INL/RPT-25-87633 Implementing Fuel Cladding Bonding and Nuclear Energy Advanced Modeling and Simulation Assessing its Impact on Axial Gas Communication S E P TEMBER 2025 Kyle A. Gamble and Daniel J. van Wasshenova Idaho National Laboratory Wenfeng Liu, Joe Y. R. Rashid, and Deena Jaber Structural Integrity Associates This information was prepared as an account of work sponsored by an agency of the U.S. Government. Neither the U.S. Government nor any agency thereof, nor any of their employees, makes any warranty, expressed or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness, of any information, appara- tus, product, or process disclosed, or represents that its use would not infringe privately owned rights. References herein to any specific com- mercial product, process, or service by trade name, trade mark, man- ufacturer, or otherwise, does not necessarily constitute or imply its en- dorsement, recommendation, or favoring by the U.S. Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the U.S. Government or any agency thereof. INL/RPT-25-87633 Implementing Fuel Cladding Bonding and Assessing its Impact on Axial Gas Communication Kyle A. Gamble and CONTENTS ABSTRACT iii ACKNOWLEDGMENTS iv ACRONYMS AND GLOSSARY ix 1. INTRODUCTION 1 2. CONSTITUTIVE MODELS FOR FUEL CLADDING BONDING 2 3. AXIAL GAS COMMUNICATION 4 4. COUPLING THE MODELS TOGETHER 6 4.1. Preliminary Coupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 4.2. Tight Coupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 5. VERIFICATION AND DEMONSTRATION 8 5.1. Implementation Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 5.1.1. Mechanical Implementation Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 5.1.2. Thermal Implementation Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 5.2. Full-Length Rod Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 5.2.1. LOCA Input and Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . 12 5.2.2. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 6. CONCLUSIONS 18 7. FUTURE WORK 18 8. PUBLICATIONS 20 9. REFERENCES 21 FIGURES Figure 1. Schematics from Khvostov et al. [11] showing the two different flow paths, where the hydraulic layer is either calculated as a function of the fuel-cladding gap or through permeable fuel. The BISON implementation has been designed to allow both pathways concurrently. . 5 Figure 2. A Moody diagram showing the regions of laminar flow, the transitional region and the turbulent flow as a function of the Reynolds number (Re). For fluid flow to be termed laminar, the relationship f = Re 64 must be met. . . . . . . . . . . . . . . . . . . . . . . . . 7 Figure 3. Final contour plots for the cracking function mechanical implementation test for (a) displace- ment, (b) packing ratio, and (c) and fracture status. . . . . . . . . . . . . . . . . . . . . . . 10 Figure 4. Verification of thermal conductivity of individual gaseous constituents. Solid lines corre- spond to BISON calculations where the symbols are analytical values. . . . . . . . . . . . 10 Figure 5. Thermal conductivity as a function of packing ratio for a solid fuel and gas mixture at 1000 K. 11 Figure 6. Axial power factor of a hot fuel assembly. . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 vi Figure 7. Cladding surface temperature history during loss of coolant accident (LOCA). i identifies the equally spaced axially position starting from the bottom of the fuel rod. . . . . . . . . . 13 Figure 8. Schematic of the R-Z mesh used for the modeling of bonded fuel. . . . . . . . . . . . . . . 14 Figure 9. Plenum pressure under different gas flow rates. . . . . . . . . . . . . . . . . . . . . . . . . 15 Figure 10. Gas pressure in fuel stack under different gas flow rates. . . . . . . . . . . . . . . . . . . . 15 Figure 11. Clad radial displacement vs. axial position with fully blocked gas flow. . . . . . . . . . . . 16 Figure 12. Clad radial displacement vs. axial position with different gas flow rates (a) = 1.0×10−4 mol/s, (b) = 1.0×10−5 mol/s, and (c) = 1.0×10−6 mol/s. . . . . . . . . . . . . . . . . . . 16 Figure 13. Evolution of gas volume in the (a) plenum and (b) fuel stack. . . . . . . . . . . . . . . . . 17 Figure 14. Clad radial displacement vs. axial position with high gas flow rate (1.0×10−3 mol/s). . . . 17 Figure 15. Evolution of the packing ratio as a function of axial position and time for the bonding informed axial gas communication simulation. . . . . . . . . . . . . . . . . . . . . . . . . 18 Figure 16. The effects of implementing a preliminary bonding model within AxialGasCommunication. (a) The rate of change in the number of moles based on a scale factor applied to the limiting factor in Equation (16). (b) The pressure post-rupture as a result of the limiting factor described in (a). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 TABLES Table 1. Coefficients for thermal conductivity of gases considered in the bonding model. . . . . . . 4 Table 2. Fuel rod and operating parameters for the base irradiation. . . . . . . . . . . . . . . . . . . 11 Table 3. Key results from base irradiation for loss of coolant accident (LOCA) run initialization. . . 12 vii Page intentionally left blank viii ACRONYMS BISON MOOSE-based multidimensional fuel performance code CIVET Continuous Integration, Verification, Enhancement, and Testing EOL end-of-life INL Idaho National Laboratory LOCA loss of coolant accident LWR light-water reactor MOOSE Multiphysics Object Oriented Simulation Environment NEAMS Nuclear Energy Advanced Modeling and Simulation PWR pressurized-water reactor ix Page intentionally left blank x Implementing Fuel Cladding Bonding and Assessing its Impact on Axial Gas Communication 1. INTRODUCTION As part of the life cycle for nuclear fuel rods, high pressures due to the continual creation and containment of gases are an area of concern for safe operation as industry pursues higher discharge burnups. The performance of high-burnup fuel rods containing restructured fuel and higher rod internal pressures during an accident scenario, such as a loss of coolant accident (LOCA), remains an active area of research. Initially, the free volume of the fuel rod consists of a plenum, voids and cracks within the fuel, and the fuel-to-cladding gap acts as a reservoir for the fill gases at an equilibrium pressure. As irradiation progresses, fission gases are released into the free volume and impact the pressure along with dimensional changes that consume some of the internal volume. At higher burnups above 45 MWd/kgU [1], a chemical and mechanical bonding between the fuel and cladding can form. This bond changes the physical properties of the system, limiting the free volume for gases by preventing separation of the fuel and cladding during power decreases. Additional cracking within the fuel or the breaking of the bond due to tensile stresses may result in the formation of new pathways for the contained gases to flow. The flow of the gas within the rod due to pressure gradients that form is referred to as axial gas communication. Axial gas communication occurs throughout irradiation. Early in life, the fuel-to-cladding gap is open, providing an almost instantaneous transport of released gases, which result in fast equilibration times. After bond formation, the gas flow between different regions of the rod that have different pressures is severely impacted as flow is restricted through the fueled region that has a relatively low permeability. The importance of reduced gas communication is magnified during a LOCA. Strong pressure differentials may be present that lead to localized ballooning of the cladding due to the impeded gas flow. As ballooning occurs, the pressure gradient may increase causing further ballooning and potentially leading to cladding rupture. Coupling this with large amounts of transient fission gas release potentially being released during the LOCA, the rod internal pressure will continue to increase. If rupture occurs, the decay of the high plenum pressure to the system pressure depends upon the severity of flow blockage. A strong pressure differential across severely fragmented fuel may lead to dispersal of the fragments to the primary coolant system. On the other hand, the presence of a fuel cladding bond may reduce ballooning and improve coolability of the rods. Therefore, the ability to capture both phenomena in fuel performance simulations of high-burnup fuel subjected to a LOCA transient is of interest. This report summarizes activities undertaken to begin the coupling of a fuel cladding bonding and axial gas communication models in the BISON ¹ code [2] to assess high-burnup fuel rod performance. The report is structured as follows: Section 2 presents an overview of the theory underpinning the implemented fuel cladding bonding model, Section 3 summarizes past developments of the axial gas communication model available in BISON, Section 4 describes the rudimentary coupling strategy adopted in this work as well as provides details into the steps required to ensure a tight coupling between the two models, and Section 5 highlights the implementation testing of the bonding model and the LOCA demonstration problem to assess impact of including bonding effects in axial gas communication studies. The report concludes with sections summarizing the key findings of the work (see Section 6), a collection of developments necessary for future work (see Section 7), and a list of publications related to this work from the current research year (see Section 8). ¹BISON is a MOOSE-based multidimensional fuel performance code. 1 2. CONSTITUTIVE MODELS FOR FUEL CLADDING BONDING This section presents the finalized constitutive models for high-burnup fuel with bonding, which have been incorporated into the main BISON code branch, building on the earlier implementation described in Reference [3]. Within a fuel rod due to the combined effect of irradiation and the external pressure imposed by the coolant, the fuel experiences an outward expansion, while the cladding undergoes a inward reduction in radius. The mechanisms causing these dimensional changes within the fuel include thermal expansion, gaseous and solid fission product swelling, densification, cracking, and creep. Within the fuel, the mechanisms include thermal expansion, irradiation growth, and creep. These dimensional changes results in a reduction of the fuel cladding gap leading to the eventual contact between the two materials. After prolonged contact between the fuel and cladding, a bond may form. The model for capture the effects of fuel cladding bonding is broken into two parts: mechanical and thermal. The mechanical framework describing the fuel-cladding bonding is formulated by assuming bonding exists and was developed to work in a 2D axisymmetric or Cartesian geometry. The transition from a bonded to un-bonded state consists of a change from a solid state to a mixed gaseous/fragmented fuel state. The gaseous mixture has no mechanical rigidity unless the fuel fragments are compressed against each other. As such, the gas and solid mixture behaves similar to a mixture of gas and granular materials under hydrostatic pressure confined by the surrounding media. The gas in this situation contributes to the overall pressure exerted outward onto adjacent materials or structures. The implementation of the mechanical model for bonding in BISON is separated into two processes: (1) the calculation of stresses and (2) the derivation of the tangent stiffness. The stress computation in the fuel material depends on the fuel cracking state, which is described using a state variable α (1 for cracked fuel and 0 for solid). The model does not differentiate between the initial occurrence of cracking and the extensive fragmentation that may occur in a LOCA. In either case, once cracked, the element is treated as a hydrostatic entity. Three criteria, in the order of precedence, are used in the code to initiate the fuel cracking state: 1. Prescribed Function: A prescribed function can be used as an input to describe the state of the fuel, either solid or cracked. 2. Temperature-based Criterion: Describes fuel cracking by T > Tth , where T is the fuel temperature and Tth is a temperature threshold defined in the user input. A suitable value for Tth is provided by Walker et al. [4]. 3. Strength-based Criterion: Fuel cracking is determined by its maximum principal stress, and it is described as max(σ1 , σ2 , σ3 ) ≥ σ f , where σ f is the fracture strength provided by the user and σi (i=1, 2, and 3) is the principal stress in direction i. A suitable fracture strength for UO2 can be found from a variety of references [5–8]. The effective elastic modulus E is determined by E = (1 − α) E0 − 3αP(1 − 2ν), (1) where α is the fuel cracking state, P is gaseous pressure, and E0 is the elastic modulus of the fuel material. The shear modulus is computed using the input Poisson’s ratio and the effective Young’s modulus. An isotropic elasticity tensor is constructed for computing the Jacobian matrix. 2 When the material is in a solid state, the stress is simply calculated using the elastic material properties. In a “cracked” material state (α = 1), the hydrostatic pressure is computed in BISON under the constraint of an instantaneous pressure equilibrium in a selected block of material. The shear stress components in such a state are optional. With regard to the pressure calculation, the temperature, volume, and mole of gas are distributed in each separate element volume. Assuming the fuel packing ratio (the ratio of solid fuel material to the total volume) in each element is λi , the total gas volume is calculated by subtracting the solid volume (determined by the packing ratio) from the element’s total volume VT = ∑(1 − λi )Vi . (2) i The ideal gas law yields P (1 − λi )Vi P VT n = ∑ ni = ∑ = (3) i R i Ti R Te f f 1 (1 − λi )Vi Te−f 1f = ∑ VT i Ti . (4) The internal pressure is RTe f f VT ∑ P= Ni . (5) i At each time step, the volume packing ratio is updated by λ0i ρ0f λi = . (6) ρi where ρ f is the density of fuel, superscript 0 represents the initial state of each variable, and subscript i represents the ith element. Implementation of the above mechanical behavior models for the fuel cladding bonding behavior in BISON is achieved through two classes: ComputeFuelStress and ComputeFuelElasticityTensor. Several regression tests have been added to verify that the method described above can sufficiently model the pressure load from solid gas mixture to its surrounding cladding (described in Section 5.1). The thermal aspects of the bond, in its solid, mixture, or gaseous state, uses a similar model to existing correlations already available within BISON. The model for thermal conductivity of the fuel fragments following the equations by Loeb [9] and Kämpf and Karsen [10], is given by ! (k f − k g )λ ke f f = k g 1 + , (7) k g λ1/3 + k f (1 − λ1/3 ) where k f and k g are the thermal conductivity of solid fuel and gas, respectively. The effective specific heat is computed using the rule of mixtures. It depends on the proportion of each component and is expressed as the weighted average of the specific heat of fuel and gas materials. C p, f ρ f λ + C p,g ρ g (1 − λ) C p,e f f = , (8) ρ where C p, f is the specific heat of fuel material and C p,g is the specific heat of gas material. The density of the mixture is ρ = ρ f λ + ρ g (1 − λ ). (9) 3 At present, four different gases are implemented in the model for specific heat and thermal conductivity using properties from MATPRO [5] including Helium, Argon, Krypton, and Xenon, which is consistent with the available options in the axial gas communication model. For the solid UO2 , the Halden thermal conductivity model for high-burnup fuel in BISON is used. The thermal effects of the bond are implemented in a model a class called ThermalGap. From Loeb [9], the thermal conductivity as a function of temperature for the gaseous species is given in the form of k = AT B , (10) where A and B constant and T is the temperature in K. Table 1 gives the coefficients for the four gases considered by the fuel cladding bonding model. Table 1. Coefficients for thermal conductivity of gases considered in the bonding model. Coefficient Helium Argon Krypton Xenon A 2.639 × 10−3 2.986 × 10−4 8.247 × 10−5 4.351 × 10−5 B 0.7085 0.7224 0.8363 0.8616 Similarly, the specific heat capacity is for a single gaseous constituent is given by A0 cp = , (11) M where A0 is a constant given in J/mol-K and M is the molar mass of the gas. For all gases considered in the bonding model, A0 = 20.786 J/mol-K. The molar masses of helium, argon, krypton, and xenon are 4.003, 39.95, 83.79, and 131.3 g/mol, respectively. The value obtain after substitution into Equation (11) is multiplied by 1,000 to obtain the SI units necessary for use in BISON. If there is a gas mixture, the thermal conductivity and specific heat are weighted based upon the gas fraction. The summation of the gas fractions must equal 1.0. 3. AXIAL GAS COMMUNICATION Axial gas communication concerns the computation of local pressure as a function of axial position due to a computed mole rate introduced due to pressure differentials. Gas no longer communicates in a given step once pressure equilibrium is reached. Using Khvostov et al. [11] as a numerical foundation, an algorithm to model gas flow within BISON under post-rupture conditions during a LOCA event was developed. Knowing that information to determine the permeability or ability for gas to flow will need to come from the axial fuel relocation model that already existed in BISON, which uses a Layered1D or Layered2D geometric formulation, a Layered1D formulation was also adopted for the axial gas communication model. Utilizing the same geometric representation enables the easier transfer of relevant quantities between the models. Notable differences between the BISON and Khvostov implementations for gas communication are the discretization of the gas flow rates and pressures as a function of height. The use of axial layers in Layered1D formulation allows for other physical phenomena to be modeled in BISON, such as double ballooning events, bonded regions, and varying fuel permeability due to cracking and fragmentation, which cannot be captured by the Khvostov framework. The starting point of the model uses Poiseuille’s law [11], which assumes a laminar, stationary flow of incompressible uniform fluid. This allows the conservation laws to be fulfilled while reducing the complexity of the equations involved. The method followed by Reference [11] was primarily to find the flow rate of atoms, 4 Ṅ, between the plenum, with pressure P1 , and the distended (ballooned) region, with pressure P2 given by the equation π P12 − P22 , (12)  Ṅ = 256IHD where the integrated hydraulic diameter function, IHD , is Z zrup + Le f f ηkT IHD = dz, (13) zrup Dh4e f f (z) where zrup is the axial location of the rupture and Le f f is the effective length between the plenum and ballooning region. The other parameters are defined as follows: η is the dynamic viscosity, V the volume, T the local temperature, k the Boltzmann constant, and Dhe f f the effective hydraulic diameter at position z. Khvostov implementation for emergent paths can be seen in Equation (14) and Figure 1. Within this model, Ṅ0 represents the mole flow rate through the initial internal paths, while δ Ṅ describes the rate through the emergent paths that form during the ballooning process. The emergent paths can be taken as one of the following two conditions: (1) a change in the gap distance between the fuel pellet and cladding or (2) through the permeable fuel (e.g., cracks, voids) if the fuel pellet has expanded and entered a bonding phase. This implementation assumes that the rate is dominated by the gap contribution over the passage through the fuel as can be proven mathematically. πDh4 e f f (z) PdP 4 πDh0 PdP πδDh4 PdP (14)  Ṅ = − = Ṅ0 + δ Ṅ = − − . 128ηkT dz 128kT dz 128kT dz Figure 1. Schematics from Khvostov et al. [11] showing the two different flow paths, where the hydraulic layer is either calculated as a function of the fuel-cladding gap or through permeable fuel. The BISON implementation has been designed to allow both pathways concurrently. In the BISON implementation, both emergent paths can be accounted for simultaneously, meaning that the change in hydraulic diameter δDh4 is dependent upon a combination of the fuel/bond state and the fuel-to-cladding-gap state. In addition, the pressure differential is more axially refined to account for local variations in the hydraulic diameter, temperature, and dynamic viscosity between the plenum and ballooned regions. The hydraulic diameter within the fuel is a function of the fuel packing ratio in situations where fuel has fragmented and relocated. For the fuel-to-cladding gap, the hydraulic diameter depends upon the size of the fuel-to-cladding gap. In the BISON implementation the dynamic viscosity is computed based upon the properties for four common gases present in nuclear fuel rods: helium, xenon, krypton, and argon. Helium is the traditional fill gas in light-water reactor (LWR) fuel rods, whereas xenon and krypton are the most common fission gases produced. Argon is generally used in refabricated experiments to simulate the conductivity degradation of fission gases that have been removed. 5 4. COUPLING THE MODELS TOGETHER The inclusion of a fuel cladding bonding model in fuel performance simulations is the most impactful when considering its influence on axial gas communication phenomena both before and after cladding rupture. From an experimental point of view, there is more data available for the post-rupture scenario, and this has been the focus of the initial implementation of the axial gas communication model [12]. Therefore, in the preliminary coupling strategy adopted in this work, the focus has been on the impact of the bond on post-rupture depressurization compared to the standalone no-bonding model cases. The tight coupling approach described below, which needs substantial further development to enable, requires the axial gas communication model to be extended to include pre-rupture gas flow in scenarios where complete blockage is not present. This will enable more accurate calculations of the ballooning process subjected to local pressure effects and the fuel bond. Ultimately, with the industry’s shift in focusing on a no-rupture condition from the minimization of fuel dispersal after a burst event, this is the required end goal for coupled gas communication and bonding modeling. Nevertheless, with the wealth of data available in the post-rupture regime from a wide variety of experimental programs, it is still useful to consider the post-rupture case when assessing the impacts of the coupling strategies. 4.1. Preliminary Coupling The preliminary coupling strategy is a one-way transfer of information from one model to the next in a staggered execution of separate simulations. Looking at the models as described, the fuel cladding bonding model provides a means of calculating the packing ratio for flow through the bonded region, which is different from the packing ratio computed from the fuel relocation model. The first is computed from assuming a mixture of fuel and gas upon rupture breakage, whereas the latter is computed from an estimation of fragment sizes due to fuel pulverization. From a gas communication perspective, the flow through the corresponding region is directly impacted by this packing ratio. Therefore, the preliminary coupling algorithm is to execute the standalone bonding model and import as a function of axial location and the time that the packing ratio computed into the axial gas communication simulation. One will note that the bonding model also contains a constant user-defined gas flow rate, which indirectly affects the packing ratio predictions. Therefore, after running the bonding informed axial gas communication simulation, it is prudent to compare the calculated gas flow rates to the selected input for the bonding model. If there is a large discrepancy, the models should be run again using more appropriate values for the gas flow rate. With Poiseuille’s equation, it is assumed the flow in the AxialGasCommunication is incompressible and laminar. Similarly, the generalized form of Bernoulli’s equation, which is also for incompressible flows, can be written as v2 P + gz + = constant. (15) 2 ρ In the limit of small height differences (∆z → 0), this can be taken as a functional “speed limit” with respect to the pressure gradient, through the dynamic pressure, in the form 1 2 q = ∆P = ρv , (16) 2 where ∆ P is the pressure gradient between two adjacent layers, ρ is the density of the fluid, and v the average fluid flow rate. By using Equation (16) to limit the effective loss in pressure, the ratio of fluid flow rate to the speed of sound in the medium—the Mach number (M = vc )— remains low, enforcing the incompressible conditions 6 for the Poiseuille’s equation and ensuring the conservation laws are met. In general, fluids with M  1 are considered to be incompressible [13]. A more efficient method of ensuring incompressible flow is through the use of the Reynolds number,R = η , where a value less than 3,000, or when considering frictional loss, a ρvl ratio of f < Re 64 should be met. This is traditionally visualized in a Moody diagram as seen in Figure 2. Figure 2. A Moody diagram showing the regions of laminar flow, the transitional region and the turbulent flow as a function of the Reynolds number (Re). For fluid flow to be termed laminar, the relationship f = Re 64 must be met. To match the testing methods utilized for the bonding method, an extensive preliminary manipulation of the AxialGasCommunication class was undertaken. As previously discussed, the bonding method was created in 2D-axisymmetric, with radial refinements with a single flow rate of plenum to rupture, while the axial gas communication model currently exists in a layered 1D model with flow rates for each axial layer. To best match the flow rate studies, only the topmost layer of the fuel stack layers was preserved instead of the several layers between the bottom of the plenum and the point of rupture and the “speed limit” discussed above was utilized. This was accomplished by reducing the number of layers within the fuel in such a manner that the topmost layer would be the layer in which the rupture occurred. Initially, the fuel was assumed to be bonded, and as it expanded, the fuel packing ratio from the bonding studies was used to give an effective permeability for the fuel. Since only the fuel elements nearest the cladding changed the packing ratio in the bonding study, it was assumed an effective new hydraulic diameter was established between the fuel and the cladding. Via this method, it was assumed the average hydraulic diameter was constant between the plenum and the rupture point. The hydraulic diameter can be expressed as Dnew_ f uel = rclad λi , (17) where λi is the packing ratio. Whereas, if the fuel is still bonded and if the gap between the fuel and cladding surpasses the limit currently within the algorithm for contact (2×10−6 m), then the expression becomes Dnew_ f uel = r f uel λi . (18) When simulating the standard approach, the same packing ratio data was used; however, the fuel was considered fully bonded, only allowing gas to escape through fragmented fuel via the equation aε V Dh f uel = a , (19) 1 − ε V + Dclad 7 where Dclad and Dh f uel are the diameter of the cladding and hydraulic diameter of the fuel pellet at a given instance. The pellet hydraulic diameter, Dh f uel , is a function of the homogenized volume fraction of the emergent voids, ε V , and the effective average diameter, a, of the fuel fragments. The smeared void is expressed as Kγ εV = 1 − (1 + ε θθ )2 !2 γ D f ueli εV = 1 − , γ0 D f uel f where Kγ = γγ0 is the relative linear density of the fuel and the outer diameter of the fuel is taken to be D f uel f = (1 + ε θθ ) , (20) D f ueli as function of the current hoop strain ε θθ . When dealing with the fuel, the hydraulic diameter is for a circular tube, which is simply the diameter of the tube. This allows for the flow to be equated to the volume of the free space within the fuel. For this work, the free volume is assumed to be directly related to the packing ratio. The results of this rudimentary implementation are shown in section 5.2.2. 4.2. Tight Coupling In the tight coupling strategy planned for future work and required for enhanced LOCA behavior predictions within BISON, the models are run concurrently in the same simulation passing necessary information such as packing ratio and gas flow rates iteratively until a converged solution is achieved. To achieve this tight coupling, several enhancements to the fuel cladding bonding and axial gas communication models are required. These enhancements are outlined below: • Extend both the axial fuel relocation and axial gas communication models from Layered1D to 2D-RZ axisymmetric geometric representations • Extend the axial gas communication model to compute gas flow evolution prior to cladding rupture • Extend the bonding model to account for evolution of the gap before bond formation • Enable the ability to have radially varying permeability in axial gas communication. 5. VERIFICATION AND DEMONSTRATION This section focuses on the verification of the fuel cladding bonding model implementation in BISON to confirm proper coding and mathematical formulations compared to analytical solutions. In addition, a full-length rod demonstration problem is presented that evaluates the effects of the bonding model on axial gas communication predictions using the rudimentary coupling strategy. 8 5.1. Implementation Testing Implementation testing refers to unit and regression tests included with a code submission to BISON to ensure proper coding of the mathematical equations for the model of interest. These tests are designed to capture changes in behavior due to future code modifications through a continuous integration process. Through this system, adequate code coverage (i.e., tests exercising all available options) is ensured. BISON utilizes the Continuous Integration, Verification, Enhancement, and Testing (CIVET) [14] platform for its testing. This is the same platform utilized by Multiphysics Object Oriented Simulation Environment (MOOSE) and all other MOOSE-based applications. Unit and regression tests for the entire code base are executed when any change is proposed to the code. Here an overview of the implementation tests added for testing of the fuel cladding bonding model described in Section 2. Details of the implementation testing for axial gas communication can be found elsewhere [12]. The testing overview is broken down into two sections, one covering mechanical behavior and the other thermal behavior. 5.1.1. Mechanical Implementation Tests The regression tests for mechanical behavior of the fuel cladding bond are designed to test the ComputeFuelStress and ComputeFuelElasticityTensor models including under internal pressure load- ing, tension, and shear. These tests are developed for both 2D Cartesian and 2D-RZ axisymmetric geometries. Additional tests are added to verify that the different crack initiation (i.e., bond breakage) criteria (i.e., pre- scribed function, temperature-based criterion, strength-based criterion) are working as expected. Most of the regression tests are completed on a single 2D QUAD4 element in the Cartesian plane or assuming 2D-RZ ax- isymmetry. A few cases explore the impact of additional mesh elements using 1x10 and 10x10 QUAD4 meshes on the mechanical response. The tests above are included in the bison/test/tests/fuel_clad_bonding directory within the repository. As an example, let’s consider the cracking function test included in the repository to indicate bond breakage through a prescribed function. Here a 1-m by 1-m square is composed of 10 elements in both the x and y directions. In both directions, the position from 0 to 0.8 consists of solid fuel, 0.8 to 0.9 makes up the bonded region, and 0.9 to 1 indicates cladding. The temperature is ramped from 300 to 3000 K in 10 seconds. Cracking can only occur in the bonded region and is set to only occur after 2 seconds. The packing ratio is computed based upon the density of the region relative to its initial density. Deformation is driven entirely by the increase in pressure due to the increasing temperature. Figure 3 presents contour plots for the displacement magnitude, packing ratio, and fracture status at the end of the simulation. As the bond breaks, the pressure can begin to exert load to the cladding, and therefore, the cladding expands. The packing ratio decreases in the regions of highest deformation. Given the symmetric nature of the analysis, the maximum displacement and stress in the x and y directions are 0.1056 m and 63.88 MPa, respectively. 5.1.2. Thermal Implementation Tests The regression tests for thermal behavior of the fuel cladding bond are designed to test the ThermalGap model. These tests are available in the same location as the mechanical tests. This model computes a weighted average thermal conductivity and specific heat for the bonded region based upon whether it is solid, gaseous, or a mixture. For simplicity, regression tests for the individual gas components available assume that the gas is entirely made of that constituent. A single 2D QUAD4 element is subjected to a temperature ramp rate from 300 to 1000 K, and the temperature dependent properties are computed by BISON. Figure 4 illustrates the computed values from 9 (a) (b) (c) Figure 3. Final contour plots for the cracking function mechanical implementation test for (a) displacement, (b) packing ratio, and (c) and fracture status. BISON for gaseous thermal conductivity (i.e., solid lines) compared to the analytical values (i.e., symbols). As expected, there is perfect agreement. From Equation (11), it is observed that the specific heat is constant and independent of temperature for the gases. As expected, BISON computes the analytical values obtained from the equation: 5192.6 J/mol-K for He, 520.3 J/mol-K for Ar, 248.07 J/mol-K for Kr, and 157.7 J/mol-K for Xe. Figure 4. Verification of thermal conductivity of individual gaseous constituents. Solid lines correspond to BISON calculations where the symbols are analytical values. For solid fuel representing the bond, it was mentioned that the Halden thermal conductivity model in BISON is used. This model has been separately verified when it was added to the repository and is not revisited here. In the case of a mixture of solid and gas representing the bond, consider a single gas constituent mixed with the fuel. In this case, the temperature is held constant at 1000 K, and the packing ratio of the bond is increased from 0 to 1. A packing ratio of 0 represents pure gas, and a value of 1 indicates an intact bond. At 1000 K, it has been verified that the Halden fuel conductivity model gives a value of 2 W/m-K. The values for the various gases at 1000 K can be seen in Figure 4. Using these verified values, Figure 5 illustrates the 10 expected output from Equation (7) for the different gas UO2 mixtures. Figure 5. Thermal conductivity as a function of packing ratio for a solid fuel and gas mixture at 1000 K. 5.2. Full-Length Rod Analysis A pressurized-water reactor (PWR) fuel rod case was prepared as a demonstration of the impact of the various models metrics of interest. The fuel design and operation parameters are listed in Table 2. The base irradiation was simulated using BISON to create an end-of-life (EOL) condition used for initializing the LOCA runs. A few key parameters from base irradiation results are shown in Table 3. Table 2. Fuel rod and operating parameters for the base irradiation. Parameter Value Units Fuel Stack Height 3.658 m Cladding Length 3.851 m Cladding Outer Diameter 0.00914 m Cladding Inner Diameter 0.00800 m Fuel Pellet Outer Diameter 0.00784 m Fuel Pellet Length 0.00940 m Subchannel Equivalent Hydraulic Diameter 0.01295 m Internal Gas Pressure (at STP), He 2.4 MPa Fuel Grain Radius 0.000005 m Fuel Enrichment 4.5 % Initial Density 95.67 % TD Heat Deposited in Fuel 0.974 – Inlet Temperature 289.0 ◦C Coolant Pressure 15.5 MPa Mass Velocity 3690.0 kg/m2 /s 11 Table 3. Key results from base irradiation for LOCA run initialization. Parameter Value Units Burnup 67.9 GWd/tU FGR 3.69 % Rod Pressure 5.65 MPa Fractional Density 0.905 – Gas Volume 10.35 cm3 Gas Fraction (He/Xe/Kr) 0.76/0.22/0.02 – 5.2.1. LOCA Input and Boundary Conditions The rod average linear power of 27 kW/m is used for LOCA modeling with an axial power profile shown in Figure 6. An initial steady-state operation for 50 s is modeled to allow the fuel to reach thermal equilibrium. The LOCA is initiated at 100 s, and a calculated decay power is used for the duration of the LOCA. The cladding surface temperature history is shown in Figure 7. Nodes 0–11 represent equally spaced axial locations along the length of the fuel rod from the bottom to the top. The cladding surface temperature history is used as the Dirichlet boundary condition in BISON input for LOCA modeling. The cladding temperature peak (1179.3 K) occurs at 136.4 sec in Node 9, which is 3.0 m from the bottom of the fuel. The coolant pressure is initially 15.5 MPa. Depressurization starts at 100 s, drops to 1 MPa at 137 s, and then further reduces to atmospheric pressure at 400 s. Figure 6. Axial power factor of a hot fuel assembly. 12 Figure 7. Cladding surface temperature history during LOCA. i identifies the equally spaced axially position starting from the bottom of the fuel rod. Three case studies are explored to evaluate the impact of axial gas flow on cladding ballooning during a LOCA: standalone fuel cladding bonding simulation, standalone axial gas communication simulation, and a simulation using the preliminary coupling strategy for the two models. For the standalone fuel cladding bonding simulation, the mechanical and thermal material models described Section 2 are used for modeling a smeared fuel column and the plenum. The thin bonding layer is not explicitly modeled; instead, the fuel and cladding meshes are directly connected at their common interface nodes. The finite element grid is shown in Figure 8. The plenum is also meshed, and the gas material properties are assigned to the elements in plenum. For high-burnup fuel, fission gas bubbles and porosities can become over-pressurized, stressing fuel grain boundaries, causing the fuel to fragment when rapidly heated to high temperatures. In this modeling case without precise knowledge when this can occur, fragmentation is assumed to occur at a prescribed time. Prior to the onset of LOCA, the gas volume in the fuel is assumed to be isolated from the plenum by blocking the gas flow from the plenum to the fuel. In the analysis, the gas volume is assumed to be 1% of the fuel column, which is equivalent to ∼1/5 of the plenum volume. At the beginning of the LOCA, the fragmentation is assumed to have occurred instantaneously in the fuel column except for the top layer of fuel elements that are used to separate fuel from the plenum. 13 Figure 8. Schematic of the R-Z mesh used for the modeling of bonded fuel. For both the standalone and bonding informed axial gas communication cases, the traditional BISON approach of modeling the fuel and cladding separately is used where the gap and plenum are not meshed. The gap starts closed to account for the effects of the base irradiation. For the preliminary coupling case, the packing ratio computed from the bonding model is used as an input to the axial gas communication as a function of axial position and time. 5.2.2. Results To examine the impact of how axial gas flow can affect the clad ballooning in bonded fuel, a constant gas flow from the plenum to the fuel stack is assumed for the standalone fuel cladding bonding study. Figure 9 and Figure 10 show the gas pressure in the plenum and the free volume in the fuel stack under different gas flow rates, respectively. 14 Figure 9. Plenum pressure under different gas flow rates. Figure 10. Gas pressure in fuel stack under different gas flow rates. With constant-rate gas flow, the pressure in both the gas flow source (plenum) and sink (fuel stack) were affected; the plenum pressure remains relatively high in an oscillatory way, following the cladding temperature change, and then decreases to some level, while the fuel stack pressure is reduced significantly. At low gas flow rates (1.0×10−5 mol/s and 1.0×10−6 mol/s), there is no apparent pressure difference. This is equivalent in a case where the gas flow is fully blocked. For reference, Figure 11 shows the clad radial displacement from a base case where the fuel is fully bonded and axial gas communication is completely blocked. 15 Figure 11. Clad radial displacement vs. axial position with fully blocked gas flow. Consistent with the pressure history at different flow rates, the cases with low flow rates (at or below 1.0e−5 mol/s) track the fully blocked fuel response with small clad deformation at peak clad surface temperature location, while the high pressure and deformation could occur at other location (e.g., the plenum). At higher gas flow rate (1.0×10−4 mol/s), the axial gas communication measurably alters the pressure distribution between the plenum and the fuel stack; the corresponding radial displacement profiles deviate from the base case accordingly as seen in Figure 12. (a) Flow rate = 1.0×10−4 mol/s (b) Flow rate = 1.0×10−5 mol/s (c) Flow rate = 1.0×10−6 mol/s Figure 12. Clad radial displacement vs. axial position with different gas flow rates (a) = 1.0×10−4 mol/s, (b) = 1.0×10−5 mol/s, and (c) = 1.0×10−6 mol/s. The evolution of gas volumes in the plenum and fuel stack is shown Figure 13. The high gas flow rate case (1.0×10−4 mol/s) has relatively larger cladding deformation, which results in larger gas volume. Therefore, this case has relatively higher pressure initially, as shown in Figure 10, but would be significantly reduced due to the rod ballooning. Further increasing the gas flow rate to 1.0×10−3 mol/s results in an unrealistic scenario where the fuel stack has higher gas pressure than plenum in later time. This unrealistic condition is due to the fact that the flow rate is imposed rather than occurring naturally. The result, however, does show that when the gas flow rate is very high, and the clad deformation matches more closely to a case with instantaneous pressure equilibrium. In this scenario, a single large ballon is observed at the peak clad temperature location, as shown 16 (a) (b) Figure 13. Evolution of gas volume in the (a) plenum and (b) fuel stack. in Figure 14. Figure 14. Clad radial displacement vs. axial position with high gas flow rate (1.0×10−3 mol/s). Figure 15 presents the computed input function for use in the bonding informed axial gas communication simulation. After depressurization to 1 MPa for the system pressure (∼137 s), the packing ratio at around 3-m drops, indicating bond breakage in a locally axially refined region. 17 Figure 15. Evolution of the packing ratio as a function of axial position and time for the bonding informed axial gas communication simulation. The effects of this rudimentary coupling of the bonding model is shown in Figure 16. The scal- ing factor to enable the Reynolds number to be within laminar flow is not fully implemented within the AxialGasCommunication model, but as seen in the trials, the flow rate is faster when utilizing the bonding method as would be expected since it allows for a hydraulic diameter instead of passage through the frag- mented fuel. These preliminary models are rough approximations, and considerable work is planned in future development, but the results show a clear path forward. 6. CONCLUSIONS This report has presented the theory behind a developed fuel cladding bonding model implemented in the BISON fuel performance code to evaluate the impact of the fuel cladding bond on fuel rod behavior during a LOCA. A summary of a previously developed axial gas communication framework was also presented. Implementation testing of the bonding model as well as a full-length PWR rod LOCA demonstration problem was highlighted. It was found that the implementation tests provided alongside the code verified proper coding of the relevant correlations provided in Section 2. The results of cases studies on the full-length PWR rod indicated that the modeling assumptions used in the bonding model can greatly influence the predicted depressurization rates and cladding displacement profiles. 7. FUTURE WORK This report has presented the preliminary implementation of a fuel cladding model and its coupling to an existing axial gas communication model. Several areas of future work have been identified to enhance the 18 (a) (b) Figure 16. The effects of implementing a preliminary bonding model within AxialGasCommunication. (a) The rate of change in the number of moles based on a scale factor applied to the limiting factor in Equation (16). (b) The pressure post-rupture as a result of the limiting factor described in (a). predictability of the individual models and the coupling between them. These areas include: • Extend BISON’s axial gas communication and axial fuel relocation models to work in 2D-RZ axisym- metric and subsequently in three dimensions (3D). The extension to 2D-RZ will enable tight coupling with the presented bonding model. The 3D capabilities will allow for more detailed simulations of upcoming and past experiments to capture that have 3D effects. • Extend the axial gas communication model to pre-rupture evolution in cases where complete flow blockage has not yet occurred. Originally, the axial gas communication model was developed for post- rupture depressurization of the plenum to the system pressure due to wealth of available experimental data. Axial gas communication experiments focused solely on depressurization rates through a porous media can also be currently studies. More recent experiments include capturing gas flow rates in ballooning fuel rods that do not experience rupture. With the industry’s recent shift from minimizing fuel dispersal to a no rupture condition, this functionality will be essential going forward. • Extend the limits of the specific heat and viscosity correlations for the helium, argon, xenon, and krypton to higher temperatures and pressures for improved accuracy. • Incorporate radially varying permeability for axial gas communication. This is important if one has partially relocated fuel, which is expected in pre-rupture scenarios where the rod consists of intact fuel and relocation fragments. This would allow for more refined computations of the depressurization rates. • Account for local changes in transient fission gas release to more accurately compute the pressure gradients that may form within the rod. • Extend the bonding model to work at burnup prior to those when fuel cladding bonds traditionally form. This will be necessary to run full base irradiation prior to experiments focused on the impact of 19 bonding on gas communication when performing a tightly coupled analysis. • Refine fuel cladding bonding model implementation within the code to remove code duplication related to the computation of thermal conductivities of the solid fuel and gaseous components within the bond as some correlations can be obtained from previously existing models. Re-evaluate the stress calculation of the bonded region to determine whether the support of evaluating at only a single quadrature point is needed compared to the standard approach of evaluating at all quadrature points in an element. • Couple the bonding model to existing models available in BISON to compute the thermal conductivity and specific heat of the solid and gaseous components to reduce code duplication. • Perform validation against experiments while incorporating the coupled fuel cladding bonding and axial gas communication framework. • Apply the coupled bonding and gas communication models to alternative high-burnup fuels for LWRs such as Cr2 O3 or Gd2 O3 doped UO2 . • Investigate turbulent flow, loss of mass, and rupture size effects on overall change in gas pressure over time. 8. PUBLICATIONS The work presented in this report has led to a conference paper to be presented at the TopFuel conference in Nashville, Tennessee October 5-9, 2025. Details of the publication are provided below: 1. Daniel van Wasshenova, Wenfeng Liu, Kyle A. Gamble, Joseph Y. R. Rashid, Ryan T. Sweet, and Stephen R. Novascone, ”Fuel Bonding and its Impact on Axial Gas Communication Behavior in Light- Water Reactor Fuel Rods,” TopFuel 2025: Nuclear Reactor Fuel Performance Conference, Nashville, TN, October 2025. 20 9. REFERENCES [1] P. Rudling, R. Adamson, B. Cox, F. Garzatolli, and A. Strasser, “High burnup fuel issues,” Nuclear Engineering and Technology, vol. 40, no. 1, pp. 1–8, 2008. [2] R. L. Williamson, J. D. Hales, S. R. Novascone, G. Pastore, K. A. Gamble, B. W. Spencer, W. Jiang, S. A. Pitts, A. Casagranda, D. Schwen, A. X. Zabriskie, A. Toptan, R. Gardner, C. Matthews, W. Liu, and H. Chen, “BISON: A flexible code for advanced simulation of the performance of multiple nuclear fuel forms,” Nuclear Technology, vol. 207, no. 7, pp. 954–980, 2021. [3] W. Liu and J. Rashid, “BISON Modeling of High Burnup Fuel Behavior with Fuel-Cladding Bonding During LOCA ,” tech. rep., Structural Integrity Associates Inc., 2024. [4] C. Walker, D. Staicu, M. Sheindlin, D. Papaioannou, W. Goll, and F. Sontheimer, “On the thermal conductivity of uo2 nuclear fuel at a high burn-up of around 100mwd/kghm,” Journal of Nuclear Materials, vol. 350, no. 1, pp. 19–39, 2006. [5] C. M. Allison, G. A. Berna, R. Chambers, E. W. Coryell, K. L. Davis, D. L. Hagrman, D. T. Hagrman, N. L. Hampton, J. K. Hohorst, R. E. Mason, M. L. McComas, K. A. McNeil, R. L. Miller, C. S. Olsen, G. A. Reymann, and L. J. Siefken, “SCDAP/RELAP5/MOD3.1 code manual, volume IV: MATPRO-A library of materials properties for light-water-reactor accident analysis,” Tech. Rep. NUREG/CR-6150, EGG-2720, Idaho National Engineering Laboratory, 1993. [6] M. Oguma, “Microstructure effects on fracture strength of UO2 fuel pellets,” Journal of Nuclear Science and Technology, vol. 19, no. 12, pp. 1005–1014, 1982. [7] J.-M. Gatt, J. Sercombe, I. Aubrun, and J.-C. Ménard, “Experimental and numerical study of fracture mechanisms in UO2 nuclear fuel,” Engineering Failure Analysis, vol. 47, pp. 299–311, 2015. [8] A. G. Evans and R. W. Davidge, “The strength and fracture of stoichimetric polycrystalline UO2 ,” Journal of Nuclear Materials, vol. 33, pp. 249–260, 1969. [9] A. L. Loeb, “Thermal Conductivity: VIII, A Theory of Thermal Conductivity of Porous Materials,” Journal of the American Ceramic Society, vol. 37, no. 2, pp. 96–99, 1954. [10] H. Kämpf and G. Karsten, “Effects of Different Types of Void Volumes on the Radial Temperature Distribution of Fuel Pins,” Nuclear Applications and Technology, vol. 9, no. 3, pp. 288–300, 1970. [11] G. Khvostov, W. Wiesenack, M. Zimmermann, and G. Ledergerber, “Some insights into the role of axial gas flow in fuel rod behaviour during the LOCA based on Halden tests and calculations with the FALCON-PSI code,” Nuclear Engineering and Design, vol. 241, no. 5, pp. 1500–1507, 2011. ICONE-17: Heavy Liquid Metal Technologies and Material Challenges for Gen IV Fast Reactors. [12] K. A. Gamble, D. J. van Wasshenova, J. I. Vermaak, M. E. Tano Retamales, and J. D. Hales, “A LOCA Analysis Tool: Coupling RELAP5-3D and BISON,” Tech. Rep. INL/RPT-25-84367, Idaho National Laboratory, 2025. [13] L. Landau and E. Lifshitz, Fluid Mechanics, vol. 6 of Course of Theoretical Physics. Elsevier, 1987. [14] A. E. Slaughter, C. J. Permann, J. M. Miller, B. K. Alger, and S. R. Novascone, “Continuous integration, in-code documentation, and automation for nuclear quality assurance conformance,” Nuclear Technology, vol. 207, no. 7, pp. 923–930, 2021. 21