Math 308, Sections 301, 302, Summer 2008 Review Before Test II. 06/23/2008 Chapter 4. Linear Second Order Equations Section 4.2 Linear Differential Operators Definition A linear second order equation is an equation that can be written in the form a2 (x) d 2y dy + a0 (x)y = b(x). + a1 (x) dx 2 dx We will assume that a0 (x), a1 (x), a2 (x), b(x) are continuous functions of x on an interval I . When a0 , a1 , a2 , b are constants, we say the equation has constant coefficients, otherwise it has variable coefficients. We are interested in those linear equations for which a2 (x) is never zero on I . In that case we can rewrite (??) in the standard form dy d 2y + p(x) + q(x)y = g (x), 2 dx dx where p(x) = a1 (x)/a2 (x), q(x) = a0 (x)/a2 (x) and g (x) = b(x)/a2 (x) are continuous on I . (1) Associated with equation (1) is the equation y ′′ + p(x)y ′ + q(x)y = 0, (2) which is obtained from (1) by replacing g (x) with zero. We say that equation (1) is a nonhomogeneous equation and that (2) is the corresponding homogeneous equation. Given any function y with a continuous second derivative on the interval I , then y ′′ (x) + p(x)y ′ (x) + q(x)y (x) generates a new function L[y ] = y ′′ (x) + p(x)y ′ (x) + q(x)y (x). What we have done is to associate with each function y the function L[y ]. This function L is defined on a set of functions. Its domain is the collection of functions with continuous second derivatives. We will call this mappings operators. Because L involves differentiation, we refer to L as a differential operator. The image of a function t under the operator L is the function L[y ]. If we want to evaluate this image function at some point x, we write L[y ](x). Lemma Let L[x] = y ′′ (x) + p(x)y ′ (x) + q(x)y (x). If y , y1 , and y2 are any twice-differentiable functions on the interval I and if c is any constant, then (a)L[y1 + y2 ] = L[y1 ] + L[y2 ], (b)L[cy ] = cL[y ]. Any operator that satisfied satisfies both properties from Lemma for any constant c and any functions y , y1 , and y2 in its domain is called a linear operator and we can say that ”L preserves linear combination”. If properties fails to hold, the operator is nonlinear. Theorem (linear combination of solutions). Let y1 and y2 be solutions to the homogeneous equation (2). Then any linear combination C1 y1 + C2 y2 of y1 and y2 , where C1 and C2 are constants, is also the solution to (2). Theorem (existence and uniqueness of solution). Suppose p(x), q(x), and g (x) are continuous on some interval (a, b) that contains the point x0 . Then, for any choice of initial values y0 , y1 there exists a unique solution y (x) on the whole interval (a, b) to the initial value problem y ′′ + p(x)y ′ + q(x)y = g (x), y (x0 ) = y0 , y ′ (x0 ) = y1 . Section 4.3. Fundamental solutions of homogeneous equations Theorem Let y1 and y2 denote two solutions on I to y ′′ + p(x)y ′ + q(x)y = 0 where p(x) and q(x) are continuous on I . Suppose at some point x0 ∈ I these solutions satisfy y1 (x0 )y2′ (x0 ) − y1′ (x0 )y2 (x0 ) 6= 0. Then every solution to (2) on I can be expressed in the form y (x) = C1 y1 (x) + C2 y2 (x) where C1 and C2 are constants. Definition For any two differentiable functions y1 and y2 , the determinant y1 (x) y2 (x) = y1 (x)y ′ (x) − y ′ (x)y2 (x) W [y1 , y2 ](x) = ′ 1 2 y1 (x) y2′ (x) is called the Wronskian of y1 and y2 . (3) Procedure for solving homogeneous equations To determine all solutions to y ′′ + p(x)y ′ + q(x)y = 0: (a) Find two solutions y1 and y2 that constitute a fundamental solution set. (b) Form the linear combination y (x) = C1 y1 (x) + C2 y2 (x), to obtain the general solution. Definition Two functions y1 and y2 are said to be linearly dependent on I if there exist constants C1 and C2 , not both zero, such that C1 y1 (x) + C2 y2 (x) = 0 for all x ∈ I . If two functions are not linearly dependent, they are said to be linearly independent. Theorem Let y1 and y2 be solutions to the equation y ′′ + p(x)y ′ + q(x)y = 0 on I , and let x0 ∈ I . Then y1 and y2 are linearly dependent on I if and only if the constant vectors y1 (x0 ) y2 (x0 ) and y1′ (x0 ) y2′ (x0 ) are linearly dependent. Corollary If y1 and y2 are solutions to y ′′ + p(x)y ′ + q(x)y = 0 on I , then the following statements are equivalent: (i) {y1 , y2 } is a fundamental solution set on I . (ii) y1 and y2 are linearly independent on I . (iii) W [y1 , y2 ] is never zero on I . Another representation of the Wronskian for two solutions y1 (x) and y2 (x) to the equation y ′′ + py ′ + qy = 0 on (a, b) is Abel’s identity: Z x W [y1 , y2 ](x) = C exp − p(t)dt , x0 where x0 ∈ (a, b) and C is a constant that depends on y1 and y2 . Section 4.4 Reduction of Order A general solution to a linear second order homogeneous equation is given by a linear combination of two linearly independent solutions. Let f be nontrivial solution to equation y ′′ + p(x)y ′ + q(x)y = 0. Let’s try to find solution of the form y (x) = v (x)f (x), where v (x) is an unknown function. y ′ = v ′ f + vf ′ , y ′′ = v ′′ f + 2v ′ f ′ + vf ′′ . Substituting these expression into equation gives v ′′ f + 2v ′ f ′ + vf ′′ + p(v ′ f + vf ′ ) + qvf = 0 or (f ′′ + pf ′ + qf )v + fv ′′ + (2f ′ + pf )v ′ = 0. Since f is the solution, fv ′′ + (2f ′ + pf )v ′ = 0. Let’s w (x) = v ′ (x), then we have fw ′ + (2f ′ + pf )w = 0, Separating the variables and integrating gives dw f′ = (−2 − p)dx, w f Z ′ Z Z f dw = −2 dx − pdx, w f w = v′ = ± e− R pdx , f2 which holds on any interval where f (x) 6= 0. Z − R p(x)dx e v =± . [f (x)]2 Section 4.5 Homogeneous Linear Equations with Constant Coefficients For the equation ay ′′ + by ′ + cy = 0 where a, b, c are constants, we try to find a solution of the form y = erx . If we substitute y = erx into (??), we obtain ar 2 + br + c = 0 This quadratic equation is called the auxiliary equation or characteristic equation. y = erx is a solution to ay ′′ + by ′ + cy = 0 if an only if r satisfies ar 2 + br + c = 0. The auxiliary equation has two roots √ √ −b − b 2 − 4ac −b + b 2 − 4ac , r2 = r1 = 2a 2a When b2 − 4ac > 0, then r1 , r2 ∈ R and r1 6= r2 . So, y1 (x) = er1 x and y2 (x) = er2 x are two LI solutions to ay ′′ + by ′ + cy = 0 and y (x) = c1 er1 x + c2 er2 x is the general solution to ay ′′ + by ′ + cy = 0. If b2 − 4ac = 0, then the auxiliary equation has a repeated root b r ∈ R, r = − 2a . In this case, y1 (x) = erx and y2 (x) = xerx are two LI solutions to ay ′′ + by ′ + cy = 0 and y (x) = c1 erx + c2 xerx = (c1 + c2 x)erx is the general solution to ay ′′ + by ′ + cy = 0. If b2 − 4ac < 0, then the equation (2) has two complex conjugate roots √ b ac − b 2 r1 = − + i = α + i β, 2a 2a √ b ac − b 2 r2 = − − i = α − i β = r¯1 , 2a 2a √ 2 b here i 2 = −1, α = − 2a , β = ac−b , α, β ∈ R. 2a Then two linearly independent solutions to ay ′′ + by ′ + cy = 0 are eαx cos βx and eαx sin βx and a general solution is y (x) = c1 eαx cos βx + c2 eαx sin βx, where c1 and c2 are arbitrary constants. Cauchy-Euler Equations A linear second order differential equation that can be expressed in the form d 2y dy ax 2 2 + bx + cy = 0 dx dx where a, b, and c are constants is called a homogeneous Cauchy-Euler equation. To solve a homogeneous Cauchy-Euler equation, we make the substitution x(t) = et . Substituting x(t) = et into an equation gives 2 dy dy d y − + cy = 0, +b a 2 dt dt dt dy d 2y + cy = 0, + (b − a) 2 dt dt which is linear second order differential equation with constant coefficients. a Section 4.7 Superposition and nonhomogeneous equations Let L be a linear differential operator L[y ](x) = y ′′ + p(x)y ′ + q(x)y . L maps function y into the function y ′′ + p(x)y ′ + q(x)y . Suppose that y1 (x) an y2 (x) are two functions, and L[y1 ](x) = y1′′ (x) + p(x)y1′ (x) + q(x)y1 (x) = g1 (x), L[y2 ](x) = y2′′ (x) + p(x)y2′ (x) + q(x)y2 (x) = g2 (x). L[c1 y1 + c2 y2 ](x) = c1 L[y1 ](x) + c2 L[y2 ](x) = c1 g1 (x) + c2 g2 (x). Theorem 1 (superposition principle) Let y1 be a solution to a differential equation L[y1 ](x) = y1′′ (x) + p(x)y1′ (x) + q(x)y1 (x) = g1 (x) and let y2 be a solution to a differential equation L[y2 ](x) = y2′′ (x) + p(x)y2′ (x) + q(x)y2 (x) = g2 (x), where L is a linear differential operator. Then for any constant c1 and c2 , the function c1 y1 + c2 y2 is the solution to the differential equation L[y ](x) = c1 g1 (x) + c2 g2 (x). Theorem 2 (representation of solutions for nonhomogeneous equations). Let yp (x) be particular solution to the nonhomogeneous equation y ′′ + p(x)y ′ + q(x)y = g (x) on the interval (a, b) and let y1 (x) and y2 (x) be linearly independent solutions on (a, b) of the corresponding homogeneous equation y ′′ + p(x)y ′ + q(x)y = 0. Procedure for solving nonhomogeneous equations. To solve y ′′ + p(x)y ′ + q(x)y = g (x): (a) Determine the general solution c1 y1 (x) + c2 y2 (x) of the corresponding homogeneous equation. (b) Find the particular solution yp (x) of the given nonhomogeneous equation. (c) Form the sum of the particular solution and a general solution to the homogeneous equation y (x) = yp (x) + c1 y1 (x) + c2 y2 (x), to obtain the general solution to the given equation. Section 4.8 Method of Undetermined Coefficients In this section, we give a simple procedure for finding a particular solution to the equation ay ′′ + by ′ + cy = g (x), (4) when the nonhomogeneous term g (x) is of a special form g (x) = eαx (Pm1 (x) cos βx + Qm2 (x) sin βx), where Pm1 (x) = p0 x m1 + p1 x m1 −1 + p2 x m1 −2 + . . . + pm1 −1 x + pm1 is a polynomial of degree m1 and Qm2 (x) = q0 x m2 + q1 x m2 −1 + q2 x m2 −2 + . . . + qm2 −1 x + qm2 is a polynomial of degree m2 , α, β ∈ R. To apply the method of undetermined coefficients, we first have to solve the auxiliary equation for the corresponding homogeneous equation ar 2 + br + c = 0 Particular solutions to ay ′′ + by ′ + cy = g (x) Type g (x) I II III IV V VI VII xk x k−1 p0 + p1 + . . . + pk deαx eαx (p0 x k + ... + pk ) d cos βx + f sin βx Pm1 (x) cos βx+ +Qm2 (x) sin βx eαx (d cos βx + f sin βx) eαx [Pm1 (x) cos βx+ +Qm2 (x) sin βx] yp (x) x s (Ax k + Bx k−1 + ... + F ) x s Aeαx x s eαx (Ax k + Bx k−1 + . . . + F ) x s (A cos βx + B sin βx) x s {(A0 x m + ... + Am ) cos βx+ +(B0 x m + ... + Bm ) sin βx} x s eαx (A cos βx + B sin βx) x s eαx [(A0 x m + ... + Am ) cos βx+ +(B0 x m + ... + Bm ) sin βx] In this table s = 0, when α + i β is not a root to the auxiliary equation, s = 1, when α + i β is one of two roots to the auxiliary equation, and s = 2, when β = 0 and α is a repeated root to the auxiliary equation; m = max{m1 , m2 }, Pm1 (x) = p0 x m1 + p1 x m1 −1 + p2 x m1 −2 + . . . + pm1 −1 x + pm1 , Qm2 (x) = q0 x m2 + q1 x m2 −1 + q2 x m2 −2 + . . . + qm2 −1 x + qm2 Section 4.9 Variation of Parameters Consider the nonhomogeneous linear second order differential equation y ′′ + p(x)y ′ + q(x)y = g (x). (5) Let {y1 (x), y2 (x)} be a fundamental solution set to the corresponding homogeneous equation y ′′ + p(x)y ′ + q(x)y = 0. The general solution to this homogeneous equation is yh (x) = c1 y1 (x) + c2 y2 (x), where c1 and c2 are constants. To find a particular solution to (1) we assume that c1 = c1 (x) and c2 = c2 (x) are functions of x and we seek a particular solution yp (x) in form yp (x) = c1 (x)y1 (x) + c2 (x)y2 (x). Let’s substitute yp (x), yp′ (x), and yp′′ (x) into (1). yp′ (x) = c1′ (x)y1 (x) + c2′ (x)y2 (x) + c1 (x)y1′ (x) + c2 (x)y2′ (x). Assume that c1′ (x)y1 (x) + c2′ (x)y2 (x) = 0 , then yp′ (x) = c1 (x)y1′ (x) + c2 (x)y2′ (x). yp′′ (x) = c1′ (x)y1′ (x) + c2′ (x)y2′ (x) + c1 (x)y1′′ (x) + c2 (x)y2′′ (x). yp′′ + p(x)yp′ + q(x)yp = c1′ (x)y1′ (x) + c2′ (x)y2′ (x) = g (x). We can find c1 (x) and c2 (x) solving the system ′ c1 (x)y1 (x) + c2′ (x)y2 (x) = 0 c1′ (x)y1′ (x) + c2′ (x)y2′ (x) = g (x) for c1′ (x) and c2′ (x). Z Z −g (x)y2 (x) g (x)y1 (x) c1 (x) = dx , c2 (x) = dx W [y1 , y2 ](x) W [y1 , y2 ](x) Section 4.11 A closer look at free mechanical vibrations A damped mass-spring oscillator consists of a mass m attached to a spring fixed at one end. Model for the motion of the mass is expressed by the initial value problem my ′′ + by ′ + ky = Fexternal , y (0) = y0 , y ′ (0) = v0 , where m is a mass, b is the damping coefficient, k is the stiffness. Let’s Fexternal = 0 Undamped free case: b = 0 The equation reduces to my ′′ + ky = 0| : m y ′′ + ω 2 y = 0 where ω = q k m. The solution of this equation is y (t) = C1 cos ωt + C2 sin ωt y (t) = A sin(ωt + φ), where A = q C12 + C22 , tan φ = The motion is periodic with period 2π/ω natural frequency ω/2π angular frequency ω amplitude A. C1 C2 . Underdamped or oscillatory motion (b 2 < 4mk) The solution to the equation my ′′ + by ′ + ky = 0 is y (t) = e αt (C1 cos βt + C2 sin βt) = Ae αt sin(βt + φ), where q b C1 1 p 4mk − b 2 , A = C12 + C22 , tan φ = α=− ,β= . 2m 2m C2 The solution y (t) varies between −Ae αt and Ae αt with 4πm 2π =√ and quasifrequency 1/P. quasiperiod P = β 4mk − b 2 y (t) → 0 as t → ∞. An exponential factor Ae αt is called a damping factor. The system is called underdamped because there is not enough damping present (b is too small) to prevent the system from oscillating. Overdamped motion (b 2 > 4mk) The solution to the equation my ′′ + by ′ + ky = 0 is y (t) = c1 er1 t + c2 er2 t , b b 1 p 1 p where r1 = − 4mk − b 2 , r2 = − 4mk − b 2 + − 2m 2m 2m 2m r2 < 0, and since b 2 > b 2 − 4mk, r1 < 0. y (t) → 0 as t → ∞. Critically damped motion (b 2 = 4mk) The solution to the equation my ′′ + by ′ + ky = 0 is b y (t) = (c1 + c2 t)e− 2m t . y (t) → 0 as t → ∞.