Hi Lammps community,
I’m a freshmen in LAMMPS, I already installed the LAMMPS in the PC and follow the tutorial. But when I follow the sample LAMMPS input script to calculate thermal conductivity (http://lammps.sandia.gov/doc/compute_heat_flux.html)
I got different result from the suggestion author. Any one could assist me ?
Best wish,
Aziz
Ahmad S. Aziz Agus | Australian Institute for Bioengineering & Nanotechnology (AIBN)
Building 75 - Cnr College and Cooper Roads
The University of Queensland | St Lucia, Queensland 4072
Telephone + 61 7 3346 3978 | Fax + 61 7 3346 3973
email a.agus@…3844…
Web address www.aibn.uq.edu.au
I think we need you to put more effort into asking your question.
Nobody can help you without more information.
I personally don't know how to do this calculation, so I can't answer you.
But please make another post on this same thread which clarifies:
1) which tutorial (which web page)
2) what are the results you are getting now?
If the problem might be explained by a picture, please take the
time to include one.
If your output files like, log.lammps, are less than 200K, you can
also attach it.
How are these results different from what you expected?
Cheers and good luck
Andrew
Thanks for your advice Andrew,
I already mention the calculation that I tested, but I forgot to put my results.
My results is 3779,311 W/mK @ 70K, 0.025744366602047559 /A^3
The result should be: average conductivity ~0.29 in W/mK (author vers)
I just copy-paste the code from the website (http://lammps.sandia.gov/doc/compute_heat_flux.html) and run it.
Which version of LAMMPS are you using to run this simulation? A bug in the trap() function, which caused it to give incorrect results, was recently fixed. If you've been using an older version, try building the most recent revision and running again. I wouldn't necessarily expect the discrepancy to be that large, but it's a start.
I've just tested this on my own machine. The correction to trap() makes a small difference (~0.30 vs ~0.29), but nothing like what you've seen. I ran the script in your e-mail without changes, so you may want to check there's no typo in your input script on disk.
I've just tested this on my own machine. The correction to trap() makes
a small difference (~0.30 vs ~0.29), but nothing like what you've seen.
I ran the script in your e-mail without changes, so you may want to
check there's no typo in your input script on disk.
as an add-on (and a warning) to that.
i've noticed repeatedly that web browsers
and pdf viewers seem to render text into
UTF-8 encoding while replacing certain
ASCII characters by similar looking UTF-8
characters. this happens when native language
support is enabled.
this will lead to errors, even when simply doing
a copy-and-paste without modifying the text at all.
LAMMPS only recognizes ASCII text and will
thus not recognize those characters as what you
might think they represent. the dash ('-') is a
common example.
axel.