Frequency response of 3 degree of freedom system

Included in the ZIP file is a SIMULINK model of a 3 degree of freedom system along with the input data and appropriate gain values a, b, & c. The SIMULINK model generates both the pole/zero locations along with the corresponding BODE plots. The first 3 maps and plots I understand, but the last one I do not. I am hoping someone with extensive control system experience might elucidate what is happening.

The first map & plot pertain to the input and output of the lowest order subsystem. The 2nd pertains to the mid order system, & the 3rd pertains to the highest order system. Each of the first 3 maps show 3 pole complex pairs along with 2 complex pairs of ZEROS that correlate with the pole pairs that belong to the adjacent subsystems. For example, the 2nd map pertaining to the MID frequency subsystem shows the 3 pole pairs, and 2 ZERO pairs belonging to the LO (1st) & HI (3rd) order subsystems. Likewise, the 1st map shows ZERO pairs that belong to the 2nd & 3rd subsystem, and the 3rd map shows ZERO pairs belong to the 1st & 2nd subsystems. I hope you, the reader, can follow what I am trying to say.

The last, 4th, map shows all 3 pole pairs along with 2 ZEROS setting on the REAL axis. The input is to the 1st subsystem & the output is from the 3rd subsystem. I was expecting to only see the 3 pole pairs, and NO ZEROS. Why is this & is it possible to get a response that is not influenced by any ZEROS?

Appreciate any insight

[Gillian Rosen]

Hi Jeff, 

I understand that you are simulating a 3 DOF system and observing the frequency response, but you are getting unexpected output of your system. 

If it seems that the output you are seeing is incorrect for the model you have built, there are steps you can take to debug this issue. The following documentation links have information about how to get accurate results when working with the Control System Toolbox, which may be helpful to you: 

https://www.mathworks.com/help/releases/R2012b/control/examples/using-the-right-model-representation.html

https://www.mathworks.com/help/releases/R2012b/control/examples/preventing-state-duplication-in-system-interconnections.html

https://www.mathworks.com/help/releases/R2012b/control/examples/index.html#how-to-get-accurate-results

[Jeff]

You are misinterpreting what I am trying to say. My model output is CORRECT, but my EXPECTATIONS were not. However, it is plausible to obtain a frequency response such that NO ZEROS exist in the numerator, and all 3 poles are distinctly visible in both the BODE & pzmap plots. So far I have not been successful at obtaining this type of response. So I am seeking counsel on how to create that situation other than directly typing in a mathematical transfer function into the block diagram.

In other words, I am trying to create a physical model that generates the mathematical situation I described above.