Menu

Discussion Forum

New Discussion
Filter

Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Results display

Posted Mar 21, 2023, 4:09 a.m. EDT Version 5.6 2 Replies

Tao Zhou
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

My simulation is based on experiments, and the geometric modeling is 5mm × 8mm, while in the experiment, particle image velocimetry was used, and it was observed with a magnifying glass (2.66× 2.66mm) at 5 times the magnification , but now the speed in the entire geometry that I have obtained, the range of the rainbow bar, compared to the speed measured in the experiment, is generally smaller, with a few orders of magnitude difference. How do I handle this?If you have any suggestions,Thank you very much.

2 Replies Last Post Mar 21, 2023, 11:04 p.m. EDT
Dave Greve
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 2 years ago Mar 21, 2023, 11:47 a.m. EDT

This question concerns the agreement between experiment and simulation, which is beyond the scope of this forum.

All that can be said is that there is a problem with the experiment, the simulation, or both.

This question concerns the agreement between experiment and simulation, which is beyond the scope of this forum. All that can be said is that there is a problem with the experiment, the simulation, or both.
Robert Koslover Certified Consultant
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 2 years ago Mar 21, 2023, 11:04 p.m. EDT

Dave Greve is right, but... I would indeed like to offer a suggestion! You say that you compared your computed to measured value and find "a few orders of magnitude difference." That's a huge difference! Any scientist knows that before you ever do an experiment, or before you ever create a model, you should have at least a rough idea/estimate of what the answer is going to be! So Dave is entirely right to say that "there is a problem with the experiment, the simulation, or both." But as a scientist yourself, you should be using your own knowledge of the subject to guide you in which one (that is, the experimental measurement vs. the model-based prediction) to investigate further, based on its huge disagreement from your own estimate! Good scientists typically make good estimates, by which I mean to say that their pre-experiment or pre-model estimates are not usually in error by "a few orders of magnitude." Thus, if the experimental result is fairly close to your estimate, the model is probably wrong. If the model result is fairly close to your estimate, there was probably an error in the experiment. But, if by some chance you didn't even make an estimate to begin with, you need to go back and do that. It's the way science is done. E.g., see https://en.wikipedia.org/wiki/Scientific_method#Elements_of_the_scientific_method

-------------------
Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
Dave Greve is right, but... I would indeed like to offer a suggestion! You say that you compared your computed to measured value and find "a few orders of magnitude difference." That's a huge difference! Any scientist knows that before you ever do an experiment, or before you ever create a model, you should have at least a rough idea/estimate of what the answer is going to be! So Dave is entirely right to say that "there is a problem with the experiment, the simulation, or both." But *as a scientist* yourself, you should be using your own knowledge of the subject to guide you in which one (that is, the experimental measurement vs. the model-based prediction) to investigate further, based on its huge disagreement from your own estimate! Good scientists typically make good estimates, by which I mean to say that their pre-experiment or pre-model estimates are not *usually* in error by "a few orders of magnitude." Thus, if the experimental result is fairly close to your estimate, the model is probably wrong. If the model result is fairly close to your estimate, there was probably an error in the experiment. But, *if by some chance you didn't even make an estimate to begin with*, you need to go back and do that. It's the way science is done. E.g., see https://en.wikipedia.org/wiki/Scientific_method#Elements_of_the_scientific_method

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.

Suggested Content