[CP2K-user] Nonsensical energies with MGGA functionals via LibXC

[Peter Kraus]

Dear Colleagues,

I am trying to do a simple calculation with some MGGA functionals. This is 
for CP2K 6.1 compiled with Libxc 4.2.3. The calculations with the built-in 
functionals (using &XC_FUNCTIONAL PBE or &XC_FUNCTIONAL TPSS) work fine. 
Calculations of Libxc driven PBE also work fine, with the attached input 
file. When switching to TPSS (or M06_L), as long as I use only one 
component (XC_MGGA_X_TPSS or XC_MGGA_C_TPSS), the calculations work OK. 
When I try to combine them (see the other attachment), the energy is 
immediately nonsensical. Combining eg. XC_MGGA_X_TPSS with XC_GGA_C_PBE 
also doesn't work.

Arch file attached.

I hope you can help,
Best,
Peter
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20190111/afa90ac3/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pbe.inp
Type: chemical/x-gamess-input
Size: 1942 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20190111/afa90ac3/attachment.inp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: tpss.inp
Type: chemical/x-gamess-input
Size: 1946 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20190111/afa90ac3/attachment-0001.inp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: openmp.psmp
Type: application/octet-stream
Size: 1677 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20190111/afa90ac3/attachment.obj>