Mary L Boas - Mathematical Methods in The Physical Sciences 2ed - Physicsma (563-600)

Download as pdf
Download as pdf
You are on page 1of 38
PARTIAL DIFFERENTIAL EQUATIONS ~ 1, INTRODUCTION Many of the problems of mathematical physics involve the solution of partial differen tial equations. The same partial differential equation may apply to a variety of physical problems; thus the mathematical methods which you will learn in this chapter apply to many more problems than those we shall discuss in the illustrative examples. Let us outline the partial differential equations we shall consider, and the kinds of physical problems which lead to each of them. (LI) Laplace's equation = V’u = 0 The function u may be the gravitational potential in a region containing no matter, the electrostatic potential in a charge-free region, the steady-state temperature (that is, temperature not changing with time) in a region containing no source of heat, or the Yelocity potential for an incompressible fluid with no vortices and no sources or sinks. (1.2) Poisson's equation — Vu = f(x, y, 2) _ The function u may represent the same physical quantities listed for Laplace's equa- tion, but in a region containing matter, electric charge, or sources of heat or fluid, tespectively, for the various cases. The function f(x, y, 2) is called the source density; for example, in electricity it is proportional to the density of electric charge 542_ PARTIAL DIFFERENTIAL EQUATIONS. chy (1.3) The diffusion or heat flow equation — V‘u = 1 a Here u may be the non-steady-state temperature (that is, temperature varying with time) in a region with no heat sources; or it may be the concentration of a diffusing substance (for example, a chemical, or particles such as neutrons). The quantity 2? is a constant known as the diffusivity. (14) Wave equation Vw Here w may represent the displacement from equilibrium of a vibrating string or membrane or (in acoustics) of the vibrating medium (gas, liquid, or solid); in electricity u may be the current or potential along a transmission line; or « may be a component of E or H in an electromagnetic wave (light, radio waves, etc.). The quantity v is the speed of propagation of the waves; for example, for light waves in a vacuum it is ¢, the velocity of light, and for sound waves it is the speed at which sound travels in the medium under consideration, (13) Helmholtz equation VF + ?F =0 As you will see later, the function F here represents the space part (that is, the time-independent part) of the solution of either the diffusion or the wave equation. We shall be principally concerned with the solution of these equations rather than their derivation. If you like, you could say that it is true experimentally that the physical quantities mentioned above satisfy the given equations. However, it is also true that the equations can be derived from somewhat simpler experimental assumptions. Let us indicate briefly an example of how this can be done. In Chapter 6, Sections 10 and 11, we considered the flow of fluid. We showed (Chapter 6, Problem 10.15) that V - v =0 for an incompressible fluid in a region containing no sources or sinks. I also true that there are no vortices (that is, the flow is irrotational), then curl v = 0, and v can be written as the gradient of a scalar function: v = Vu. Combining these two equations, we have V « Yu = V?u = 0. The function w is called the velocity poten- tial and we see that (under the given conditions) it satisfies Laplace’s equation as we claimed. A few more examples of such derivations are outlined in the problems. In the following sections, we shall consider a number of physical problems to illus- trate the very useful method of solving partial differential equations known as separation of variables (no relation to the same term used in ordinary differential equations, Chapter 8), In Sections 2 to 4, we consider problems in rectangular coordinates leading to Fourier series solutions—problems similar to those solved by Fourier. In later sections, we consider use af other coordinate systems (cylindrical, spherical) leading to solutions using Legendre or Bessel series. PROBLEMS, SECTION 1 1, Assume from electricity the equations V + D = p (D = electric displacement and p = charge density) and D = —eV@ (¢ = electrostatic potential and € = dielectric constant). Show that Sec.2 LAPLACE $ EQUATION 543 the electrostatic potential satisfies Laplace’s cauation (1.1) in a charge-free region and satisfies Poisson’s equation (1.2) in a region of charge density 9. 2. Show that the expression w= sin (x — vt) describing a sinusoidal wave, satisfies the wave equation. Show that in general uafix—v) and w= f(x +0 satisfy the wave equation (1.4), where fis any function with a second derivative. 3, Assume from electricity the following equations which are valid in free space. (They are called Maxwell’s equations.) From them show that any component of E or H satisfies the wave equation (.4) with v= (es) "2, Hint: Use vector identity (e) in the table at the end of Chapter 6, 4. Obtain the heat flow equation (1.3) as follows: The quantity of heat Q flows surface is proportional to the normal component of the (negative) temperature gradient, (-VT) - n. Compare Chapter 6, equation (10.4), and apply the discussion of flow of water given there to the flow of heat. Thus show that the rate of gain of heat per unit volume per unit time is proportional to V « WT. Bur 87/21 is proportional to this gain in heat; thus show that T satisfies (1.3) 2. LAPLACE’S EQUATION; STEADY-STATE TEMPERATURE IN A RECTANGULAR PLATE We want to solve the following problem: A long rectangular metal plate has its two long sides and the far end at 0° and the base at 100° (Figure 2.1). The width of the plate is 10 cm. Find the steady-state temperature distribution inside the plate. (This problem is mathematically identical to the problem of finding the electrostatic potential in the region 0 0, if the given temperatures are replaced by potentials— see, for example, Jackson, p. 72.) To simplify the problem, we shall assume at first that the plate is so long compared to its width that We may make the mathematical approximation that it extends to infinity in the y direction. It is then called a semi-infinite plate. This is a good assump- tion if we are interested in temperatures not too near the far end. The temperature 7 satisfies Laplace’s equation inside the plate where there are no sources of heat, that is, or oF oT oT ax? vere coe We have written V? in rectangular coordinates be- cause the boundary of the plate is rectangular and FIGURE 2.1

You might also like