gbacsf Messages 15 Reaction score 0 Thread starter May 13, 2006 #1 What method would be used to solve this DE, it look like a Bernoulli but isn't. I'm lost. y'cosx = 1-y^2 Thanks, Gab
What method would be used to solve this DE, it look like a Bernoulli but isn't. I'm lost. y'cosx = 1-y^2 Thanks, Gab
neutrino Messages 2,093 Reaction score 2 May 13, 2006 #2 It looks like you can separate the variables.
Pseudo Statistic Messages 391 Reaction score 6 May 13, 2006 #3 Treat y' as the limiting ratio dy/dx. Your aim is to get the xs (and dx) on one side, and the ys (and dy) on the other side and integrate...
Treat y' as the limiting ratio dy/dx. Your aim is to get the xs (and dx) on one side, and the ys (and dy) on the other side and integrate...
gbacsf Messages 15 Reaction score 0 May 14, 2006 #4 I'm not sure if I can solve it that way. I did some digging and found its in the form of a Riccati eq. http://en.wikipedia.org/wiki/Riccati_equation But I'm not sure how to apply it to this problem, any help? Thanks
I'm not sure if I can solve it that way. I did some digging and found its in the form of a Riccati eq. http://en.wikipedia.org/wiki/Riccati_equation But I'm not sure how to apply it to this problem, any help? Thanks
FrogPad Messages 801 Reaction score 0 May 14, 2006 #5 Hint: [tex]\frac{dy}{dx} \,\, \frac{\cos x}{1-y^2}=1[/tex]
gbacsf Messages 15 Reaction score 0 May 14, 2006 #6 Well from that I can say that y=sin(x) is a solution. Then I get: z' +(-2*tan(x))*z = 1/cos(x) So then I solve this linear equation: (sin(x) + C)/(cos(x))^2 Last edited: May 14, 2006
Well from that I can say that y=sin(x) is a solution. Then I get: z' +(-2*tan(x))*z = 1/cos(x) So then I solve this linear equation: (sin(x) + C)/(cos(x))^2
J77 Messages 1,092 Reaction score 1 May 15, 2006 #7 The next step should be to substitute: [tex]y=\sin(u)[/tex] into the equation FrogPad gave.