# Site map

A phenomenon occuring in nearly every sample is the preferred orientation (PO). Because we cannot avoid this we have to provide a model for its description. BGMN uses the spherical harmonics model up to the 10th order which allows the refinement of multiple and strong PO.

If one phase exhibits PO, a simple statement is necessary:

```GEWICHT=SPHARx
```

`x` determines the order of the spherical harmonics model and can be 0, 2, 4, 6, 8 or 10. Except the triclinic lattice there are constraints between the parameters of the model. BGMN itself determines these constraints automatically.

Lets have a look at the gypsum structure file, lines 5 and 6:

```PHASE=Gypsum SpacegroupNo=15 HermannMauguin=I12/a1 //
PARAM=A=0.6522_0.645^0.658 PARAM=B=1.5202_1.51^1.53 //
PARAM=C=0.5679_0.56^0.575 PARAM=BETA=118.4_118^119 //
RP=3 PARAM=B1=0_0^0.008 //
GEWICHT=SPHAR6 GOAL:gypsum=GEWICHT
E=CA+2 Wyckoff=e TDS=0.0088 y=0.4213
E=S    Wyckoff=e TDS=0.0066 y=0.9216
E=O-2  Wyckoff=f TDS=0.0149 x=0.5505 y=0.1319 z=0.9632
E=O-2  Wyckoff=f TDS=0.0141 x=0.6671 y=0.0223 z=0.7582
E=O-2  Wyckoff=f TDS=0.0219 x=0.4588 y=0.1821 z=0.3796
```
There is the `GEWICHT=SPHAR6` statement in line 5 which determines the use of a 6th order spherical harmonics model.

If the phase content is too low, the order of the model will be reduced down to `SPHAR4`, `SPHAR2` or even `SPHAR0` (isotropy), if necessary. Thats why you can use the same structure file for all analytical problems even if nearly nothing of the phase is contained in the sample.