University of Illinois Urbana-Champaign
Illinois Digital Environment for Access to Learning and Scholarship RepositoryNot a member yet
123813 research outputs found
Sort by
Function over form in contemporary media: The animated family as case
No full textFamily is increasingly defined by what it does, not how it looks. It is less about biological ties and more about bonds and roles. This paper tests that claim with the anime SPY×FAMILY, which portrays a “fake” household made by necessity. Using the Olson Circumplex Model, I assess cohesion, flexibility, and communication across key scenes and arcs. The findings show how the Forgers move from cover to care. It transforms from a facade into a loving, functional unit that coordinates roles, manages conflict, and the most importantly basic act, talks more openly. On functional grounds, they meet the criteria for “family.” Social Constructionism frames this shift. Shared meanings and cultural scripts build family over time. Animation’s imaginative space helps norm‑breaking legible and safe, inviting viewers to rethink kinship and embrace diversity. When function is present, non-traditional families can thrive. This is a theory‑driven case that links on‑screen practice to public acceptance, showing how popular media model inclusive family forms and contributing to the discussion on modern family forms
Investigation of machine learning techniques in stabilization of co-propagating polarization encoded photons
Get PDFQuantum communication protocols and networks have advanced considerably in experimental laboratories; however, their operation is frequently constrained by short time frames and significant environmental noise. These challenges affect both time-coded and polarization-encoded photon systems. Polarization encoding has become the prevalent method despite its intrinsic issues with stability. This thesis investigates the integration of classical and quantum signals through the application of modern machine learning techniques, with the goal of overcoming the stochastic challenges that have historically impeded system performance. We evaluate three distinct schemes: (1) direct prediction of the quantum signal from the classical signal, (2) variable recalibration queue for the quantum channel, and (3) calibration of the quantum signal via reinforcement learning. Each approach is characterized by unique strengths and limitations in terms of measurement overhead, implementation complexity, and operational stability. Through extensive experimental testing over durations ranging from hours to days, we analyze the performance of attention-based models, sliding window time series predictors, and reinforcement learning frameworks in predicting and stabilizing polarization states. Our findings indicate the potential of machine learning approaches to enhance the longevity and reliability of quantum communication systems, while also highlighting the challenges of model generalization and data requirements for robust performance. This research contributes to the advancement of quantum communication infrastructure by demonstrating practical methods for maintaining stable polarization encoded quantum channels, essential for the development of long-distance secure quantum networks.Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2026-02-19 without embargo termsThe student, Yueze Liu, accepted the attached license on 2025-07-22 at 04:18.The student, Yueze Liu, submitted this Thesis for approval on 2025-07-22 at 04:21.This Thesis was approved for publication on 2025-10-02 at 09:35.DSpace SAF Submission Ingestion Package generated from Vireo submission #22688 on 2026-02-19 at 18:24:0
The power and limitations of restricted quantum learning models
Get PDFQuantum technologies have rapidly advanced over the past decade or so. However, noise and decoherence are still challenges. At the same time, quantum learning theory has quickly grown and many techniques have been developed with applications in verifying and benchmarking quantum devices. However, known generic but optimal protocols may require entangled measurements across many copies of a state, rendering experimental implementations rather difficult. With this mind, it is interesting to ask what learning tasks can and cannot be achieved with limited models of measurements, a question which has received much attention in recent years. In this thesis we answer several questions in this vein.
First, we consider entangled versus separable measurements for exactly learning a function c ∈ C from a quantum example state |c⟩. We show that, if T copies suffice to learn f using entangled measurements, O(n^2) copies suffice to learn f using only separable measurements.
Second, we study the Quantum Statistical Query model. We exhibit a concept class C based of degree-2 functions with an exponential separation between QSQ learning and quantum learning with entangled measurements (even in the presence of noise). This proves the “quantum analogue” of the seminal result of Blum, Kalai, & Wasserman. [1] that separates classical SQ learning from classical PAC learning with classification noise. To obtain these bounds, we introduce a quantum statistical query dimension (QSD). We prove superpolynomial QSQ lower bounds for testing purity of quantum states, reconstructing a state via shadow tomography, learning coset states for the Abelian hidden subgroup problem, approximating degree-2 functions, finding planted bi-cliques, and learning the output states of Clifford circuits of depth polylog(n). We also show that an extension of QSD characterizes the complexity of general search problems. Lastly, we give an unconditional separation between weak and strong error mitigation and prove lower bounds for learning distributions in the QSQ model. Prior works by Quek et al. [2], Hinsche et al. [3], and Nietner et al. [4] proved analogous results assuming diagonal measurements and our work removes this assumption.
Next, we consider the fundamental task of distributed inner product estimation when allowed limited communication. Suppose Alice and Bob are given k copies of an unknown n-qubit quantum state |ψ⟩ and |ϕ⟩ respectively. They are allowed to send q qubits to one another with the goal of estimating |⟨ψ|ϕ⟩|^2 up to constant additive error. We show that k = Θ(√(2^(n−q)) copies are essentially necessary and sufficient for this task (extending the work of Anshu, Landau and Liu (STOC’22) who considered the case when q = 0). Additionally, we also consider the task when the goal of the players is to estimate |⟨ψ|M|ϕ⟩|^2, for arbitrary Hermitian M. For this task we show that certain norms on M determine the sample complexity of estimating |⟨ψ|M|ϕ⟩|^2 when using only classical communication.
Returning to separable vs entangled measurements, we make progress on proving a hierarchy of copy complexity separations. To that end, we prove that estimating Tr[ρ^k], for k a constant, requires Ω(log d) samples with (k − 1)-copy measurements. By contrast, O(1) samples suffice with k-copy measurements. To obtain our bounds, we extend Harrow’s [5] method of PPT constraints on sums of permutations.
Lastly, we consider measuring the entanglement in multipartite systems using two-copy measurements, in particular the SWAP test. Multipartite entanglement is one of the hallmarks of quantum mechanics and is central to quantum information processing. We show that the concentratable entanglement (CE), an operationally motivated measure from SWAP tests, induces a hierarchy upon pure states from which different entanglement structures can be experimentally certified. In particular, we find that nearly all genuine multipartite entangled states can be verified through the CE. Interestingly, GHZ states prove to be far from maximally entangled according to this measure. Instead we find the exact maximal value and corresponding states for up to 18 qubits and show that these correspond to extremal quantum error correcting codes. The latter allows us to unravel a deep connection between CE and coding theory. Finally, our results also offer an alternative proof, on up to 31 qubits, that absolutely maximally entangled states do not exist.Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2026-02-19 without embargo termsThe student, Louis Schatzki, accepted the attached license on 2025-08-20 at 14:56.The student, Louis Schatzki, submitted this Dissertation for approval on 2025-08-20 at 17:16.This Dissertation was approved for publication on 2025-08-25 at 15:52.DSpace SAF Submission Ingestion Package generated from Vireo submission #22767 on 2026-02-19 at 18:24:1
Shear force impacts type IV pilus and flagellar dynamics during Pseudomonas aeruginosa surface adhesion
Get PDFSurface-attached bacteria experience forces associated with fluid flow in nature. However, many studies on bacterial adhesion have been conducted in conditions either lacking flow or using forces not typically found in natural systems.
In Chapter 2, I examine how Pseudomonas aeruginosa cells use type IV pili to interact with surfaces under host-relevant shear forces. I demonstrate that cell tilting, driven by type IV pilus retraction, predicts surface departure at low shear forces. Conversely, higher host-relevant shear forces counteract cell tilting and enhance adhesion. Thus, P. aeruginosa couples type IV pilus dynamics and cell geometry to tune adhesion to its mechanical environment.
In Chapter 3, I investigate how host-relevant shear force impacts flagellar rotation during P. aeruginosa initial surface interactions. Experiments using strains having fluorescently-labeled flagella reveal that flow bends upstream-facing flagella and blocks rotation. Additionally, cells with upstream flagella exhibit fewer surface departures than cells with downstream flagella, showing that flagellar rotation drives surface departure. Therefore, these results reveal a novel role for flagellar rotation and determine how bacteria optimize surface interactions in dynamic environments.
Together, these findings emphasize the need to study bacteria under mechanically relevant conditions. Shear force exerts a mechanical stress on surface attached bacterial cells, leading to prolonged surface adhesion. By integrating host-relevant conditions into bacterial research, my work has provided a clearer understanding of how P. aeruginosa optimizes surface behavior in dynamic environments.Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2026-02-19 without embargo termsThe student, Jessica-Jae Palalay, accepted the attached license on 2025-10-13 at 07:28.The student, Jessica-Jae Palalay, submitted this Dissertation for approval on 2025-10-13 at 07:29.This Dissertation was approved for publication on 2025-10-14 at 14:41.DSpace SAF Submission Ingestion Package generated from Vireo submission #22824 on 2026-02-19 at 18:24:2
Energy and time efficient federated learning
Get PDFOver the past decade, the volume of data generated by edge devices has grown exponentially as the number of edge devices surges. Federated learning (FL) enables on-device training while preserving privacy, but edge devices typically operate under tight time and energy budgets, highlighting the need for time- and energy-efficient FL algorithms. While prior work focuses on either compute or communication costs, our hardware characterization shows that, although per-bit communication is far more expensive than per-bit computation, the relative importance of compute and communication costs depends on the communication period. Therefore, both costs must be considered before we find the optimal communication period.
In this work, we formulate the optimal time and energy efficiency of FL systems into two separate optimization problems, characterized by four design variables: communication period C, GPU frequency f_GPU, number of clients K, and number of global rounds G. We solve both problems analytically and empirically using our simulation framework featuring a maximum of 10 Jetson clients with IID Cifar-10 dataset. Our empirical results show that time efficiency is optimal at the maximum available GPU frequency f^(max)_GPU, while energy efficiency is optimal at the highest GPU frequency with minimum supply voltage. Ablation studies reveal that optimal K for time efficiency increases with target accuracy; while optimal K for energy efficiency is always the smallest feasible K. The ablation also shows that optimal C for both time and energy efficiency increase with target accuracy.
Based on the optimal hyperparameter setting, we find that optimal communication latency dominates only in low-accuracy regions (20%−29%); whereas optimal compute latency dominates in high-accuracy regions (30%−69%). On the other hand, optimal compute energy dominates across all accuracy regions. These results imply that compute costs are the primary concern for system efficiency under homogeneous FL settings.Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2026-02-19 without embargo termsThe student, Jing-Teng Hwang, accepted the attached license on 2025-11-26 at 22:29.The student, Jing-Teng Hwang, submitted this Thesis for approval on 2025-11-26 at 23:00.This Thesis was approved for publication on 2025-12-02 at 10:13.DSpace SAF Submission Ingestion Package generated from Vireo submission #22975 on 2026-02-19 at 18:25:3
Advancement of liquid lithium plasma-facing components for next-generation fusion devices
No full textFusion power generation is clearly an attractive solution to the problem of global energy production. It produces no greenhouse gases, is incredibly energy dense and could offer access to limitless energy. Generating and controlling the extreme conditions required to achieve fusion however, poses several large technical challenges. One of these challenges is handling the large heat, plasma and neutron fluxes exhausted from the core plasma. Current tokamak design points assume a steady state heat flux of 10MW/m2 increasing to values >1GW/m2 in transient events and particle fluxes > 1 × 1023m−2s−1. Flowing liquid lithium Plasma Facing Components (PFCs) are a promising candidate for plasma exhaust. A flowing PFC presents a constantly refreshing face to the plasma greatly reducing the erosion and sputtering of the solid substrate below whilst also protecting against catastrophic damage during transient events. These factors are predicted to extend the component lifetime in comparison to solid walls, increasing the duty cycle and thus the economic output of a reactor. Lithium also brings improvements to the plasma performance in a reactor by gettering impurities to reduce the zeff and thus boost confinement time or by offering access to high performance operating modes such as the ’low-recycling regime’. By reducing the recycling of hydrogenic species at the walls of the reactor, the collisionality of the edge will drop and, as such, the temperature will rise. This increased edge temperature can bring a myriad of benefits such as a reduction in anomalous transport, an increase in the burning region of the tokamak and an increase in the scrape off layer power width λq.
The integration of liquid lithium into the divertor region of a fusion power plant requires several key issues to be addressed. This work will outline these key issues, the progress made so far to overcome them and the future tasks required to bring the maturity of the liquid lithium technology to a level where it can be installed into a reactor. The experimental test bed used throughout this work is the Actively Pumped Open Surface Lithium LOop (APOLLO) at the University of Illinois at Urbana Champaign (UIUC). Due to lithium’s high chemical reactivity, lithium PFCs must be included in an integrated loop where the fluid can be pumped into a reactor, removed and then diverted to processing chambers for impurity removal and the re-capture of fuel species. APOLLO was commissioned in January 2024 to further develop the technology readiness level (TRL) of lithium components. At the time of writing the key components installed on APOLLO are an open surface lithium divertor module within a magnetic field, an ECR plasma source, an electron beam heat source and a hydrogen distillation column.
A new 3D ordered mesh flow plate has been 3D printed from tantalum and installed into the APOLLO PFC. With this upgraded flow plate, 100% lithium coverage has been achieved across the entirety of the flow plate. Complete coverage was observed across the full fluid velocity and magnetic field ranges achievable in APOLLO and was able to be regained after freezing and restarting the flow multiple times. These tests also achieved some impressive engineering feats. Over the 91 days these experiments were run: lithium was molten in the loop for 101 hours, was recirculated for 44 hours and was frozen and restarted 17 times all of which are key steps in showing the reliability of lithium systems.
With uniform flow achieved on the plate, experiments studying Thermo-Electric MagnetoHydroDynamics (TEMHD) drive began. This involved the re-commissioning of an electron beam assembly to generate fusion relevant heat fluxes at the lithium PFC. The heat flux produced at the PFC was determined using a calorimeter and a maximum value of 6.1 ± 0.63MW/m2 was achieved. Exposures of flowing lithium to this heat flux revealed that the fluid dynamics were driven by the plate normal field and the truncation of the heat flux across the plate. Throughout the test transient peaks on the order of 0.1g/s were observed with applied heat flux and appear to correlate to the expulsion of fluid from the plate. To interpret the results from the TEMHD experiments, a fluid dynamics simulation was created in COMSOL multiphysics. Once the model had been benchmarked against the experimental observations it was extended up to magnetic fields of 5T to investigate the performance of TEMHD drive in a fusion relevant environment. The final part of this work was the design, construction and characterization of an Electron Cyclotron Resonance (ECR) plasma source. This source was installed so that the retention of hydrogen species into liquid lithium and their subsequent transport around the loop could be elucidated. The source also provides a controlled influx of said species into the lithium flow; allowing for the benchmarking of impurity measurement diagnostics and the determination of the removal efficiency of the distillation column installed into APOLLO. The characterization of the plasma centered around tuning external parameters to maximize the flux of ions and radicals to the lithium surface. An optimized solution was found which generated a fluence of 1.7 × 1018s−1 of hydrogenic species to the plate. Studies of the uptake of helium ash were also studied using the ECR plasma source. Helium fluxes on the order of ×1015s−1 were observed to have been transported from the open surface PFC to the bulk lithium reservoir.Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2026-02-19 without embargo termsThe student, Daniel O'Dea, accepted the attached license on 2025-12-03 at 13:34.The student, Daniel O'Dea, submitted this Dissertation for approval on 2025-12-03 at 13:50.This Dissertation was approved for publication on 2025-12-04 at 16:17.DSpace SAF Submission Ingestion Package generated from Vireo submission #23045 on 2026-02-19 at 18:26:4
A reactor physics framework to detect anomalies in HTGR core
Get PDFThis dissertation extends the application of neutron noise analysis as a tool for detecting anomalies in High Temperature Gas-cooled Reactors (HTGRs). Neutron noise is defined as the fluctuations in the neutron detector signal caused by perturbations in a nuclear reactor. Neutron noise analysis is performed by obtaining the neutron flux signals in the time domain, which is then converted to the frequency domain to perform spectral analysis of the neutron noise. This analysis is possible only for steady-state conditions. Neutron noise analysis has been used in the past in LWRs and has proven effective in several applications. This includes the application for core monitoring and surveillance, model validation, and core diagnostics for anomaly detection.
Considering the effectiveness of the analysis, computational methods and models of neutron noise were developed from the neutron transport equation. The neutron noise models can be divided into two categories, namely the absorber of variable strength (AVS) and the vibrating absorber. The AVS is modeled as fluctuations in the macroscopic cross-section at a fixed location . This includes the cases of perturbations traveling with the coolant flow in molten salt reactors (MSRs) and perturbations due to unseated fuel assemblies in boiling water reactors (BWRs). The vibrating absorber is modeled as a displacement of the assembly from its nominal position, with the assumption that neutron noise is induced by the vibrating interface. Several tools have been developed to solve neutron noise equations, using both Monte Carlo and deterministic methods.
This dissertation is divided into two contributions. First, we determined if neutron noise analysis and current neutron noise unfolding methods are applicable to prismatic-type HTGRs. To make such determination, a neutron noise equation solver was developed for both hexagonal and rectangular geometries. The developed solver solves the multigroup neutron diffusion equation in the frequency domain, using a finite volume method for spatial discretization. The solver has been verified using several benchmarks. Then, we simulated neutron noise in a prismatic-type HTGR core to obtain its neutron noise characteristics, including the point-kinetic and spatial components of neutron noise, as well as the zero-power reactor transfer functions. The results show that, generally, the neutron noise in HTGRs exhibits a similar behavior to that in LWRs. However, there are a few aspects that distinguish HTGRs. From the zero-power reactor transfer function, HTGRs have a lower plateau, which means that HTGRs are more sensitive to perturbations within a certain range of frequencies. Neutron noise in HTGRs is largely dominated by the point-kinetic component of neutron noise, which means that the neutron noise is largely driven by the reactivity effect of the perturbation.
The second contribution is to introduce two new methods to unfold neutron noise. The challenge of neutron noise unfolding is that the neutron noise is not known (measured) at all positions and energy, it is only known at the detector positions. There are three primary methods for neutron noise unfolding: the inversion method, the zoning method, and the scanning method. These methods were applied to unfold the neutron noise source in the 2D HTTR core. The results show that the inversion method lacks accuracy due to the use of linear interpolation. The scanning method shows satisfactory results but is limited to only one AVS-type source. The zoning method incorrectly identifies the location and magnitude of the neutron noise source. This dissertation developed two novel methods for neutron noise unfolding that use Green’s function, namely the brute force method and the greedy method. The brute force method solves the combinations of Green’s function and its coefficient until it finds the correct combination of Green’s functions. The greedy method iterates Green’s function to find the Green’s function matrix and the local minimum of the residual. This results in several valid solutions to the noise sources. The correct solution is the solution that requires the least number of Green’s function combinations to minimize the residual. The results show that both brute force and greedy methods can simultaneously unfold multiple AVS-type noise sources, which is not possible by the currently used methods (inversion, zoning, scanning methods).Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2026-02-19 without embargo termsThe student, Harun Ardiansyah, accepted the attached license on 2025-12-04 at 11:06.The student, Harun Ardiansyah, submitted this Dissertation for approval on 2025-12-04 at 11:14.This Dissertation was approved for publication on 2025-12-04 at 16:23.DSpace SAF Submission Ingestion Package generated from Vireo submission #23046 on 2026-02-19 at 18:28:0
Laboratory and field evaluation of efficacy of potential fungicides for managing phytophthora blight (Phytophthora capsici) of processing pumpkin (Cucurbita moschata)
Get PDFPhytophthora blight, caused by the oomycete Phytophthora capsici Leonian, is a destructive disease that severely limits the production of processing pumpkin (Cucurbita moschata Duchesne). This study was conducted to evaluate the efficacy of the potential eight fungicides or their combinations (cyazofamid, dimethomorph, dimethomorph + ametoctradin, ethaboxam, fluopicolide, mandipropamid, mandipropamid + oxathiapiprolin, and potassium phosphite) for managing P. capsici in the laboratory and fields. The effectiveness of the fungicides on mycelial growth of six Illinois isolates of P. capsici in the laboratory was determined. The field efficacy of the fungicides in fourteen different treatment combinations was evaluated in commercial processing pumpkin fields with high natural inoculum of P. capsici in Tazewell County, Illinois, during 2024 and 2025. The laboratory assays revealed that mandipropamid + oxathiapiprolin (Orondis Ultra) fungicide was the most effective in suppressing mycelial growth of P. capsici isolates [0.0036 µg/ml for 50% reduction in mycelial growth (EC50)]. In field trials, all fungicide treatments significantly (P = 0.05) reduced vine and fruit infection. Incidence of vine infection in control plots was 38.33 and 86.25% in 2024 and 2025, respectively; whereas vine infection in fungicide treated plots ranged from 1.67 to 20.00% and 1.25 to 35.00% in 2024 and 2025, respectively. Fruit infection with P. capsici in untreated plots was 45.40 and 76.10% in 2024 and 2025, respectively. The incidence of fruit infection in the fungicide treated plots ranged from 3.49 to 32.15% in 2024, and from 0.83 to 41.40% in 2025. Spray applications of mandipropamid + oxathiapiprolin (Orondis Ultra) alternated with cyazofamid (Ranman 400SC) alternated with mandipropamid (Revus 2.09SC) resulted in low vine and fruit infection. The findings of this field research emphasized using fungicides with different modes of action for effective protection of processing pumpkins against P. capsici.Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2026-02-19 without embargo termsThe student, - Sahil, accepted the attached license on 2025-12-08 at 13:39.The student, - Sahil, submitted this Thesis for approval on 2025-12-08 at 13:50.This Thesis was approved for publication on 2025-12-09 at 13:47.DSpace SAF Submission Ingestion Package generated from Vireo submission #23103 on 2026-02-19 at 18:30:0
Lion-hearted Zezva: musical memorial at a funerary horse race
No full textThis thesis is focused on the memorial practices of the Tushetian community of northeastern Georgia, in particular their annual, commemorative festival known as Zezvaoba. Through an analysis of Zezvaoba’s structure, rites, and sounds; an explanation of local histories and beliefs; and personal accounts derived from my ethnographic experience, I contextualize Tushetian practices (musical, equestrian, and culinary) within a broader conversation about regional identity, the politicization of memorial practices, and the utilization of musical performance as a means to both remember the past, and comment on the present and future.Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2026-02-19 without embargo termsThe student, Benjamin Michels Wheeler, accepted the attached license on 2025-12-09 at 15:46.The student, Benjamin Michels Wheeler, submitted this Thesis for approval on 2025-12-09 at 15:57.This Thesis was approved for publication on 2025-12-10 at 07:31.DSpace SAF Submission Ingestion Package generated from Vireo submission #23114 on 2026-02-19 at 18:30:1
Impact of nuclear data libraries for criticality calculation of HTGR
Get PDFThis work examines how variations among evaluated nuclear data libraries affect criticality calculation for High-Temperature Gas-cooled Reactors (HTGR). Accurate nuclear data is essential for reactor safety analysis. However, previous studies have reported significant differences in calculated keff for HTGR systems when different nuclear data libraries are employed. This discrepancy can be attributed to a lack of applicable benchmark experiments for biasing the nuclear data for HTGR. To assist nuclear data evaluators in improving the accuracy of nuclear data, this thesis quantifies the reaction channels, isotopes, and energy range that have the largest contribution towards uncertainty in keff. It is also important to quantify the Δk caused by differences in nuclear data between nuclear libraries by specific reaction channels and isotopes.
This research is based on the High Temperature engineering Test Reactor (HTTR) and HTR-10 reactor models. Serpent 2 was used to calculate criticality and its uncertainty due to nuclear data, using data from evaluated nuclear data libraries ENDF/B-VIII.1 and JENDL-5. Uncertainty quantification demonstrates that for both HTTR and HTR-10, keff uncertainty due to ENDF/B-VIII.1 data is larger than uncertainty due to JENDL-5 data because of larger thermal covariance of U-235 ̅_ data in ENDF/B-VIII.1. In addition, HTR-10 uncertainty is larger than HTTR uncertainty for both libraries. This could be attributed to smaller fuel volume ratio in the active core (V_fule/V_active core) of the HTR-10, therefore softer spectrum, causing its criticality to be more sensitive to C-12 capture and elastic scattering cross section.
For criticality calculations, using JENDL-5 instead of ENDF/B-VIII.1 leads to a Δk≈−200 pcm for the HTTR and a Δk≈−360 pcm for HTR-10. The most significant contributors are the C-12 capture cross section, C-12 elastic scattering cross section, and angular distribution of C-12 elastic scattering. For the HTTR, compared with the nominal case (all ENDF/B-VIII.1), using C-12 capture cross section data from JENDL-5 causes Δk≈−180 pcm. The difference in thermal region dominates criticality variation for the C-12 capture cross section. For the C-12 elastic scattering cross section, Δk≈−140 pcm. For the C-12 elastic scattering angular distribution, Δk≈150 pcm. The fast region dominates criticality variation for both the C-12 elastic scattering cross section and its angular distribution. For the HTR-10, the C-12 capture cross section causes Δk≈−200 pcm. For the C-12 elastic scattering cross section, Δk≈−320 pcm. For the C-12 elastic scattering angular distribution, Δk≈320.Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2026-02-19 without embargo termsThe student, Runxia Wen, accepted the attached license on 2025-12-11 at 09:08.The student, Runxia Wen, submitted this Thesis for approval on 2025-12-11 at 09:18.This Thesis was approved for publication on 2025-12-11 at 14:28.DSpace SAF Submission Ingestion Package generated from Vireo submission #23133 on 2026-02-19 at 18:30:1