COALITION MATH 152, SPRING 1998 (S. A. Fulling, assisted by Vera Rice) Day 23.T 1. The usual paper shuffling: Old homework, CAPA, revised new homework, lab sheet if available. 2. Remarks on technical part of project proposal/report (Point 4 of previous day). 3. First-order differential equations (accelerated to make up a lost day) A. Solve a (nonlinear) separable equation with initial data. (This would have been a team exercise if we had time.) y' = x^2/(1 + y^2), y(2) = 1. B. The integrated relation is hard to solve for y in this example, so use it as an occasion to demonstrate: i. Verification by implicit differentiation ii. Maple commands DEplot1 and implicitplot MINOR DISASTER: DEplot1 IS A RELEASE 3 COMMAND -- NO LONGER EXISTS IN RELEASE 4. ALSO, dfieldplot NO LONGER DOES WHAT IT DID IN RELEASE 3. MY SYNTAX EXAMPLES NEED TO BE UPDATED (TO RELEASE 5?) C. Quickly mention other ODE-related Maple commands: dfieldplot, dsolve, plot, display D. Linear equations i. Homogeneous case a. Separable! Solve as discussed above. b. This is not the same meaning of "homogeneous" as on pp. 972-974 of Stewart 3. That meaning is LESS important than this one. ("Homogeneous" also refers to materials with no spatial dependence -- still another meaning.) ii. Nonhomogeneous case, such as y' + (2/x) y = (cos x)/x^2 a. Solve the corresponding homogeneous equation. y_h = x^{-2} in example. b. Multiply equation by integrating factor mu = 1/y_h c. By magic, the left side is d(something)/dx (by the product rule in reverse). d/dx (x^2 y) = cos x in the example. d. Evaluate the antiderivative and solve for y. y = (sin x)/x^2 + C/x^2 in example. The term with the arbitrary constant is the most general solution of the homogeneous equation. E. If time permits, show prepared transparencies with example of Euler's method. Emphasize: 1. Common error of miscounting the steps. 2. Euler is NOT the method of choice for solving ODEs numerically. More complicated methods give greater accuracy for the same computation time. Euler is to Runge-Kutta or Predictor-Corrector as Riemann sums are to Simpson's rule. ALL THIS WAS ACCOMPLISHED WITH ONLY A BIT OF RUSH.