Forum Navigation
You need to log in to create posts and topics.

Ovito installation problems

After ovito is installed on the desktop computer and notebook, the dislocation of the close packed hexagonal structure is different, so it needs to be solved.

Hi,

Could you please be more specific? Are you saying that the DXA function produces different outputs on two different computers for the exact same input configuration and the same parameters?

-Alex

Ovito version is 2.9.0-win 64. They are all win 10 systems with the same input parameters (close-packed hexagonal structure). The dislocation displayed on the portable notebook is 1 / 2 < 110 >, 1 / 6 < 112 > and so on, while the dislocation displayed on the desktop computer is < 0001 >, < 1-100 >, 1 / 3 < 1-100 > and so on.I think the desktop computer is more accurate, but I want to use the portable notbook now. How can I solve this problem?

Yours

ChunXue

Uploaded files:
  • You need to login to have access to uploads.

Hi Chun,

From the pictures you posted, it seems that there is no dislocation in your structure. In the first picture, the Dislocation analysis results  show that "No dislocation found"; besides, the segement and lenghth of all of the types of dislocations are all "0". Check if you chose the correct input structure type when you were trying to analyze the dislocation in your hexagonal structure.

Hi Li,

There are dislocations in the model. My screenshot is just to show that they show different types of dislocations.These are the types of dislocations that follow.

 

My suggestion for you is to try the same analysis with the current OVITO version 3.1. It might be that the old DXA implementation in OVITO 2.9 does not yield fully reproducible outputs, because the algorithm uses random number sampling internally. At some point in the past (I don't remember when exactly - perhaps after the release of OVITO 2.9 in 2017) this behavior was changed. Since then DXA uses fully reproducible pseudo-random numbers in the algorithm and the output should always be identical (assuming the input structure and algorithm parameters are identical).

Hi,

I have used the version of ovito 3.1 to analyze under the notebook win10 system, and the result is still the same as the version of ovito 2.9. What I want to achieve now is to see the four-axis representation of the dense six-way structure on the notebook (the modeling is also the four-axis representation).

-ChunXue

Generally, DXA uses 4-component Miller indices to express the Burgers vectors of dislocations embedded in hexagonal (HCP) crystal regions and 3-component Miller indices to express Burgers vectors of dislocations embedded in cubic (FCC) crystal regions. Your dataset may contain a mix of both crystal regions. HCP and FCC lattices can form coherent interfaces, in particular, stacking faults in FCC parent crystals are formed by HCP-like atoms and --vice versa-- stacking faults in HCP parent crystals are formed by FCC-like atoms.

What I want to say is that ambiguous situations can easily arise in close-packed systems, because a single dislocation line may lead through several different crystal regions (FCC- and HCP-like, stacking faults of variable thickness). In such a situation, DXA has to decide in which crystal lattice frame to express the Burgers vector of that dislocation line. The selected input structure for the DXA modifier controls the priority that is given to FCC and HCP respectively, but even if you select "hexagonal close-packed" as input crystal structure, it may be impossible for DXA to express the dislocation's Burgers vector in the hexagonal form if that (short) dislocation segment is completely contained inside a wide (FCC-like) stacking fault region. DXA will then automatically switch to the cubic, 3-component vector representation.

To tell you more about the behavior in your specific situation, I would need to take a look at your input dataset. Please share it here if you would like me to investigate this issue in more detail.

Thank you, Alex and Li. I have found this problem. It's not a software problem,which is that I didn't choose the type of input crystal. I'm very sorry for my mistake and trouble you.

-Chun