Konstruktion 8
− Frauenkirche-Turmuhr-Modell
Sommersemester 2007
Professor: Frank Owen, http://www.calpoly.edu/~fowen/,
089/1265-3348, Büro: R4.059
Meeting time and place: Tuesdays, 1030-1130, R4.068
Current draft of final report.
Organizational tasks (08.04.07):
1. Prepare Gantt chart for project. Google "Gantt chart" and/or see my website for the Cal Poly course ME 428 under "Project schedule" to see a sample Gantt chart. Define project tasks, assign durations for project work, decide on the project milestones, etc.
2. If you can get access to a copy of SolidWorks before the FHM opens again on Wednesday, go through as many of the tutorials as possible.
3. Those with SolidWorks experience, if you are finished with your own models, get measurement data from the other groups and work on models of what they measured.
4. Someone needs to figure out how to share files by posting them to a website. We also talked about blogging, which may also be useful to us. Figure out how to share files and how to run a blog.
5. Before too long we shall want to start an outline of the final report. I wouldn't hurt if someone would start on an outline. It won't be complete by any means. But a start would be nice. It can always be changed and supplemented as we go along. Look into View/Outline in MS Word.
Recap of 23.05.07 meeting with Museum personnel:
The meeting went very well, in my opinion, especially for one that we did not coordinate any more than we did. So thanks to all for your contribution. I am amazed at the progress we've made in such a short time. Since some of the conversation was in German, I'd like to bring up important issues for us to keep in mind.
1. We are always faced with the dilemma of how detailed a model we should make. The reason to make as detailed a model as reasonably possible is for documentation reasons primarily. Take the example of fasteners and retainers, like pins, for example. These are really not necessary to make an animation work. You can put a gear on a shaft and without a pin make it not turn on the shaft, using a SolidWorks mate. In fact you could even just make the gear a part of the shaft, as if the whole thing was just machined out of one block of metal. By modeling in detail, you are actually placing a computing burden on the computer. With our myriad of parts, we may be putting an unnecessary computing burden on the computer. Another good example are the ridges in the drum that I made. That was hard to figure out. And then in the end, I asked myself, "Now why did I do that?" What that does is that it adds a whole bunch of irregular surfaces that the computer has to calculate the light and shading on as the part turns or as you revolve the part around to view it. In my opinion, that's load that's not justified. Of course all this talk is speculation at this stage. We don't know how such detail will affect the model animation until we try it. That is coming. On the other hand, the Museum is interested in how these parts are made. The documentation part of our task is important. So detailed part models have a worth of their own. One thing I mentioned to them during the meeting is that perhaps we can have both. We can have a complicated part that we make invisible when we run the entire model. Invisible parts place no load on the computer during the animation. Then we would have a simplified version that we made visible to use in the animation.
2. Another important point related to the documentation function was brought up by Herr Petzold. If we do take short-cuts or make assumptions that we don't have a basis for, we should note this. So I emphasize that every part should have an accompanying spreadsheet, used to figure out dimensions for modeling, but also used to note assumptions or unknowns. This is very important! If you are not doing this, you need to start. I propose discussing these spreadsheets, showing examples at our next meeting.
3. One of the important points I brought up as a "coming challenge" is that SolidWorks is purely a solid modeller. It does not model physical reality. So, for example, how do we set a gravity-driven clock in motion? How do we make the little flat, beam-like springs click against the pawls on the drum ratchets? How do we make rope around a drum unwind naturally under the influence of a hanging weight? I asked similar questions and then said that I think we can fake all this. How, I don't know. But my gut feeling is that we do not need a physical model to do this. We can describe the kinematics of the motion somehow to appear realistic. SolidWorks does have a macro language, though it's limited. I think we should learn something about this, so that we know how to use it and what its capabilities and limitations are. Apparently you record a macro. But then I think you can go back and modify it by adding to the commands that are written by SolidWorks in the macro. It looks also like one can access SolidWorks models from Visual Basic and Visual C++. It may be (I think I read it somewhere) that SolidWorks does have some simple, pseudo-physical modeling capabilities. What are these? There may also be other possibilities, like third-party, animation software that uses SolidWorks models and allows a user to manipulate them or set them in motion. Anyway, we will need to be looking into these possibilities in the near future.
4. There are several technologies we need to prove. To do this, I suggest making simple, arbitrary, proof-of-concept models.
a. We need to look at how two meshing spur gears look when we use the gear mate. The information on the gear mate says that it does not check interference. That means that the teeth could actually be grinding into each other and SolidWorks wouldn't care. Can we avoid this or at least make it so that it is not obvious? SolidWorks also has interference checking. But I think this is a one-time thing that you turn on to make sure you don't have parts running into each other as you design a mechanism.
b. After getting to spur gears to mesh, try a worm.
c. Build a simple pendulum and get it to swing on its on. This will test how to set a mechanism in motion. After doing this, maybe we can add more complicated geometry, driven by the pendulum.
d. Get intermittent motion to work. Animate some simple mechanism that converts steady motion into intermediate motion. Maybe a simple Geneva mechanism. With gear mates, you specify the speed ratio of two meshing shafts. As you rotate one, the other one moves at the relative speed defined in the gear mate. But this doesn't apply to a steady revolution that drives an intermediate output.
5. I always wondered whether or not there is a way to use parts in an assembly as references when making a new part. That is, some parts depend on their shape by their relation to other parts. This is possible. One can create a new part within an assembly and use the assembly geometry to define dimensions on the part. Also important are that there are "mate references", as opposed to just regular mates. These have to do with inter-subassembly kinematic constraints.
6. At some point I may want to go up to Dresden to talk to the professor there whose students built the animation of the steam engine of an old steamboat at the TU. They apparently used 3rd party, virtual-reality, animation software to stage the motion. We are aiming in this direction.
Tasks after 02.05.07 meeting:
Siegfried:
1. Will use his clock knowledge to work up a parts list for the clock. Maybe should work with Herr Rebényi in doing this. May need to give parts numbers. This list will then be given to Olli, Pyry, and Mika to finish and to maintain. The parts list will track what we've done and not done.
2. Will use his clock knowledge to understand the intermittent motion of the clock and try to explain it to the rest of us. This may also be a group effort. We'll try to get Rebényi to give us another tour of the clock in the clocktower and the model that he has. These may be just the thing for us to understand how the pieces fit together and how the clock works.
Hunter and Jesse:
1. Make a simple set of spur gears, use the gear mate, and see how the gears look when they move relative to one another.
2. Do the same with a simple worm gear.
3. Look at possibility of intermittent motion. I suggested making a Geneva mechanism and setting it into motion. I also suggested a simple pendulum above.
4. I'll check into the Dresden trip. I'll also translate the article about the Dresden steam engine model for the group (see here). Hunter and Jesse expressed an interest in going.
Olli, Pyry, and Mika:
1. Will work on modeling what I've measured but not yet modeled. This includes the curved frame and also the Windflügel shaft. Check and see whether the shaft previously modeled by Olli can be used for this shaft.
2. Will send me the project planning spreadsheet for review.
3. Will work with Siegfried on the part tracking list.
Everybody:
1. Start trying to piece assemblies together and get sub-assemblies working.
2. Look into modeling physics, like gravity, mechanical stiffness, etc. Look into "dynamic clearance", "collision detection", and "physical dynamics", all features of SolidWorks.
3. Look at macros as a possibility of setting things into motion. Other motion-creation possibilities.
4. Make sure spreadsheets with calculations accompany parts. Hunter will try pasting drawings into spreadsheets to clarify spreadsheet calculations. See sample spreadsheet.
5. We will not make more detailed parts at this stage. More detail can be added later. We want to proceed quickly and get something working. We can add more detail as we need to.
Final Report
Here's the current version of the final report. Send me contributions to it and I'll add them in. I need to know where they go. So maybe the best thing to do is to download a copy of this, add your contributions in red, then email me the revised report back. I'll add the contributions to the master report.