maxschalz / flexicamore

Flexible Cycamore. Legacy repository (incl. issues and PRs) available at https://git.rwth-aachen.de/nvd/fuel-cycle/flexicamore

Flexible Cycamore

This repository contains Cyclus archetypes, strongly based on Cyclus' Cycamore library. These archetypes are designed to be more versatile by allowing input variables to be time-dependent, e.g., the SWU capacity of an enrichment facility can be configured to change during the simulation. This behaviour is implemented for few cases in Cycamore, and the goal of this repository is to facilite the expansion of this behaviour to other variables.

This repository is developed by the Nuclear Verification and Disarmament group from RWTH Aachen University, Germany.

How to use Flexicamore

Installing flexicamore

1. Make sure you have Cyclus installed.
2. Clone this repository.
3. Run the install-script: $ python3 install.py.
4. Done! You can run the unittests using $ flexicamore_unit_tests which should result in all tests passing successfully. Currently, there are 96 tests from 14 test cases and 3 disabled tests.

Hands on: example input file

An input file and the corresponding output using all flexicamore archetypes can be found in the input directory. The input file has been tested successfully on Fedora35 using the following versions:

$ cyclus --version
Cyclus Core 1.5.5 (1.5.5-57-gc1910b90)

Dependencies:
   Boost    1_74
   Coin-Cbc 2.9.9
   Coin-Clp 1.16.11
   Hdf5     1.10.6-
   Sqlite3  3.35.5
   xml2     2.9.10
   xml++    2.40.1

A bit more in-depth

The underlying class, FlexibleInput, can be used in two ways:

1. Define a std::vector of values and a std::vector<int> containing the changing times, or,
2. Define a std::vector containing values for all timesteps.

In addition, one always has to pass a pointer to the agent in question. Explained in code:

// Consider a uranium mine with a production rate that increases every three
// timesteps by 1, s.t.:
// Simulation timestep: 0  1  2  3  4  5  6  7  8
// Production rate:     1  1  1  2  2  2  3  3  3

// Method 1
std::vector<int> change_times({0, 3, 6});
std::vector<double> new_throughputs({1, 2, 3});
FlexibleInput<double> flexible_production(&my_source, new_throughputs, change_times);

// Method 2, more verbose
std::vector<double> throughputs({1, 1, 1, 2, 2, 2, 3, 3, 3});
FlexibleInput<double> flexible_production(&my_source, throughputs);

Important note: All changing times are defined relative to the facility's simulation entry time. Consider for example a uranium mine entering the simulation at time t = 5 with the following FlexibleInput:

std::vector<int> change_times({0, 3, 6});
std::vector<double> new_throughputs({1, 2, 3});
FlexibleInput<double> flexible_production(&my_source, new_throughputs, change_times);

This would result in the following production rates:

// Facility entering at simulation time t = 5
// Simulation timestep: 0  1  2  3  4  5  6  7  8  9 10 11 12 13
// Production rate:     0  0  0  0  0  1  1  1  2  2  2  3  3  3

FlexibleEnrichment

Flexible variables:

• SWU capacity. To use method 2, set swu_capacity_times to -1 and indicate all SWU values in swu_capacity_vals.
• Multiple feed commodities can be specified (at the same time, including preferences). See the example input file in input/.
• Note: currently, order_prefs should be set to false if feed commodity preferences are used due to a possible bug, see issue 4. Due to this, the unit tests also show two disabled tests.

FlexibleSource

Flexible variables:

• Throughput (the production rate). Currently allows using both methods, similar to FlexibleEnrichment.

FlexibleStorage

Flexible variables:

• inventory size: total amount of material present in the facility at a given moment.

FlexibleSink

Flexible variables:

• Throughput: maximum amount of material requested and (if available) accepted per timestep.