# Fixed position structure breaking apart when water is near

**URL:** <https://gromacs.bioexcel.eu/t/fixed-position-structure-breaking-apart-when-water-is-near/519>\
**Category:** User discussions\
**Created:** [July 17, 2020, 3:35pm UTC](https://gromacs.bioexcel.eu/t/fixed-position-structure-breaking-apart-when-water-is-near/519 "2020-07-17T15:35:08Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![ca10nt](https://avatars.discourse-cdn.com/v4/letter/c/eb9ed0/32.png) [@ca10nt](https://gromacs.bioexcel.eu/u/ca10nt)\
**Post date:** [July 17, 2020, 3:35pm UTC](https://gromacs.bioexcel.eu/t/fixed-position-structure-breaking-apart-when-water-is-near/519/1 "2020-07-17T15:35:08Z")

</div>

GROMACS version:  
GROMACS modification: No  
Here post your question

Hello all

I am trying to fille a fixed structure, a tube, with water.

I have employed these sigma and epsilon parameters:  
0.115817 0.011000

Whenever water reaches the inner surface of the tube the simulation crashes. My hope was to create parameters that repel the water, but also don’t crash the system.

I am unsure why water near the inner surface will make my simulation crash.

So the question is why does my simulation crash when water reaches the inner surface, and how do you fix that problem?

Thanks!

---

<div class="post-metadata">

**Author:** ![jalemkul](https://dub1.discourse-cdn.com/flex017/user_avatar/gromacs.bioexcel.eu/jalemkul/32/18_2.png) [@jalemkul](https://gromacs.bioexcel.eu/u/jalemkul)\
**Post date:** [July 19, 2020, 10:51am UTC](https://gromacs.bioexcel.eu/t/fixed-position-structure-breaking-apart-when-water-is-near/519/2 "2020-07-19T10:51:03Z")

</div>

If the simulation is unstable, then the physical interactions are poorly parametrized. What is the source of your LJ parameters?

---

<div class="post-metadata">

**Author:** ![ca10nt](https://avatars.discourse-cdn.com/v4/letter/c/eb9ed0/32.png) [@ca10nt](https://gromacs.bioexcel.eu/u/ca10nt)\
**Post date:** [July 19, 2020, 3:30pm UTC](https://gromacs.bioexcel.eu/t/fixed-position-structure-breaking-apart-when-water-is-near/519/3 "2020-07-19T15:30:06Z")

</div>

I have used multiple LJ parameters, some have been imitating other atomtypes(carbon alpha) looking for just a stable combination. One LJ set drawn from a paper had a sigma of 0.04 and an epsilon as 233 that did not produce even a stable initial structure. I have tried just arbitrarily changing each parameter by orders of magnitude and see how that changes the behaviour but I have yet to find a realistic working scenario.

---

<div class="post-metadata">

**Author:** ![jalemkul](https://dub1.discourse-cdn.com/flex017/user_avatar/gromacs.bioexcel.eu/jalemkul/32/18_2.png) [@jalemkul](https://gromacs.bioexcel.eu/u/jalemkul)\
**Post date:** [July 20, 2020, 2:09am UTC](https://gromacs.bioexcel.eu/t/fixed-position-structure-breaking-apart-when-water-is-near/519/4 "2020-07-20T02:09:14Z")

</div>

Now you see why it is so difficult to obtain well-balanced force field parameters. You can’t necessarily expect arbitrary values to produce physically meaningful results. There are tons of published carbon nanotube simulations that you should be able to use to draw some inspiration (i.e. try their parameters, because they work in those cases) but if your system is geometrically different, there’s no reason to necessarily expect those parameters will be valid.

---

<div class="post-metadata">

**Author:** ![ca10nt](https://avatars.discourse-cdn.com/v4/letter/c/eb9ed0/32.png) [@ca10nt](https://gromacs.bioexcel.eu/u/ca10nt)\
**Post date:** [July 20, 2020, 11:41am UTC](https://gromacs.bioexcel.eu/t/fixed-position-structure-breaking-apart-when-water-is-near/519/5 "2020-07-20T11:41:18Z")

</div>

my next step was carbon nanotubes, thanks for your confirmation it is most helpful!