I suggest you “close” this topic now so that nobody can reply beyond this and we stop having multiple discussions on different topics in the same topic. Thanks.
1 Like
Thank Steven for the topic and just some more words.
- Thank you very much Mr. Kohlmeyer for your suggestion. I feel a bit sorry that I did not explain well about the thermostat. I just though about the situation of indentation from Steven. If we use NVE for all atoms in the sample structure, the temperature of the sample will continue going up during the process of indentation. The fix temp/scale is just one way to dissipate the heat from the sample. Your discussion have greatly helped to think differently about thermostat. I sincerely appreciate it.
- For learning MD, I agree that most of us have to learn by ourselves and it is not easy (if not extremely difficult) to find a tutor for MD, because the most important thing is about the physics behind our simulations. We absolutely can learn a lot from any software document we used. But I myself find the Lammps forum from Matsci.org is a great opportunity for me to express my though about some specific topic. In exchanging my ideas with professional people in LAMMPS community, I can see my weaknesses and I can learn from others.
Again, thank you Mr. Kohlmeyer and Steven, I really learnt a lot from you!
2 Likes
I strongly disagree with this statement. Learning MD on your own is the worst way to learn it. MD is not a new technique and there are plenty of people around that know it and use it; also at your institution.
This is a systemic problem and there are two major reasons for it: a) the advisers and b) the students.
about point a): If this was not computer simulations, but experimental research using machinery and equipment like a UHV apparatus, a laser, or chemicals or anything similar, there would be no question at all that you would not be doing anything without supervision and without a proper introduction into how to use the machinery and devices and how to do your work. Now, few advisers provide that kind of training themselves but arrange for it. But with computer simulations this kind of responsibility goes out of the door for too many of them and their students now show up in forums like this and expect that the developers of the software do their job for them. Mind you, I am not saying that all of them do this (it cannot be because otherwise the number of messages here would be much higher), but they are still too many.
The part that I don’t understand about this is how careless this is also for their own reputation, since if they put their name on publications of flawed results because the students they advise make easily avoidable mistakes it should seriously concern them, but apparently the risk of being called out for low quality research results is not much of a motivation (or risk).
about point b): this issue would be half as bad if the affected students would own up to their needs and be more vocal about this and remind their advisers about their responsibilities and realize that they have much more power than they are led to believe they have. yes, they are dependent on their advisers to graduate, but the advisers need their labor and results (and particularly good results even more so) in a highly competitive research environment and that also gives the students some power over their advisers. if an adviser says that he has no knowledge of a method and cannot teach you, then you should think twice about whether it is a good idea to do such a project and better ask for a different project. or you have to be pro-active and ask around. there will be people around that can help you, and if you ask the right way, they will at least get you started and that will make things much better than just making things up on your own and being guaranteed to have many mistakes and thus much extra work to correct for them. It won’t happen automatically, you have to become active yourself. Just complaining, especially in a forum like this one, is not going to help you much and not getting much sympathy. If you want to get help you have to show that you can do your part - however little - and be pro-active, not tell how bad things are and how little help you get.
I am using the term student in a very loose sense here, since a lot of what I stated here applies to postdocs as well.
Forums like this can be a great resource, but they cannot be and should not be the first resource for advice and help. People like me should not be required to explain the same basic things over and over again and tell people to bug their advisers and get proper training before doing complex simulation projects. I could go on about this for hours, but I better stop before I start using offensive language and really turn off people.
Just let me reiterate one more time. This is a forum about the software LAMMPS, not about your research and you have to learn to tell the difference. We are willing to help you master getting LAMMPS to do what you need it to do, but we cannot also teach you what that is.
2 Likes