Research on Topology Optimization Design of Load-bearing Structure of Space Nuclear Reactor Power

The design of truss structures for space nuclear reactor power systems, a critical energy solution for deep-space exploration, faces two significant challenges under extreme launch loads (axial 6g, lateral 3g) while supporting critical components totaling ～9 850 kg, where the reactor core alone constitutes over 35% of the mass. First, conventional optimization struggles with the “component suspension issue”: the extreme mass concentration causes material to cluster excessively around the core support regions, leaving low-mass components like turbogenerators and recuperators without effective load paths due to their minor mechanical contribution. Second, a “lightweight-manufacturability conflict” exists: while a low volume fraction (≤10%) is mandated for lightweight truss-type designs, overly low fractions yield indistinct load paths, and higher fractions tend to produce inefficient, thin-shell structures incompatible with manufacturing. To overcome these challenges and achieve a high-stiffness, lightweight, and manufacturable design, this study proposed a multi-design domain (referred to as multi-domain) collaborative topology optimization method. Researchers addressed this by developing an innovative optimization model. They partitioned the structure into independent inner and outer design domains. The outer domain was dedicated to directly supporting the massive reactor core and interfacing with the launch vehicle cylinder, while the inner domain exclusively supported the low-mass components. They implemented multi-load case integrated optimization, eight critical lateral acceleration directions identified via modal analysis. Volume fraction constraints were applied independently to each domain. A comparative analysis was conducted using both the solid isotropic material with penalization (SIMP) method and the level-set method (LSM) to investigate their influence. For validation, a high-fidelity finite element model was established based on the Project Prometheus (JIMO) layout, with modal analysis guiding lateral load case distribution. An orthogonal experimental design was adopted to parametrically study the interactive effects of inner/outer domain volume fraction combinations on performance metrics. The results demonstrate that the multi-domain collaborative optimization effectively resolves the core challenges. At a combined 5% volume fraction (e.g., 5% inner + 5% outer), material in the outer domain forms direct load paths from the reactor core to the launch interface, while the inner domain independently supports low-mass components, completely eliminating the suspension issue. Compared to single-domain optimization, this approach significantly enhances minor component support reliability and yields clearer load paths under identical volume constraints. Volume fraction allocation proves critical: specific stiffness (stiffness-to-mass ratio) increases substantially when the outer domain’s fraction approaches or exceeds the inner’s (e.g., 5% outer + 2.5% inner) because outer domain support enables shorter, straighter, and mechanically more efficient load paths than curved paths transferring load between domains. A stiffness-volume relationship emerges: specific stiffness rises, then decreases with increasing total volume fraction; below 5%, complete paths fail to form, while exceeding 10% yields diminishing returns due to marginal material utility. Furthermore, the LSM exhibits clear engineering advantages over SIMP: it generates cleaner, manufacturable truss-type structures, avoids irregular plate-like features, and achieves higher specific stiffness with straighter paths at identical volume fractions. Consequently, LSM optimization with volume fractions like 0.025×0.025 or 0.05×0.05 (inner×outer) represents the recommended solution, achieving the optimal balance among lightweight design, high stiffness, and manufacturability. In conclusion, this study establishes a significant methodological innovation: multi-domain collaborative topology optimization successfully resolves the component suspension problem under extreme mass concentration and stringent volume constraints by decoupling support for the core and low-mass components, providing a novel structural design approach for complex space systems. Key design guidelines are derived: volume fraction allocation should favor the outer domain (outer≥inner recommended) to achieve high-specific-stiffness linear paths, and the LSM is the preferred optimization approach due to its superior performance in generating manufacturable structures and enhancing stiffness. The optimized configurations (5%-7.5% total volume fraction) deliver direct engineering application value, satisfying all critical requirements for space nuclear power systems-structural stiffness specifications, lightweight objectives (using TC4 titanium alloy), radiation shielding compatibility, and assembly feasibility-thereby providing a directly implementable technical solution for nuclear power load-bearing structures in future deep-space exploration missions.