PHYS 201 Mathematical Physics, Fall 2015, Homework 4 Due date: Tuesday, October 27th, 2015 1. In this exercise, we will prove the existence of a single-valued and analytic local inverse function for any analytic function whose derivative is non-zero. i. Prove that if w = f (z) is analytic at z0 , with w0 = f (z0 ) and f 0 (z0 ) 6= 0, then there exists a contour C around z0 such that f (z) does not take the value w0 on or within C except at z0 . This shows that |f (z) − w0 | has a lower bound, say m, on C. (Hint: use the identity theorem, which states that if f is analytic, we have f ≡ 0 in a domain D if and only if you can construct a sequence an → a within D such that f (an ) = 0 for all n and a lies in D.) ii. Next, consider w such that |w − w0 | < m. Show that for any such w, say w1 , there is only one point inside C such that w1 = f (z1 ). Conclude the proof. (Hint: use RoucheĢ’s theorem) 2. Most of the special functions in mathematical physics can be generated by a generating function of the form X g(t, x) = fn (x)tn n Show that we can give an integral representation of the special function fn (x) as I 1 g(t, x)t−n−1 dt fn (x) = 2πi where the contour encloses the origin and no other singular points. The generating function for Bessel functions Jn (x) is given by gB (t, x) = e(x/2)(t−1/t) . Expand gB (t, x) as a Laurent series near the origin and find the first three terms of Jn (x) by computing the residue. 3. Evaluate using rectangular contours (or otherwise): R∞ 2 i. −∞ e−αx eikx dx for real α, k. R ∞ eax ii. −∞ 1+e x dx for 0 < a < 1. By making an appropriate substitution, show that this integral is the beta function B(a, 1 − a) = Γ(a)Γ(1 − a). 4. Show that π cot πz has simple poles at integer values of z. Consider the integral I 1 f (z)π cot(πz)dz 2πi CN 1 with f (z) = 1/z 2 and CN a rectangular contour with diagonally opposite vertices −N − 1/2 andPN + 1/2. Argue why the integral goes to zero and consequently show π2 1 that the sum ∞ n=1 n2 = 6 by computing the residue at 0. 5. Evaluate the integral π Z 0 cos nθ dθ 1 − 2a cos θ + a2 R −iz n−1 dz iθ for a real and > 1. (Hint: Consider the real part of 1−a(z+z −1 )+a2 where z = e . Many integrals involving trigonometric integrands can be solved this way.) 6. Show that, for α > 0, Z ∞ 0 t sin αt π dt = e−α 2 1+t 2 7. The function f (z) = (1 − z 2 )1/2 can be made single-valued using cuts running along the real axis for |x| > 1. Using these cuts and a suitable contour, evaluate the integral Z ∞ dx 2 x(x − 1)1/2 1 Verify your answer by substituting x = sec θ. 8. Use a semicircular contour in the upper half plane (with a bump at the origin) to compute the integral Z ∞ (ln x)2 dx 1 + x2 0 and deduce, as a byproduct of your calculation, that Z ∞ ln x dx = 0 1 + x2 0 9. Use a keyhole contour to show that Z ∞ √ ln x dx = − 2π 2 3/4 x (1 + x) 0 10. Evaluate the integral 1 x1/2 (1 − x)1/2 dx 2−x 0 (Hint: Apply Cauchy’s Integral Formula to the entire complex plane excluding the branch cut between 0 and 1. Be careful about your choice of branch - verify that the branch you’ve chosen is continuous across the negative real line and the positive real line > 1. To compute the residue at infinity, use the formula Res(f (z)) at z = ∞ is equal to Res(−z −2 f (1/z)) at z = 0.) Z 2