin(\omega t)sin(\phi)] = \sqrt{(G^{2} + H^{2})(cos^{2}(\omega t)cos^{2}(\phi) - 2cos(\omega t)cos(\phi)sin(\omega t)sin(\phi) + sin^{2}(\omega t)sin^{2}(\phi))}$$
 
Last edited:
ContagiousKnowledge said:
@TSny Again, I've been rearranging trig functions for a long time, but it hasn't been very productive.
$$ Acos(\omega t + \phi) = A[cos(\omega t)cos(\phi) - sin(\omega t)sin(\phi)] = Gcos(\omega t) + Hsin(\omega t) $$
Look at the second equality. What is the coefficient of ##\cos \omega t## on each side of this equality? What is the coefficient of ##\sin \omega t## on each side of this equality?
 
@TSny
$$ Acos(\phi)cos(\omega t) = Gcos(\omega t) $$ $$ G = Acos(\phi) $$
$$ -Asin(\omega t)sin(\phi) = H sin(\omega t) $$ $$ H = -Asin(\phi) $$
$$ \sqrt{G^{2} + H^{2}} = \sqrt{A^{2}cos^{2}(\phi) + A^{2}sin^{2}(\phi)} = A $$
Okay, I should have seen that; thank you for your help. Now, as for ϕ
$$ \frac{H}{G} = \frac{-sin(\phi)}{cos(\phi)}$$
I think ϕ should have been negative, which is the default since a negative ϕ conventionally indicates a wave moving in the positive direction. This would get red of the negative sign in the final expression, and the resultant expression would prove that
$$ arctan(\frac{H}{G}) = \phi $$
This question was much simpler than I thought. Thank you for your help.
 
  • #10
OK. Good.

I've never heard of choosing the sign of ##\phi## based on the direction of travel of a wave. But, go with the convention for your course. In this problem, you are dealing with standing waves (normal modes) rather than traveling waves. ##\phi## is just a phase angle associated with the time dependence of the standing wave.
 

Similar threads

Replies
4
Views
1K
Replies
4
Views
2K
Replies
1
Views
2K
Replies
19
Views
3K
  • · Replies 2 ·
Replies
2
Views
796
  • · Replies 30 ·
2
Replies
30
Views
3K
  • · Replies 20 ·
Replies
20
Views
3K
Replies
16
Views
3K
Replies
9
Views
2K
  • · Replies 5 ·
Replies
5
Views
2K