MAE 5735, Fall 2012

Old news

12/11: Sign up times for Extended Final project are here.

12/5: Candidate Final exam problems, written by some of you, are posted. At least one actual final exam problem will be taken from this bunch with, perhaps, slight modifiction.

12/5: Office Hours in study week: Mon Dec 3, 2:30-3:30, Wed Dec 5 10-12, Fri Dec 7, 10-12.

12/3: Project panic? Talk to me at your signed up time on 12/4. Once I hear your reason, I can give an extension to Dec 14 for both the project and the files. After I give the say so, you can schedule a time for Dec 14. If you are not available then, or can't do the project by then tell me and you can get a course incomplete. Course incompletes are a pain for you and I both, especially for you, so I don't recommend that.

12/3: Anonymous Course Survey. Please do this survey before Wed December 5 at 6 PM. Andy will get a list of who and who has not done the survey (but not who did which one). The yet-to-be-determined grading formula will have one point added for the people who do the survey. Most important are comments about what should be kept or changed for this course in the future. If you haven't done it yet, now would be a good time.

12/2: Clarification. Final project due midnight at the end of Tuesday Dec 4.

12/1: The JACOBIAN command in Matlab. In class on Wed Nov 28 I mentioned that this could be used to extract a "mass matrix" when deriving equations of motion symbollically. This is included as an option double-pendulum MATLAB example that is posted (corrected on 12/4). This approach removes the symbolic algebra demanded of the SOLVE command (a huge reduction of computer effort if there are many degrees of freedom). A student just wrote to me: "Using the Jacobian method you suggested makes it possible to solve systems of 10+ links with the AMB method. The resulting solutions are also exactly equivalent to the DAE solutions to numerical precision..." Previously he/she had been limited to 3 links because the Matlab SOLVE command had been overwhelmed.

11/30: Notes of last course lecture by R. Rand (on Friday Nov 30).

11/28: What will be on the Final Exam? Things from lecture, readings, HWs and prerequisite courses. Some gaurantees, perhaps overlapping, at least: One question based on a student-submitted question. One question of a dynamics Q-exam type. One Dynamics question. One vibrations question. One Matlab-related question. There will be no extra time. There will be about 5 questions in total. It will be possible to do all the algebra for complete credit within 2.5 hours (although there is no extra time, I will try not to make this a speed-based test). No books. No notes. No calculators.

11/28: Sign up now works (for final project demo, below).

11/27: Final homework, probs 36-38, due Mond Dec 3. Hand in to Joe Rogan in Kimball 210 (or is it 212?, at the North end of hallway).

11/26: Office Hours in study week: Mon Dec 3, 2:30-3:30, Wed Dec 5 10-12, Fri Dec 7, 10-12. (Office hrs on Tues Nov 27, as usual)

11/26: Sign up for FINAL PROJECT DEMONSTRATION (this works now! - 11/28) Times available all day Tuesday December 4. Your presentation should be 5 minutes long.
Edit the google doc sign up sheet linked above. EVERYONE should sign up, no matter the status of the project. If you need an extension until January, we can discuss this at your signed up time.

11/25: SPECIAL PROBLEM-SOLVING OFFICE HOURS: Thurston 201, Monday Nov 26, 2012, 7-9 PM.
You will do a problem or two on the board, in pairs, with my guidance. Iterating until you have a good clear solution. Regular Monday office hours cancelled.

11/25: No homework due this week. Final homework due Monday Dec 3. To be posted by Tuesday Nov 27 (done).

11/19: Prelim 2 solution now posted. You can get your grade improved, on any problem, to (old score + 20)/2 if you redo the problem perfectly. Hand in correction on Wednesday Nov 28 (not Monday).

11/16: We are now essentially done with the multi-DOF chapter 4 of Inman as well as section at the end of the book on trusses. You should be able to read and understand almost all of that chapter, or an equivalent chapter of another book, on modes, modal coordinates, modal damping, modal resonance, etc.

11/16: SPECIAL PROBLEM-SOLVING OFFICE HOURS: More than 8 people promised to come. Thurston 201, Monday Nov 19, 2012.
You will do a problem or two on the board, in pairs, with my guidance. Iterating until you have a good clear solution.

11/15: HW due Wed Nov 21 is posted. Problems 34 & 35. Spend as much time on problem 35 as you have time for.

11/8: HW due Wed Nov 1, probs 32 and 33, are posted below (re-posting from Nov 7).

11/7: Emergency office hours: Thursday Nov 8, 2-4 PM, in Thurston 102

11/7: Solutions to HW 25-28 are posted below.

11/2: Truss vibrations MATLAB code from lecture from 11/2 is posted. You should learn to play with it, and then understand how it works. You can use it to check your homework due Wed Nov 7.

11/2: PRELIM II on Thursday Nov 8:

-Covers through the HW due on Nov 7 and the lecture on Nov 2. It is "cumulative" and _WILL_HAVE_ some basic mechanics of the type we have been doing since Aug 22. Think of it as a final for the course, up to Nov 2. Perhaps with emphasis on more recent material.

-Need to take prelim early? Let's meet after class on Monday and pick a time Wed or Thursday that is good for all.

-Want review? Tell Andy by email, soon, any topics ,from any part of the semester, that you would like reviewed. These may be covered on Monday or Wednesday in lecture.

11/2: Grading rubrick: The grading on the prelim 1 was neither additive nor subtractive. Rather, I told our great grader Pat to try to make the grade x so the p=x/25 would be the probability p, as best he could estimate or guess, that you would get the problem right if you were given effecively infinite time (say, 2 days), but no external resources. In that way of grading, a good free body diagram, good and relevant basic equations, dimensionally consistent results, and good consistency checks are much more convincing than lots of good algebra.

11/2: Problems 28-31, due Wed Nov 7, now posted (not 27-29, as incorrectly posted on 11/1).

11/1: Coming guest lecture: Nov 30 Rand. (updated from Oct 25)

10/25: Homework due Wed Oct 31, problems 25-28, is posted below.

10/24 (updated from 10/13): Final computation project. This is an extension of the double pendulum homework. This project is due December 3. The minimal version is to simulate and animate both a triple pendulum and also a 4-bar linkage. In both cases the equations of motion should be found two different ways. In both problems the numerical solutions should be checked as many ways as possible. Optional extras are a) to simulate and animate more complicated mechanisms of your choice (e.g., 4,5, n link pendulum or closed kinematic loop) and b) to find periodic motions.

10/24: In course syllabus/notes there was an error in the formula for omega_max, now corrected.

10/24: One office hour added on Mondays, 2:30-3:30.

10/18: HW due Wed Oct 24 is now posted below (problems 20-24).

10/17: Prelim 1 and solution are posted, below and here: Prelim 1,   solution. Each problem will be graded out of 25. You can redo any of the problems at home and resubmit them and get 20 points for the redo if you do it perfectly. Your recorded grade will be the average of your original grade and your redo grade. This will be due a few days after we return your graded prelim to you.

10/17: (updated from 10/15) Class Challenge. Find Simulation Software.Find low cost software that runs on Macs, PCs, smart-phones or are web-based that do dynamic simulations easily. A program used to exist, called WORKING MODEL, that was good for such. But it is barely available any more. High cost simulators associated with, say, Auto-Cad, Adams, etc. are not so easy to use. Ideally one would like to be able to set up and animate a double pendulum in about a minute. Does an open source program like ODE (open Dynamics Engine) have a good simple GUI (Graphical User Interface). Some Matlab toolbox? Tell me anything you find. Yes, you are learning how to write such software, or at least learning the core ideas. But it would be good to have a platform for demonstrating various simple machines quickly.

One source for animations:

Ideas are coming in. Candidate softwares so far (Try them. Any good?): 1) Algodoo, 2) Sim Mechanics

10/15: CLASSROOM CHANGED TO PHILLIPS 403. It's the room we used to be in. Starting Wednesday Oct 17.

10/13: No hand-in homework due this coming week. Please read Inman chapter 1, and Ruina/Pratap Sections 10.1-2, and review the analytic formulas and methods therin. Next week you will be asked to generate synthetic data by simulating a forced damped oscillator, you will then compare the analytic methods to numerical measurements of your synthetic data. Continue also, to the extent that you need to, to review or develop skills for problems of the type that are in Ruina/Pratap and/or old Q-exam problems (see below).

10/13 (repost from 10/1 and 10/10) Midsemester survey: Please Click here. You can do this now and you can do it again if you get new thoughts. Survey closes on Oct 13 (TODAY, extension to noon Sunday Oct 14). The Engineering College is also doing a different survey. Please do both. Thanks.

10/10: Prelim 1 information (partial repost from 9/25): Thursday Oct 11 7:30-9PM (well, ok, until 10:30 if you like), Kimball B11. The prelim will be do-able in 90 minutes, but I will give you 3 hours. If you can do all HW problems and ones very similar to them, and explain each step, and all lecture theory and examples, and can reproduce all the reading material, then you should do very well. No notes, no calculators. You should know all needed Matlab syntax (be able to write or read relevant code), but minor syntax errors will have only minor penalty. Covers through the HW handed in on Oct 10.

10/6 About the HW to do the double pendulum with Lagrange Equations.If you find the hand algebra too long and hard to debug. And if you have trouble mastering the computer algebra (even with the program posted), there are a few ways to bail out. 1) Only do point masses (a "simple" double pendulum), 2) Only do part of the algebra. 3) Don't solve explicitly for the angular accelerations (leave them mixed together). I would, however, like you to, somehow, compare the solutions of double-pendulum ODEs that were obtained two different ways. So you might instead, for full credit, set up the DAEs instead.

10/6 Solutions to HWs 15-16 posted below.

10/2 Solutions to HWs 11-14 posted below.

10/1 Both homework problems due Wed Oct 10 are now posted below.

9/30 About the braking car, here are three questions and answers: 1) Is there friction at the locked wheels? Yes, they become robber blocks sliding with friction on the road; 2) Is the engine exerting any force on the car? No, the engine is off and in neutral; 3) At what angle is the steering wheel locked? Straight ahead. This is now clarified in the HW, so if you download the HW on Monday or later, this will already be clear.

9/30 Three Matlab samples posted (below): controlling error in ODE23 and ODE45; "events" (ending integration when, say, the canonball hits the ground); and an example of 2D animation.

9/30 2:40 PM. HW error. If you downloaded the homeowork after this instant, skip this message. Otherwise, read on. 1) The gold standard for he rate of change of angular momentum about point C is

not with acceleration of points i with respect to C. See column (d) of Ruina/Pratap page 1046 (online version), if no subscript is given it is implicitly 0. 2) some hints have been added.

9/29 Final exam is Monday Dec 10 from 9:00 - 11:30 AM. Room not yet known.

9/28 Computer work for homeworks? At a glance computer typing looks neater than most people's handwriting. But often people using computers don't draw enough pictures (e.g., FBDs for every system and subsystem used for momentum or angular momentum balance), or enough labels of graphs (which curve is which? arrows marking interesting features). Or, if you do put enough illustration on the computer, it takes too much time. Consider at least a) getting good at making quick scans of hand drawings and/or b) of marking up your computer solutions by hand to make them more clear.

9/28 Solutions posted (below) for HW though problem 9.

9/28 Please make requests for 2D dynamics problems you would like to see solved in class. Make the requests now or by Wed Oct 10. Please estimate the thoroughness you want spent (e.g., just a minute for the FBDs, a couple minutes for the general setup, or 50 minutes for full solution with simulations and animations).

9/28 (2 AM). HW due Oct 3 now posted. Extension to Oct 5 allowed if you were counting on a full 6 days to work on it. Your choice.
There will be HW over the fall break. As noted below, it will be on prelim 1. It will be posted by Wed Oct 5.

9/25: Prelim 1 information: My sense is that there isn't enough will to change the date and time. There will be an early makeup, one day early, on Wed Oct 10, from 1-4 in my lab, Kimball 306. Let me know if you plan to take that. The prelim will be do-able in 90 minutes, but I will give you 3 hours. If you can do all HW problems and ones very similar to them, and explain each step, and all lecture theory and examples, and can reproduce all the reading material, then you should do very well. No notes, no calculators. You should know all needed Matlab syntax (be able to write or read relevant code), but minor syntax errors will have only minor penalty. Covers through the HW handed in on Oct 10.

9/24: Clarifications and corrections for HW due Wed are now posted as part of HW. In particular:
            * the pendulum equation should have in it sin theta, not just theta
           * the awkward parameterization should be with y, not x, but with the assumption that x>0.
           * one problem asks for 3 2nd order equations to solve DAEs, the other for a single 2nd order ODE.

9/17: Read and master Ruina/Pratap 14.1, 14.2, 15.1, 15.2.

9/17: HW for today's (Monday's) lecture now posted.

9/17: Due to administrator ineptness, HW from Mon 9/10 was not posted until Sunday 9/16. If that was too late for you, you can delay handing that in for a week. After all delays etc, you always have at least 6 days from when HW is posted until it is due.

9/16: HW for Friday's lecture now posted, due in 10 days.

9/14: "finesse: verb - avoid or try to avoid fulfilling, answering, or performing (duties, questions, or issues)"

9/14: HW solutions will be posted under "Resources", far below. Now posted through problem 5.

9/14: Another short falling cat video, recommended by a student. I chatted with, and perhaps helped, the maker of this excellent example of computer graphics for education.

9/14: Please think about your likelihood of taking MAE 6700 in the Spring. I would/will be teaching it. On Monday I'll collect from you an anonymous vote where you express the probability that you will take the course.

9/14: Spyros writes: "The currently set date for the 1st preliminary exam is the day after you come back from Fall Break (Thursday Oct. 11). Some students want this date changed. Please vote on the dates that are convenient to you by following this link:"

9/10: Possibly interesting seminar Tuesday at 4. I'm going..

9/10: Falling cat video here. See also starting at 20:30 here. Both videos also linked on the syllabus pdf.

9/10: Old Q-exam questions posted under "Resources" far below.

9/10: HW from today's lecture is posted. Matlab code associated with the blackboard work is posted.

9/8: A goal (end) of this course is mastery of TAM 2030. It's a pre-requisite, yes. But now real mastery is demanded. Towards that end, you

* need to make sure that, within a few weeks, you can masterfully do all of the prelim and exam questions from, say, TAM 2030 in Spring 2011. You might start now, looking at those problems related to material we have already covered.

* should eventually be a master of all of the material in Ruina/Pratap (well, not all of the statics will be re-inforced here). By now you should have read Chapters 1, 2, 3, 9, 11 & 12. Even if they are review, I recommend that you read them carefully enough to shore up your weak spots (for example, do you get all 17 parts of Sample 2.1and Box 3.8 correct?)

9/8: Greenwood will be on reserve at Uris Library. (I can't find my own copy, so can't put that on informal reserve.)

9/8:  Homework from 9/7 lecture, due Sept 19, is posted.

9/7: Today's matlab code is posted. See link below under resources.

9/5: Today's HW, due Sept 12, is posted.

9/5: Pat Moran's new office hours: 5PM to 6PM in Upson 109.

* 9/3: A TA generated solution for HW 1 is posted. In the future we'd like to select HW solutions from the HW you hand in. Your goal is todo work that is good for that. This posted example and annotations should help you. Mostly, think of your HW as a self-contained docunent that a person can learn from. This person does not have the problem statement in front of them and doesn't know how to solve the problem.

* 9/1: When is homework due? Associated with every lecture are one or two homework problems. They are listed on the chronological syllabus (pdf below) at the end of the lecture description. They are posted within a day of the lecture (actually, a few hours late today). HW problems for a given lecture are due on the first Wednesday after the given lecture that is at least 6 days away after the lecture. That is, HW assigned on Mon Wed and Friday are due 9, 7 and 12 days after their associated lectures, respectively. That is a week of homework problems are assigned on Friday, Monday and Wednesday, all to be handed in a week after the Wednesday. Also, starting now, the due date of homeork is listed on the chronological syllabus.

* 9/1: HW for lecture from Friday 8/31 is posted (on Syllabus).

* 8/29: Homework for today's lecture is posted. Syllabus includes a link to the Feynman video (which the last fragment of today's lecture duplicated).

* 8/29: If you find good internet sources for any of the course material, send me a link and I will consider posting it for others.

* 8/28 Sloppy in-class board work on conservative forces clarified, perhaps, in the chronological syllabus.

* 8/28: CLASSROOM CHANGED to 253 Malott. Starting NOW (Wed Aug 29).
It's at the South end of the building, down 1.5 levels from the Barton parking lot.
Walk in the Tower Road doors labeled Math and Statistics. 253 is the first room on
the right. A 5 minute (clocked) walk from the North end of Phillips.

* 8/27: Office hours. For now, Tuesday August 28, they are
    Tuesdays 1-3 PM in Thurston 102 "The Conway Room" (big room with tables).
    Later weeks they will be 1:50-3:50.

* 8/27: Homework for todays lecture is posted.

* 8/25: About the final course grade, if you don't care about how it is done, you can
skip this message. If you do, we have two choices we will discuss in class on Wednesday August 29.
1) For large undergraduate courses I design and post a formula and then strictly grade by it. Please read this
page about such a formula and my feelings about it.
2) For small graduate courses at the end of the semester I have looked at all of the
data in hand, make up a formula to rank the students and then give a grade impressionistically.
Sometimes the rank order of grades will be different than from that of the formula's number because
of an exceptional final-exam performance, an exceptional project, etc.

In both cases, the actual distribution of As, Bs and Cs is determined at the end, impressionistically.
Some people don't like that "grading on a curve", but actually I believe ALL professors do that, no
matter what they claim. I think (2) is more accurate, in terms of having grades correlate with knowledge.
I think (1) is more comfortable for people trying to budget their time and energy. I am willing to do

* 8/24: Today's HW now posted (on the "Coverage, Syllabus " pdf, linked below.

* 8/22: The Coverage etc pdf (link below) includes HW 1 due next Wednesday.

* 6/15/2012 This WWW page will develop over the summer.

* 6/15/2012 There is no old news yet.