Dear FLUKA experts, I want to create a source distribution as an arbitrary function, e.g. f(x,y) =A exp(-x^2 - y^2) and then implement it in my FLAIR simulation, but I don’t know how to go about it. Any help would be appreciated. Best Regards, Carlos

Dear Carlos, you need to use a source user routine, to code your arbitrary function. If you didn’t use it before, I suggest to try the updated version: source_newgen.f The following lecture introduces its usage, and it has some basic information on Fortran programming: 17_Source_routine…

Dear David, Thank you for your answer and sorry for my late reply. I used source_newgen.f as you recommended, but when I try to use an arbitrary distribution (or even the sample_flat_distribution) I get this error. [Screenshot from 2021-01-25 13-56-51] I attached my files for clarity. basicsour…

Dear Carlos, the explanation is simple. Your beam is not where you expect it. So, the all values in the USRBIN scoring are zeros, and Gnuplot can’t plot it with a logarithmic color bar. Luckily, you are using the new source_newgen.f user routine, and this has a debugging option. If you set the deb…

Dear David, Thank you for your answer. I did the modifications that you told me and It worked, but now I am trying to use sample_annular_distribution subroutine but I don’t think it’s working unless I am missing something. I attach my files. basicsource.f (4.9 KB) basicsource.flair (3.2 KB) basic…

Dear Carlos, you do not need to put your variables into “[”, “]” brackets, they only supposed to indicate which variables has to be replaced by a user defined one. Probably, this needs to be more clearly written in the guide. There are two other issues: In the source routine you divide the radi…

Dear David, Thank you for your input, about the issues: It worked perfectly after I realized what you said by chaging my input parameters 2.a. It worked perfectly when getting one layer 2.b It also worked perfectly by setting the divergence to zero. Now, going back to my first post, If I wante…

Dear Carlos, could you elaborate what distribution you want to implement? The f(x,y) distribution you mentioned, looks like a 2D Gaussian distribution, where X and Y variables are independent. In that case this distribution can be replaced with two separate 1D Gaussian ones. Cheers, David

Dear David, Thank you for your answer and please ignore my last question, it’s badly worded. I have been messing around with my cosine_distribution function and it seems to make sense, but if I change my input parameters starts behaving strangely. I would think since I am using f(x)=cos(x) as my f…

Dear Carlos, unfortunately I can’t understand the cosine distribution in your source routine. Is there any reference what you want to code? However, I can elaborate on how to sample from arbitrary 1D distribution. Let’s try to sample the X random variable from the cos(x) function as a distributi…

Dear David, Thank you for your explanation, it does help a lot. I have been messing around with your implementation and what I am thinking is if there is way to show multiple times the same distribution, e.g.: I want to sample from 0 to 20 and the beam is centered at 10. Without rescaling the dis…

Dear Carlos, you can set up a random number, which would give -1, 0, or 1 like nrandom = int( FLRNDM(xdummy) * THRTHR ) - 1 sample = random_cosine + TWOPIP * nrandom Cheers, David

FLUKA User Forum

Arbitrary distribution for a user defined source

Advanced Features and User Routines

show post in topic

Home
Categories
Guidelines
Terms of Service
Privacy Policy

Powered by Discourse, best viewed with JavaScript enabled