Fracture assessment for reactor circuit (FRAS)

The objectives of the project for fracture risk assessment comprise (i) calculation of design and unforeseeable loads and their effects on a structure by applying numerical modelling; (ii) development of advanced fracture mechanics assessment tools and analysis methods based on material characterisation, damage mechanisms models and structural performance, in order to control structural failure both in cases of postulated initial flaw and environmentally assisted (internal) material damage; (iii) determination of degradation in material properties during service.

During recent years several structural analysis assessment methods have been established.

Nonetheless, need for more accurate methods in load definition, true 3D flaw assessment, as well as irradiation embrittlement evolution assessment, still remains. Traditionally, numerical simulation of loads has treated different components more or less as separate even though they belong to the same aggregate, thereby disregarding the interaction of support loads on the entire system. Applicability and limits of sub-modelling techniques in numerical simulation of crack growth need to be investigated further. Feasible ‘engineering assessment tools’ cannot be reliably applied, unless they have been tailored and verified for the particular plant and component in question. Traditionally, material’s fracture toughness estimates have been determined applying deep-notched ‘high-constraint’ specimens enabling conservative estimates to be derived. Sophisticated constraint corrections are required for shallow surface cracks with lower constraint, particularly in the case of asymmetric crack fronts during crack growth. Present limitations for specimen’s measuring capacity in fracture resistance testing and under ductile crack growth are presumably unrealistic and hence need revision. Unified model for irradiation embrittlement is still lacking, despite of intensive previous research.

Recently, advanced multi-scale modelling techniques of material damage micro-mechanism have allowed micro-scale investigation of relevant damage mechanisms, without a necessity to postulate a pre-existing flaw. This provides means for more realistic material damage assessment for environmentally assisted failure, such as stress corrosion cracking, irradiation embrittlement, ageing embrittlement and hydrogen embrittlement that do not require the existence of pre-existing flaw. In conjunction with modern FEM structural analysis methods, the application of these advanced modelling techniques enables structural integrity assessment of a component, or structure, over an entire ‘chain’ from micro- to macro-scale on a more realistic basis than previously.

Specific goals in 2008

The project was realised in three sub-projects: 1) Definition of loads, 2) Advanced fracture mechanical assessment methods and 3) Advanced surveillance techniques.

The external loads transferred to the reactor circuit components by supports were studied by simulating numerically the behaviour of a selected component under some unforeseeable, critical accident situation. The stiffness of pipes and its supports were calculated with FEM and substituted with simpler special purpose elements. A pipe break was further simulated with different kinds of models which included more non-linearities in the models.

Computational fluid dynamics analysis CFD was performed for studying temperature

fluctuations in the interface between cold and hot water, i.e. thermal striping, in a horizontal

feed-water pipe. The Large-Eddy Simulation (LES) and Detached-Eddy Simulation (DES)

turbulence models will be used. Results of the simulations will be compared with a simplified

validation case and briefly with a HDR (HeißDampfReaktor) experiment.

Weld residual stress (WRS) procedures were implemented in numerical simulations to reactor circuit components: RPV nozzles, safe-ends and connecting pipes. The models were created and elastic-plastic analyses performed with suitable FEM based analysis code, both 2D and 3D aprroached were applied. WRS definitions with, and with-out, PWHT were considered. The FEM analyses has been clarified how WRS distributions alter/decrease over the years during plant operation due to various yearly transient load cases. Crack growth in welds including the WRS distributions have been examined with fracture mechanics based analysis tool VTTBESIT, with stress distributions taken from results of the FEM analyses. Thus obtained analysis results has been compared against each other and against some handbook solutions.

Figure 1. Calculated temperature distribution near RPV inlet nozzle in a HDR experiment.

The development work on advanced fracture mechanical assessment methods consisted of finite element assessment tools and fracture mechanics analysis methods based on material characterisation, damage mechanisms models and structural performance. The aim was to increase understanding on the behaviour of postulated initial flaws (surface-flaws) and, on the other hand, on environment assisted material damage like irradiation embrittlement and stress-corrosion cracking. The reactor circuit integrity is to be further studied by analysing cracked T-junctions with 3D FEM model. The solution for integrity assessment of T-junction is sought by using FEM (ABAQUS, ANSYS) and integrating its application to existing geometric and loading data, e.g., plant database. Firstly, the needed initial data has been specified and the inter-faces to model the geometries and define the input data was studied. A case computation for a T- joint was performed.

Transferability of fracture mechanics test data associated with different levels of specimen’s

constraint was to be investigated by performing fracture mechanics tests using specimens with

both deep and shallow surface notches. The task includes fracture mechanics tests on selected

materials using surface cracked specimens at various degrees of tension and bending. This

should provide input for future development of FEM analysis methods taking into account

constraint effects in structural analysis. Numerical work was performed to assess the fracture

toughness transition between standard CT and SENB -type specimens and 3D surface cracks

for cleavage initiation and propagation using the WST cleavage fracture model. The

assessment clarifies how fracture toughness is affected by different crack types as well as

'realistic' crack features (such as asymmetric crack front). Micromechanical modelling of cleavage fracture was to be performed using multi-scale models and the Master Curve method. The WST model implementation was finalized. Development of models for predicting cleavage fracture toughness will be accomplished using micromechanical modelling methods. The development of sub-modelling techniques includes a comparison of crack growth results obtained with and without sub-modelling in some typical NPP applications, such as RPV nozzle and safe end. The limits of sub-modelling techniques were defined in the first part of the study.

Fracture resistance measurements on different materials using various specimen types were conducted to produce fracture resistance data in the ductile area to be applied in numerical calculations on stress-strain states. The overall aim was to define realistic criteria for specimen’s measuring capacity. Improved quantitative models were also to be developed for describing the microstructural changes due to irradiation using the embrittlement data from previous model alloy studies. As part of this work, an extensive VVER-440 surveillance data base was analyzed applying recently developed, advanced non-linear methods.

Deliverables in 2008

• Modelling methods (ABAQUS) for different types of supports and restraints were delineated and critical accidents cases reviewed. Stiffness of pipes and its supports were calculated with FEM and substituted with simpler special purpose elements. A pipe break was further simulated with different kinds of models which included more non-linearities.

The results were compared with each other and their reliabilities evaluated.

• Validation data for Computational Fluid Dynamics (CFD) calculations of turbulent mixing layers has been searched from the literature. Simulations have been performed in a simplified validation case by using different turbulence models and comparison with the experiment has been made. Calculations of the HDR experiment have been carried out by using different geometries and meshes. The vortex method has been implemented into the Star-CD code for generating turbulent inlet conditions for Large-Eddy Simulation (LES).

Results obtained with the vortex method have been validated for fully developed pipe flow.

• Relevant weld residual stress (WRS) procedures / standards / codes have been implemented to numerical simulations concerning reactor circuit components. WRS definitions with and without post weld heat treatment (PWHT) are considered. The selected components are a feedwater nozzle and connecting safe-end and pipe in a Finnish BWR RPV: the examined locations are the two circumferential welds joining these components. The models were created and analyses performed with FEM code ABAQUS 2D with elastic-plastic material properties. Analyses consider typical anticipated yearly plant load transient cases, and the aim is to see how the WRSs alter as their function over the simulated years in operation.

Crack sensitivity analyses including WRSs with fracture mechanics based analysis tool VTTBESIT were performed as well.

• The Zencrack code, working like a “cracked submodel”, was studied and tested. Among others a former benchmark computation concerning a possible fast fracture of a RPV was studied. After basic applications, a sample T-joint, generated by macros running with ANSYS FE code, was studied. A procedure to run a cracked model together with ANSYS and Zencrack code was created.

• The WST cleavage fracture model implementation has been finalised in co-operation with

the ‘Perfect’–IP. The model was applied to numerically compute the shape of the fracture

toughness ductile-to-brittle transition, i.e., the ‘Master Curve’. Work as of late was focusing on determination of WST model parameters, which resulted in computational demonstration of both the (i) statistical size effect and (ii) temperature dependence to be in line with the experimentally determined Master Curve.

• The experimental programme (fracture mechanical and tensile testing) and the test specimens for the fracture resistance measurements on different materials were in part completed. The numerical modelling and computations for analysing specimen’s measuring were reported as a MSc(Tech) thesis.

• Analysis of extensive data-base of VVER-440 surveillance data was performed applying recently developed, advanced non-linear methods. Non-linear analyses of the surveillance data were completed and reported as a conference paper. Participation to the IGRDM meeting was carried out.

• A detailed test matrix on ATOM Probe and PA characterisation of irradiated materials was agreed with Tohoky University. Samples representing weld 501 material in eight different IAIA-conditions and ten different model alloys in irradiated conditions were prepared by VTT with EDM and transported to Japan.

2.6.6 Influence of material, environment and strain rate on environmentally