PDA

View Full Version : Questions about ANC and Transport Theory


Comdrpopnfres
Sep22-09, 03:06 PM
I'm currently enrolled in a course in which we will be modeling parts of a nuclear reactor utilizing Westinghouse ANC code. Here are my questions:

My professor said we will be doing coding on Unix-based machines; so, does that mean the ANC code is FORTRAN 90 based?

Secondly, when performing calculations in ANC using neutron transport theory, does the process involve a lot of computational overhead? Could I install a Unix distro on my PC and (on the off chance we are allowed to have private copies of ANC) work on a project, or just tinker around with it, on my own time and computer?

Lastly, if one does not reduce the number of equations coming from transport theory (via simplifications and assumptions), are the calculations primarily parallel; seeing as the same equations are being used, just with inputs and dimensions as the variances? I'm wondering if using a multi-core PC makes enough of a beneficial difference in chugging through the calculations to justify having my primary computer inaccessible for some time.


Additionally- it is terribly difficult finding any information out about ANC. From what I deduced from some face-time with google, ANC9 is FORTRAN 90 based, where as ANC-H (as far as I know we are not going to use the hexagonal ANC) is written and worked with in either FORTRAN 77 or FORTRAN 95.

Astronuc
Sep22-09, 03:52 PM
See if your professor can get a copy of:
Nguyen, T. Q., et al., “Qualification of the PHOENIX-P/ANC Nuclear Design System for
Pressurized Water Reactor Cores,” WCAP-11597-A (Nonproprietary), June 1988.

My company uses F77 codes on UNIX/LINUX systems, so just because one is running a code on UNIX/LINUX doesn't mean the source code is F90 or higher.

Usually Westinghouse provides only the executable. The source does is highly proprietary.

ANC 9 is just the latest version of ANC.

ANC-H is a 3-D nodal code for space-dependent, few group diffusion, nodal nuclear design and steady-state safety related calculations for VVER-1000 and VVER- 440 cores. The H refers to Hexagonal lattice (triangular pitch). PHOENIX-H is the H-version of PHOENIX.

vanesch
Sep23-09, 10:58 AM
My professor said we will be doing coding on Unix-based machines; so, does that mean the ANC code is FORTRAN 90 based?


I don't know how interesting this is, but there's no particular link between the operating system and the source language. In fact, the most native language for all UNIX/LINUX kind of software is C and its derivatives, but there exist compilers for other languages like fortran.

Once you have an executable, whatever was the source code to generate doesn't matter much (unless there are API-related issues, which might play a role). Apart from open-source projects, you rarely get your hands on the source code.

Astronuc
Sep24-09, 07:22 PM
This paper might be of interest

Reactor Physics Methods Development at Westinghouse

http://www.djs.si/proc/port2007/htm/pdf/101.pdf

BishopUser
Sep28-09, 02:24 PM
I am curious as to what university you attend. If it is PSU, I could answer most/all of your questions.

QuantumPion
Sep28-09, 05:35 PM
I am curious as to what university you attend. If it is PSU, I could answer most/all of your questions.

Go kitties! ^_^

CFDFEAGURU
Oct3-09, 05:42 PM
Good ole' Westinghouse. I live about 45 minutes from their new headquarters in Cranberry, PA and the Energy Center is minutes from me. If you want I can sneak in for ya. LOL just kidding for anyone who is watching. LOL

Thanks
Matt