Hewitt, Bluck & Walker (Imperial) (i) Water reactor design basis accident studies (Large break LOCA, re-flood, interactions of thermal hydraulics and mechanical deformation of the core) (ii) Inorganic salt deposits (‘crud’) on PWR fuel ((Altered heat transfer modes,, thermal hydraulic y and neutronic effects)) (iii) Computational methods development for elastodynamic NDT Geoff Hewitt with more details on (i)
Academic Staff Geoff Hewitt (Chem Eng), Simon Walker, Mike Bluck (Mech Eng) KNOO ssupported: pported Colin Hale, Despoina Chatzikiriakou, Caroline Masson, Soleman Maudarboucas, Jessy Zeng, Panos Sfikas KNOO has already attracted others to it (free!) Has caused us to be invited to join two IAEA CRP CRP’s s. Two particular cross-Workpackage topics (both WP2: high Nb clad, NDT)
Water reactor design basis accident studies
Analysis of the reflood process: •
Multi pin rather than single pin models (heterogeneity) Multi-pin
•
Coupled Co pled anal analysis: sis pin ballooning modifies coolant flo flow passages passages, and hence ffurther rther cooling, and hence further ballooning…..
Coupling of multi-pin structural mechanics and threedimensional transient two p phase thermal hydraulic y analysis y for the study of design basis accidents -develop models to incorporate the systematic and stochastic heterogeneity of pin ballooning under reflood conditions -reduction of conservatism implicit in single-pin models would relax operating envelope constraints -particularly p y important p in context of drive for higher g burnups p -better understanding is crucial to permit use of advanced, high ductility clad
-The multi-pin model -Flow diversion studies -Droplet field modelling -Fundamental studies of rewetting -High Nb cladding materials (WP 2) Collaborations: BE, BE Serco & Nexia Nexia, USNRC (TRACE (TRACE, ‘CAMP’ CAMP agreement)
Ballooning computed for a PWR subassembly undergoing a loss of coolant accident. Systematic and stochastic differences between pins have been incorporated. Distinct differences in the ballooning behaviour of pins is clear, clear as is the absence of large regions of coherent ballooning. ballooning
Crud •
Crud (Metal oxides, NiO, NiFe2O4, ) deposits on fuel pins
Very close collaboration with Nexia & BE
Crud Issues in PWRs - Problems • Hinders heat transfer • More activated material (58Co, 60Co …) around circuit • “Axial Offset Anomaly” problems (neutron flux & power depression towards the top of the core) • Power output reduced • More fuel f l failures f il
• Surface heat transfer mechanism – Wick Boiling
Crud Modelling g
CLADDING
From the Nexia work: • Understand basic AOA mechanism • Wick boiling/chemistry model – explains lot of crud scrape data
Now: • Need to understand crud deposition mechanism better • Need to understand the thermal hydraulics of crudcoated cores
Thermal hydraulics of crud-coated cores •
•
Extent/distribution of boiling - Local thermal hydraulics (heat transport within crud) - Bulk thermal hydraulics - Coupling Wick boiling model: >99% heat removed by evaporative losses
•
Heat transfer correlations – clean surfaces
•
More boiling on crud covered surface compared to clean surface ?
•
What are the most appropriate heat transfer correlations ?
•
How are the local and bulk heat transport processes coupled ?
•
Possible crud formation experiments? p
Computational methods for elastodynamic nuclear NDT
QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture.
•
• •
QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture.
Essentially E ti ll allll acoustic ti NDT is i about b t sending di acoustic ti waves into i t structures, t t and d seeing i how h the th behaviour of the wave is modified by the geometry of the structure (ie by whether or not it contains a defect) To be able to predict how any given structure will cause a wave to scatter is an aid in understanding and interpreting observations Recent advances in other areas may have made possible a step increase in the capability to model wave fields over multi-wavelength domains.
C Cost off scattering i analyses l
•
Depending on the numerical approach, conventional methods have costs that rise with a high power, typically the 5th or 6th power of the size (in wavelengths) of the region analysed power, analysed.
•
If we can just about manage say a 5 wavelength region, a 50 wavelength l th one will ill b be up tto 1 1,000,000 000 000 times ti more expensive(!) i (!)
2 10 10
23,000 centuries!
IEFD FDTD IEFD-FMM
Time e (CPU hrs)
1.5 10 10
1 10 10
5 10 9
1 000 centuries! 1,000 t i ! 0 0
100
200
300
Length (wavelengths)
400
500
R d stealth Radar t lth
• • • •
Almost identical, but worse, cost and cost scaling problems arise in radar stealth modelling. Need to compute the scattered field in the space surrounding some target (in the free space around a plane; comparable to the material surrounding a defect) Targets (aeroplanes) are tens to hundreds of wavelengths long Existing methods could just manage a plane say 5 wavelengths long
Wh t iis now possible: What ibl
2 10 10
23,000 centuries!
IEFD FDTD IEFD-FMM
Time e (CPU hrs)
1.5 10 10
1 10 10
5 10 9
1 000 centuries! 1,000 t i ! 0 0
100
200
300
Length (wavelengths)
400
500
10 10
1,000 centuries 10 8
IEFD FDTD IEFD-FMM
Time (CPU hrs)
10 6 10 4 100 1 0.01 0.0001 10 -6 1
10
100 Length (wavelengths)
1000
Presentt project P j t is i attempting tt ti to t develop d l a frequency domain elastodynamic analysis too, employing the ‘fast’ methods mentioned, for nuclear (and more general) NDT applications.