# Relaxation with Symmetry Preserved

**URL:** <https://discourse.abinit.org/t/relaxation-with-symmetry-preserved/3775>\
**Category:** Ground state\
**Created:** [November 4, 2025, 10:22pm UTC](https://discourse.abinit.org/t/relaxation-with-symmetry-preserved/3775 "2025-11-04T22:22:38Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![sms](https://discourse.abinit.org/letter_avatar_proxy/v4/letter/s/ecb155/32.png) [@sms](https://discourse.abinit.org/u/sms)\
**Post date:** [November 4, 2025, 10:22pm UTC](https://discourse.abinit.org/t/relaxation-with-symmetry-preserved/3775/1 "2025-11-04T22:22:38Z")

</div>

Hi,

I am looking for suggestion where I want to relax a structure by preserving the symmetry group of the crystal structure of it. The structure I am doing is somewhat a modified crystal, which might not be the ground state structure of it. My goal is to relax the volume of the crystal by keeping the symmetry preserved and let the atoms to move in ab-plane but not in c-direction.

Thanks

---

<div class="post-metadata">

**Author:** ![mverstra](https://discourse.abinit.org/user_avatar/discourse.abinit.org/mverstra/32/92_2.png) [@mverstra](https://discourse.abinit.org/u/mverstra)\
**Post date:** [November 5, 2025, 9:58pm UTC](https://discourse.abinit.org/t/relaxation-with-symmetry-preserved/3775/2 "2025-11-05T21:58:36Z")

</div>

several questions here.

1. abinit always preserves the initial space group it finds, unless you set nsym 1. The positions, then the forces and stress are symmetrized before being applied to step through the relaxation

2. If you want to impose a constraint on the atoms and directions which can be relaxed, it’s easy with the variables natfix, iatfixx etc. See their definitions

3. if you want to impose constraints on the cell relaxation (const volume, only change c parameter etc) you can choose different values of optcell.