Query reg Cylinder rotation
Aswath Narayanan R.R.B na20b009
Radu Serban
Aswath,
A few things:
- Please provide a more detailed description of *what* you are trying to model. None of your posts makes that clear; you talk about a cylinder, but don’t say anything if it is joined in any way, what kind of forces are supposed to act on it, etc, etc
- Please keep all messages related to the same issue/question in a single email thread. Reply to your own post if you need to add information
- If you decide to provide code, please
- Provide the full code, including the important parts related to what you model
- Attach code as text, not images. Someone may want to test the code you provide, but nobody will retype everything from a picture
So, please provide full information on what it is you are trying to model and simulate with Chrono and I’m sure someone will be able to help you
--Radu
--
You received this message because you are subscribed to the Google Groups "ProjectChrono" group.
To unsubscribe from this group and stop receiving emails from it, send an email to
projectchron...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/projectchrono/52f4b435-8863-40c4-9a16-c56c0a21fec5n%40googlegroups.com.
Radu Serban
Aswath,
Assuming your data was obtained experimentally, even if it was indeed “100% accurate” (which is arguably an incorrect statement and no fallacious appeal to authority will fix that) you still have the issue of all the other modeling data being just an approximation to the real phenomenon. For example, was your object a perfect cylinder? Do you know its exact mass? Can you guarantee that the material was perfectly homogeneous so that your moments of inertia are exact? Was the gravitational acceleration exactly 9.81 m/s2? Etc., etc. These are all rhetorical questions: the answer to every one of them is no.
Having said that, one can still set up a good simulation model and use it for meaningful predictions. But with care.
For instance, you expect your object to “not fall”. Why is that? Supposedly because in the experiment it was precented to do so, either through a joint or else through contact (i.e., your cylinder was resting on a support). Why then do you not model that part of the experiment? With Chrono (and many other multibody simulation packages) you could model either case.
Alternatively, if you are only interested in this oscillatory motion you are trying to reproduce in simulation, why not ignore gravity altogether? Set g=0 in your Chrono simulation and do not apply any of the vertical forces from your dataset (you may also need to take care of some moments, but that depends on how the cylinder is kept from falling in the actual experiment – see above).
Speaking of this vertical force: you say the data is 100% accurate while claiming the cylinder should not fall. Yet, glancing at your data, I see that the vertical force is not constant…
One last thing: everything I said above is based on some guesses I had to make as to what you are actually trying to model. All I’m saying is that you need to go through a process of understanding your experiment, understanding what you want to model (what is important), and understand all assumptions and approximations inherent in any mathematical modeling.
--Radu
From: Aswath Narayanan R.R.B na20b009 <na20...@smail.iitm.ac.in>
Sent: Friday, June 24, 2022 3:36 AM
To: Radu Serban <ser...@wisc.edu>
Subject: Re: [chrono] Query reg Cylinder rotation
I am trying to simulate a cylinder's response to the force and moment dataset i am passing. The dataset I am passing is 100% accurate as it's verified on a thesis. The expected outcome of the cylinder is Simple harmonic oscillation but it's falling down as I had shown in the other video. I am sending my entire code and dataset.
To view this discussion on the web visit https://groups.google.com/d/msgid/projectchrono/PH0PR06MB8237F6FFB297176B167F5C2FA7B29%40PH0PR06MB8237.namprd06.prod.outlook.com.