The terminology 'conformal mapping' should have a familiar sound. In 1569 the Flemish cartographer Gerardus Mercator (1512--1594) devised a cylindrical map projection that preserves angles. The Mercator projection is still used today for world maps. Another map projection known to the ancient Greeks is the stereographic projection. It is also conformal (i.e., angle preserving), and we introduced it in Section 2.5 when we defined the Riemann sphere. In complex analysis a function preserves angles if and only if it is analytic or anti-analytic (i.e., the conjugate of an analytic function). A significant result, known as Riemann mapping theorem, states that any simply connected domain (other than the entire complex plane) can be mapped conformally onto the unit disk. Let f(z) be an analytic function in the domain D, and let be a point in D. If , then we can express f(z) in the form (10-1) , where . If z is near , then the transformation has the linear approximation , where . Because when , for points near the transformation has an effect much like the linear mapping . The effect of the linear mapping S is a rotation of the plane through the angle , followed by a magnification by the factor , followed by a rigid translation by the vector . Consequently, the mapping preserves angles at the point . We now show that the mapping also preserves angles at . For a smooth curve that passes through the point , we use the notation , for . A vector tangent to C at the point is given by , where the complex number is expressed as a vector. The image of C under the mapping is the curve K in the w plane given by the formula . We can use the chain rule to show that a vector tangent to K at the point is given by . Therefore the effect of the transformation is to rotate the angle of inclination of the tangent vector at through the angle to obtain the angle of inclination of the tangent vector at . This situation is illustrated in Figure 10.1. A mapping is said to be angle preserving, or conformal at , if it preserves angles between oriented curves in magnitude as well as in orientation. Theorem 10.1 shows where a mapping by an analytic function is conformal. Theorem 10.1 (Conformal Mapping). Let f(z) be an analytic function in the domain D, and let be a point in D. If , then f(z) is conformal at . Example 10.1. Show that the mapping is conformal at the points , , and , and determine the angle of rotation given by at the given points. Solution. Because , we conclude that the mapping is conformal at all points except , where n is an integer. Let f(z) be a nonconstant analytic function. If , then is called a critical point of f(z), and the mapping is not conformal at . The next result shows what happens at a critical point. Theorem 10.2. Let f(z) be analytic at the point . If and , then the mapping magnifies angles at the vertex by the factor k, as shown in Figure 10.3. Example 10.2. Show that the mapping maps the unit square onto the region in the upper half-plane , which lies under the parabolas and as shown in Figure 10.4. Solution. The derivative is , and we conclude that the mapping is conformal for all . Note that the right angles at the vertices , , and are mapped onto right angles at the vertices , , and , respectively. At the point , we have and . Hence angles at the vertex are magnified by the factor . In particular, the right angle at is mapped onto the straight angle at . Another property of a conformal mapping is obtained by considering the modulus of . If is near , we can use the equation and neglect the term . We then have the approximation (10-9) . From Equation (10-9), the distance between the images of the points and given approximately by . Therefore we say that the transformation changes small distances near by the scale factor . For example, the scale factor of the transformation near the point is . We also need to say a few things about the inverse transformation of a conformal mapping near a point , where . A complete justification of the following assertions relies on theorems studied in advanced calculus. (See, for instance, R. Creighton Buck, Advanced Calculus, 3rd ed. (New York, McGraw-Hill), pp. 358-361, 1978.) The mapping in Equations (10-10) represents a transformation from the xy plane into the uv plane, and the Jacobian determinant, , is defined by (10-11) . The transformation in Equations (10-10) has a local inverse, provided . Expanding Equation (10-11) and using the Cauchy--Riemann equations, we obtain Consequently, Equations (10-11) and (10-11) imply that a local inverse exists in a neighborhood of the point . The derivative of g(w) at is given by the familiar expression Source.


Яндекс.Метрика Рейтинг@Mail.ru Free Web Counter
page counter
Last Modified: March 28, 2014 @ 12:00 am