Why is the input signal to my suspension model settling at a non-zero value and drift away? Why does the drifting go away when I change a nonlinear damper to a linear one?

Question

Ryan Vorster 2018-8-7

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/413825-why-is-the-input-signal-to-my-suspension-model-settling-at-a-non-zero-value-and-drift-away-why-does

回答： Christoph Hahn 2018-8-7

Hello, my name is Ryan. I am a student at the University of Victoria in Canada. I have been trying to model a suspension corner of my FSAE car with Simscape blocks and I have run into an issue with how the model is behaving. I would attach the file but apparently the Simulink model filetype is not supported by this forum. I've attached a picture below for reference. The values in the model are realistic for the type of car, but the model behavior described happens with any values.

</matlabcentral/answers/uploaded_files/127942/Pictureofmodel.PNG> The Model

This is the input signal to the system, swept from 0-20hz of velocity peaking at 100mm/s. I have attached only the first 20s for clarity as to what the input is as it becomes a big yellow block much past that point.

</matlabcentral/answers/uploaded_files/127945/inputvelocity.PNG> Input Velocity

When I run the model as shown above, these are the body acceleration and body position outputs I get. The model is run using the ode23t (mod. stiff/Trapezoidal) solver.

</matlabcentral/answers/uploaded_files/127943/bodypos.PNG> Body Position

</matlabcentral/answers/uploaded_files/127944/bodyacc.PNG> Body Acceleration

As you can see, the acceleration is behaving as expected, vibrating with the input and having a magnified mode somewhere in the middle at its damped natural frequency. The position however, does not behave as expected and I need help understanding why. The next picture is the measured position of the input signal. You can see it also behaves in a similar manner.

</matlabcentral/answers/uploaded_files/127946/inputpos.PNG> Input Position

The weirdest part about the whole system is that the input position signal changes when I change the nonlinear damper to a linear damper: it doesn't drift off anymore. See picture below.

</matlabcentral/answers/uploaded_files/127947/inputposlinear.PNG> Input Position: Linear Suspension Damper

In summary, my simscape model of a suspension system is not behaving the way I expect it to and I need help understanding why and how to model the system properly.

Thanks for your help, Ryan