What’s New in 0.14.0

Summary

This release of OpenMC includes many bug fixes, performance improvements, and several notable new features. Some of the highlights include projection plots, pulse height tallies for photons, weight window generation, and an ability to specify continuous removal or feed of nuclides/elements during depletion. Additionally, one of the longstanding annoyances of depletion calculations, namely the need to include initial “dilute” nuclides, has been eliminated. There are also a wide array of general improvements in the Python API.

Compatibility Notes and Deprecations

New Features

  • A new openmc.ProjectionPlot class enables the generation of orthographic or perspective projection plots. (#1926)

  • The openmc.model.RightCircularCylinder class now supports optional filleted edges. (#2309)

  • Continuous removal or feed of nuclides/elements between materials can now be modeled during depletion via the openmc.deplete.abc.Integrator.add_transfer_rate() method. (#2358, #2564, #2626)

  • The MAGIC method for global weight window generation has been implemented as part of the C++ API. (#2359)

  • A new capability for pulse height tallies (currently limited to photons) has been added and can be used via the “pulse-height” tally score. (#2452)

  • A openmc.model.CruciformPrism class has been added that provides a generalized cruciform prism composite surface. (#2457)

  • Type hints have been added in various places throughout the Python API. (#2462, #2467, #2468, #2470, #2471, #2601)

  • Voxel plots can now be generated through the openmc.Plot.to_vtk() method. (#2464)

  • The openmc.mgxs.EnergyGroups class now allows you to alternatively pass a string of the group structure name (e.g., “CCFE-709”) instead of the energy group boundaries. (#2466)

  • Several enhancements have been made to the openmc.Universe.plot() method (addition of axis labels with units, ability to show legend and/or outlines, automatic determination of origin/width, ability to pass total number of pixels). (#2472, #2482, #2483, #2492, #2513, #2575)

  • Functionality in the Python dealing with bounding boxes now relies on a new openmc.BoundingBox class. (#2475)

  • Users now have more flexibility in specifying nuclides and reactions in the openmc.plot_xs() function. (#2478)

  • The import time of the openmc Python module has been improved by deferring the import of matplotlib. (#2488)

  • Mesh clases in the Python API now support a bounding_box property. (#2507, #2620, #2621)

  • The Source class has been refactored and split up into three separate classes: IndependentSource, FileSource, and CompiledSource. (#2524)

  • Support was added for curvilinear elements when exporting cylindrical and spherical meshes to VTK. (#2533)

  • The openmc.Tally class now has a multiply_density attribute that indicates whether reaction rate tallies should include the number density of the nuclide of interest. (#2539)

  • The wwinp_to_wws() function now supports wwinp files with cylindrical or spherical meshes. (#2556)

  • Depletion no longer relies on adding initial “dilute” nuclides to each depletable material in order to compute reaction rates. (#2559, #2568)

  • The openmc.deplete.Results class now has get_mass() (#2565), get_activity() (#2617), and get_decay_heat() (#2625) methods.

  • The openmc.deplete.StepResult.save() method now supports a path argument. (#2567)

  • The openmc.deplete.MicroXS has been completely redesigned and improved. First, it no longer relies on the openmc.mgxs module, no longer subclasses pandas.DataFrame, and doesn’t require adding initial “dilute” nuclides into material compositions. It now enables users to specify an energy group structure to collect multigroup cross sections, specify nuclides/reactions, and works with mesh domains in addition to the existing domains. A new openmc.deplete.get_microxs_and_flux() function was added that improves the workflow for calculating microscopic cross sections along with fluxes. Altogether, these changes make it straightforward to switch between coupled and independent operators for depletion/activation calculations. (#2572, #2579, #2595, #2700)

  • The openmc.Geometry class now has merge_surfaces and surface_precision arguments. (#2602)

  • Several predefined energy group structures have been added (“MPACT-51”, “MPACT-60”, “MPACT-69”, “SCALE-252”). (#2614)

  • When running a depletion calculation, you are now allowed to include nuclides in the initial material compositions that do not have neutron cross sections (decay-only nuclides). (#2616)

  • The CylindricalMesh and SphericalMesh classes can now be fully formed using the constructor. (#2619)

  • A time cutoff can now be specified in the openmc.Settings.cutoff attribute. (#2631)

  • The openmc.Material.add_element() method now supports a cross_sections argument that allows a cross section data source to be specified. (#2633)

  • The Cell class now has a plot() method. (#2648)

  • The Geometry class now has a plot() method. (#2661)

  • When weight window checks are performed can now be explicitly specified with the openmc.Settings.weight_window_checkpoints attribute. (#2670)

  • The Settings class now has a max_write_lost_particles attribute that can limit the number of lost particle files written. (#2688)

  • The CoupledOperator class now has a diff_volume_method argument that specifies how the volume of new materials should be determined. (#2691)

  • The openmc.DAGMCUniverse.bounding_region() method now has a padding_distance argument. (#2701)

  • A new openmc.Material.get_decay_photon_energy() method replaces the decay_photon_energy attribute and includes an ability to eliminate low-importance points. This is facilitated by a new openmc.stats.Discrete.clip() method. (#2715)

  • The openmc.model.Model.differentiate_depletable_mats() method allows depletable materials to be differentiated independent of the depletion calculation itself. (#2718)

  • Albedos can now be specified on surface boundary conditions. (#2724)

Bug Fixes

  • Enable use of NCrystal materials in plot_xs (#2435)

  • Avoid segfault from extern “C” std::string (#2455)

  • Fix several issues with the Model class (#2465)

  • Provide alternative batch estimation message (#2479)

  • Correct index check for remove_tally (#2494)

  • Support for NCrystal material in from_xml_element (#2496)

  • Fix compilation with gcc 5 (#2498)

  • Fixed in the Tally::add_filter method (#2501)

  • Fix meaning of “masking” for plots (#2510)

  • Fix description of statepoint.batches in Settings class (#2514)

  • Reorder list initialization of Plot constructor (#2519)

  • Added mkdir to cwd argument in Model.run (#2523)

  • Fix export of spherical coordinates in SphericalMesh (#2538)

  • Add virtual destructor on PlottableInterface (#2541)

  • Ensure parent directory is created during depletion (#2543)

  • Fix potential out-of-bounds access in TimeFilter (#2532)

  • Remove use of sscanf for reading surface coefficients (#2574)

  • Fix torus intersection bug (#2589)

  • Multigroup per-thread cache fixes (#2591)

  • Bank surface source particles in all active cycles (#2592)

  • Fix for muir standard deviation (#2598)

  • Check for zero fission cross section (#2600)

  • XML read fixes in Plot classes (#2623)

  • Added infinity check in VolumeCalculation (#2634)

  • Fix sampling issue in Mixture distributions (#2658)

  • Prevent segfault in distance to boundary calculation (#2659)

  • Several CylindricalMesh fixes (#2676, #2680, #2684, #2710)

  • Add type checks on Intersection, Union, Complement (#2685)

  • Fixed typo in CF4Integrator docstring (#2704)

  • Ensure property setters are used in CylindricalMesh and SphericalMesh (#2709)

  • Fix sample_external_source bug (#2713)

  • Fix localization issue affecting openmc-plotter (#2723)

  • Correct openmc.lib wrapper for evaluate_legendre (#2729)

  • Bug fix in Region.from_expression during tokenization (#2733)

  • Fix bug in temperature interpolation (#2734)

  • Check for invalid domain IDs in volume calculations (#2742)

  • Skip boundary condition check for volume calculations (#2743)

  • Fix loop over coordinates for source domain rejection (#2751)

Contributors