24.3 Exercises

Exercise 24.1 Through direct computation, verify the following calculations:

  1. When f(t)=14πDt, then f(t)=2πD(4πDt)3/2
  2. When g(t)=x24Dt, then g(t)=x24Dt2
  3. Verify that (12t+x24Dt2)=(x22Dt(2Dt)2)
 
Exercise 24.2 The equation for the normal distribution is f(x)=12πσe(xμ)2/(2σ2), with mean μ and variance σ2. Examine the formula for the diffusion equation (Equation (24.3)) and compare it to the formula for the normal distribution. If Equation (24.3) represents a normal distribution, what does μ equal? σ2?

 

Exercise 24.3 For this problem you will investigate p(x,t) (Equation @ref(eq:diffusion-equation}) with D=12.

  1. Evaluate 11p(x,10)dx. Write a one sentence description what this quantity represents.
  2. Using desmos or some other numerical integrator, complete the following table:
Equation Result
11p(x,10)dx=
11p(x,5)dx=
11p(x,2.5)dx=
11p(x,1)dx=
11p(x,0.1)dx=
11p(x,0.01)dx=
11p(x,0.001)dx=
  1. Based on the evidence from your table, what would you say is the value of lim?
  2. Now make graphs of p(x,t) at each of the values of t in your table. What would you say is occuring in the graph as \displaystyle \lim_{t \rightarrow 0^{+}} p(x,t)? Does anything surprise you? (The results you computed here lead to the foundation of what is called the Dirac delta function.)

 

Exercise 24.4 Consider the function \displaystyle p(x,t) = \frac{1}{\sqrt{4 \pi D t}} e^{-x^{2}/(4 D t)}. Let x=1.

  1. Explain in your own words what the graph p(1,t) represents as a function of t.
  2. Graph several profiles of p(1,t) when D = 1, 2, and 0.1. How does the value of D affect the profile?

 

Exercise 24.5 Consider the function \displaystyle p(x,t) = \frac{1}{\sqrt{\pi t}} e^{-x^{2}/t}:

  1. Using your differentiation skills compute the partial derivatives p_{t}, p_{x}, and p_{xx}.
  2. Verify p(x,t) is consistent with the diffusion equation \displaystyle p_{t}=\frac{1}{4} p_{xx}.

 

Exercise 24.6 For the one-dimensional random walk we discussed where there was an equal chance of moving to the left or the right. Here is a variation on this problem.

Let’s assume there is a chance v that it moves to the left (position x - \Delta x), and therefore a chance is 1-v that the particle remains at position x. The basic equation that describes the particle’s position at position x and time t + \Delta t is:

\begin{equation} p(x,t + \Delta t) = (1-v) \cdot p(x,t) + v \cdot p(x- \Delta x,t) \end{equation}

Apply the techniques of local linearization in x and t to show that this random walk to derive the following partial differential equation, called the advection equation:

\begin{equation} p_{t} = - \left( v \cdot \frac{ \Delta x}{\Delta t} \right) \cdot p_{x} \end{equation}

Note: you only need to expand this equation to first order