Move Office Hours to August-2, 2-3 pm IST (tentative) | Feedback for Office Hours | Agenda for meeting

[Mechatronics School]

Hi all,

Please fill out this form - https://forms.gle/jZcvwuu2W5EgEGh37 for a survey on what time works best for the meeting and suggestions for topics for discussion.

-------------------------

Hope all is well!

Few updates:

1. I have some other commitments this Friday and therefore, will be unable to make it to the office hours this Friday. How about we tentatively move the meeting to next Friday, Aug-02, 2-3 pm IST? 
1. We can reschedule the time and day for the meetings based on what works best for people, please fill out this form - https://forms.gle/jZcvwuu2W5EgEGh37 to give your preferences and suggestions of topics for future discussions.
2. Meeting Details:
   Mechatronics Office Hours

   Friday, Aug-02, 2019 2:00 pm IST | 1 hour 

   Recurrence: every two weeks on Friday

   Meeting number: 642 098 210

   Password: MSD_2019

   Meeting link: http://bit.ly/ohMSD
2. For the next meeting, we will take up the ball-screw drive system modeling that we left off in the course. For those who are attending the office hours, please attempt working on the model yourselves and send me your solutions latest one day before the meeting, so we can discuss the various approaches. This is your task (please feel free to work in teams and indicate the team members in the document): 
1. Build a mathematical model for the transfer functions for position output to motor current input (assume we have current control implemented) and position output to disturbance force input. Also, get the transfer functions for velocity instead of position.
2. Start simple, consider everything perfectly rigid. Then add flexibility, primarily axial and torsional flexibility of the coupling--ball-screw--nut drive system.
3. Implement the transfer function model in the control toolbox and attach a few representative bode plots of the system. Play with the parameters and see what happens as you vary mass and stiffness parameters.
4. Based on this and what we learned in the course, what implications does this have on the bandwidth you can achieve in the system? How would this relate to your overall system specifications - disturbance force, motion profiles, error allowed, stiffness, bandwidth?
5. Few references that may help:
1. https://my.mech.utah.edu/~me7960/lectures/Topic4-BallscrewCalculations.pdf
   
2. https://pdfs.semanticscholar.org/2d30/53264372a2ae6d27bc07f320d8f179ef6d34.pdf
3. Please indicate any other references that you referred to.
3. Based on this work, we will look at controller design and implementation in a future office hours session. For that, it is important to translate between the series controller design we learned in the course to the form they implemented in your application. If you can send me the block diagram of overall controller implementation in your applications (showing the various cascaded loops and the tuning gains if possible, also include any feedforward times which may be there), that would be very helpful. This information will not be shared with anyone without your permission.
4. Also, we need to look at the measured bode plots. It would be very valuable if you could send me some sample open-loop, closed-loop, and sensitivity bode plots of your systems along with a schematic of the drive system indicating where the actuator is and where the sensor is located. This information will not be shared with anyone without your permission.  We need to relate our modeling to the measured bode plots and also be able to identify interesting features that we see when we measure real bode plots related to non-linearities, unmodeled behavior, measurement errors, other modes from other axes in a compound slide, etc. 
5. Based on the feedback from the survey, we will reschedule the office hours to the time and date which works for everyone. My notes from the sessions will be sent to all the attendees attending a particular session.

Thank you!

Best,

Brij