Thank you very much Robert, it is my mistake about initial conditions

Greetings

Dear Robert, thank you for your fast and competent answer, all works perfect:) Also i found the following way to compute the secon derivative:

>dtheta[2]:=subs(res,diff(theta[2](t),t));

>ddtheta[2](t):=(dtheta[2](t+0.01)-dtheta[2](t))/0.01;

>plot([ddtheta[2](t),ddtheta[3](t)],t=0..1);

Thanks again