Question regarding VASP 5.4.1

Vasp transition state theory tools

Moderator: moderators

Post Reply
mh11
Posts: 9
Joined: Fri Jul 31, 2015 1:10 pm

Question regarding VASP 5.4.1

Post by mh11 »

Hello,

I saw R158 of VTST includes changes for VASP 5.4.1 support, but the download still includes two patches for 5.3.2. Could you clarify if VTST supports 5.4.1 and also what the proper steps for compilation are?

My understanding currently is:

1. overwrite VASP's src/neb.F with VTST's neb.F
2. copy all other .F files to VASP's src/
3. patch main.F with vasp-5.3.2-main.patch
4. patch main_mpi.F with vasp-5.3.2-mpmd.patch
5. run 'mkbdrpro.pl bdr_changes' in VASP's src/
6. modify the makefile to add additional .o files before chain.o (<-- can't find where to do this in 5.4.1 since build system changed)
7. compile as normal

I'd be very grateful for any insight/guidance.

Best regards,

Matthew
gump_813
Posts: 25
Joined: Sat Mar 14, 2009 3:29 am

Re: Question regarding VASP 5.4.1

Post by gump_813 »

Hello Matthew,

please compile without step 3,4,5, the patches doesn't work for new vasp.
for step 6, add .o file in src/.objects

Zhiyao
mh11
Posts: 9
Joined: Fri Jul 31, 2015 1:10 pm

Re: Question regarding VASP 5.4.1

Post by mh11 »

Hi Zhiyao,

Many thanks for the reply, I'll try the compilation as instructed. Just to check, can you confirm VTST (specifically NEB/climbing NEB) will still function correctly even without the patches to main.F/main_mpi.F? (I wasn't sure what the patches did so just wanted to make sure.)

Alternatively, are there any tests available to run after compilation to check everything works as intended?

Edit: can confirm it seems to compile fine following your instructions. (I missed the .objects file the first time I was looking for it since it was invisible on my system, so thanks for the tip!)
gump_813
Posts: 25
Joined: Sat Mar 14, 2009 3:29 am

Re: Question regarding VASP 5.4.1

Post by gump_813 »

Hi Matthew,

So for as I know, the vtstcode functions normal with vasp5.4.1. If you see any abnormal behavior, please report it here.
The patch for main_mpi.F is only for vasp working with our eon code.
There is no test cases for the vtstcode right now, I think it is a good idea to include one.

Zhiyao
ashutama
Posts: 7
Joined: Mon Aug 25, 2014 4:04 am

Re: Question regarding VASP 5.4.1

Post by ashutama »

Hello,

Can you give an example of output files of successful run using 5.4.1 ? In my post here viewtopic.php?f=2&t=1696 , Prof. Graeme said that the VTSTTools is not linked into VASP. My understanding so far, I have followed the instruction above from Mr. Zhiyao.

Tommy
ashutama
Posts: 7
Joined: Mon Aug 25, 2014 4:04 am

Re: Question regarding VASP 5.4.1

Post by ashutama »

Dear Prof. Grimme and other VTST developers,

I am using vasp 5.4.1 and vtst 158, I have followed the instruction above and in my understanding, compiled successfully. However, after running an SSNEB test, it doesn't seem to converge. Prof. Graeme said that the TST is not linked to VASP. I installed the VTST as follows:

1.change the src/main.F . Replace

CALL CHAIN_FORCE(T_INFO%NIONS,DYN%POSION,TOTEN,TIFOR, &
LATT_CUR%A,LATT_CUR%B,IO%IU6)
with

CALL CHAIN_FORCE(T_INFO%NIONS,DYN%POSION,TOTEN,TIFOR, &
TSIF,LATT_CUR%A,LATT_CUR%B,IO%IU6)

2. copy vtstcode/* to src/

3. edit src/.objects . Add

bfgs.o \
dynmat.o \
instanton.o \
lbfgs.o\
sd.o \
cg.o \
dimer.o \
bbm.o \
fire.o \
lanczos.o \
neb.o \
qm.o \
opt.o \

before chain.o \

4. Compile as normal (I used mpifort)

Please find the attached log file of the compilation, whether I made mistakes or not. Your helps are greatly appreciated.
Attachments
all.log
(199.16 KiB) Downloaded 22271 times
graeme
Site Admin
Posts: 2290
Joined: Tue Apr 26, 2005 4:25 am
Contact:

Re: Question regarding VASP 5.4.1

Post by graeme »

That build looks fine to me.
Post Reply