laterally graded thicknesses

[leverglow]

Hi,
Depth graded multilayer can be set by xrtQook easily,but xrtQook says that'If you need laterally graded thicknesses, modify get_t_thickness and/or get_b_thickness in a subclass.'
My question is how to set the laterally graded multilayer.I don't know how to'modify get_t_thickness and/or get_b_thickness in a subclass'

[kklmn]

Hi,
The method's signature can be found in the docs. x and y are in mm, the returned thickness is in Å.

[leverglow]

Thanks for your reply,I found what it means in the docs.So may I set the laterally graded multilayer directly in xrtQook by commenting or uncommenting codes?Or I need to wirte codes myself to set it.

[kklmn]

If you want it in xrtQook then it's a bit harder. You have to create a derived class of Multilayer yourself where you implement the methods get_t_thickness and get_b_thickness, xrtQook cannot know the lateral thickness variation you are going to implement. There is a mechanism in xrtQook for registering custom (user) optical element classes, but not custom material classes. So you will have to also make a derived class from the optical element you're using, in whose __init__ you specify the custom multilayer material. Then you register this OE class in xrtQook, as explained in xrtQookStart.py.

[leverglow]

OK,that's too difficult for me.So how I can directly use 'get_t_thickness'.If I use gradedMultilayer,the codes are'gradedMultilayer02 = rmats.GradedMultilayer(
tLayer=Ru,
tThickness=16.5,
bLayer=C,
bThickness=16.5,
nPairs=60,
tThicknessLow=19.9,
bThicknessLow=19.9,
name=None)'
I can determine depth graded multilayer's thickness through' tThickness=16.5',so how to determine laterally graded thicknesses through'get_t_thickness'

[kklmn]

So how I can directly use 'get_t_thickness'.

It is not you who uses the method, xrt uses it. You create it.

[leverglow]

So in which program(.py) should I modify the code to implement laterally graded thicknesses.

[kklmn]

You should subclass Multilayer from xrt/backends/raycing/materials.

[leverglow]

I write the code
'class LMultilayer(Multilayer):
def get_t_thickness(self, x, y, iPair):
global XY # to extract x,y values
XY = np.array([x,y])
return self.dti[iPair]
def get_b_thickness(self, x, y, iPair):
return self.dti[iPair]'in materials.py.
Then I tried to use it through
'gradedMultilayer01 = rmats.LMultilayer(
tLayer=Ru,
bLayer=C,
tThickness=19.5,
bThickness=19.5,
nPairs=150,
tThicknessLow=15.2,
bThicknessLow=15.2,
name=None)'
but it does't work,the results are the same when I change'gradedMultilayer01 = rmats.LMultilayer(' to 'gradedMultilayer01 = rmats.GradedMultilayer('.
I don't know if I'm calling the function(class LMultilayer) in a wrong way or if the function itself is wrong

[kklmn]

How do you want to define the graded thickness? How does it change in space as a function of x and y?

[leverglow]

The laterally graded multilayer consists of two materials and their thickness only varies with x,so no matter where the laser hits, it can have a high reflectivity.There is a function between x and thickness.So do I need to write complex computational relationships between them as part of the code?Or just modify parts of depth graded multilayer's code.

[kklmn]

There is a function between x and thickness.

You need to implement this dependence inside the two methods get_t_thickness and get_b_thickness in your subclass.

[leverglow]

Okay, so how does the depth gradient multilayer set up the function between x and depth? I would like to refer to it to set up the lateral gradient multilayer.

[kklmn]

If I understand your question correctly (I am not sure about it) then my first reply in this thread is the answer to it.

[leverglow]

I know the function'get_b_thickness(x, y, iPair)' and what each parameter represents,but how can I use it in my python program.For example, my lateral gradient multilayer goes from 30nm to 20nm, and there are 100 pairs.I mean what exactly is the code to use 'get_b_thickness(x, y, iPair)'and 'get_t_thickness(x, y, iPair)'.

[kklmn]

Any reason why somebody should code for you?

[kklmn]

One such reason could be your will to share experimental data that you're trying to model. Another reason for me could be in seeing that this study is a part of a bigger and attractive project.