Problem with double derivative

Hello, I have a simulink model where a signal is integrated twice, then some calculations are done and the output is derivated twice and fed back into the integrator. Now I got an error at the first integrator, that the signal is not finite. With only one derivative block it has no problem, therefore I think the second one produces a very high peak. With a delay block it is also working but I am not sure if that is the right way to do it. So do you have ideas how to solve this problem ? Greetings Lucas

[Ameer Hamza]

Can you attach an example model which is causing the problem? Also if the calculations between double integration and differentiation are linear, then it might be better to avoid integration and differentiation at all.

[Ameer Hamza]

@Lucas Maiß's answer moved here

What I do is that I have a clutch and I use the angular position and subtract the portion that is caused by the stiffness of the clutch. Afterwards I use my position to calculate the friction and inertia an subtract it from the input of the first integrator.

[Ameer Hamza]

I don't completely understand the issue, because the simulation run fine even if I make the delay equal to 0, but removing delay block brings a singularity in the model. You might want to look at this answer: https://www.mathworks.com/matlabcentral/answers/99084-why-am-i-receiving-error-messages-about-singularities-in-my-simulink-model.