[Objective] This study aims to address the difficulty of accurately representing the uniaxial compressive behavior of low-strength high-fluidity self-compacting concrete (HSCC) and natural rock in meso-scale simulations of rock-filled concrete (RFC). Existing simulation approaches tend to overestimate the post-peak residual capacity of C15 HSCC, while conventional rock models may suffer from convergence problems and may not adequately reproduce brittle shear-spalling failure. To fill this gap, HSCC and limestone taken from an RFC engineering project in Guizhou were investigated, and constitutive parameters for the constituent materials of RFC were established and validated to provide reliable inputs for subsequent three-phase meso-scale simulations. [Methods] Uniaxial compression tests and supporting mechanical tests were conducted according to relevant standards. During uniaxial compression, digital image correlation (DIC) was used to obtain full-field strain evolution and complete stress-strain responses, with particular attention to post-peak deformation. On this basis, the descending-branch parameter of the HSCC constitutive curve was recalibrated using the least-squares method. The damage variables obtained from the code-based method and the Najar energy method were compared, and an inelastic-strain amplification coefficient was introduced to modify the input parameters of the Concrete Damaged Plasticity (CDP) model. Numerical validation of the HSCC and rock specimens was then performed using the ABAQUS explicit dynamic solver. [Results] The code-based HSCC constitutive curve agrees well with the experimental curve before the peak, but exhibits a clear deviation in the post-peak stage because the calculated descending-branch parameter is too small. For C15 HSCC, the damage factor obtained by the Najar energy method is more consistent with the actual damage evolution, indicating that the code-based method may overestimate stiffness degradation when plastic deformation is significant. Direct use of the inelastic strain calculated by the conventional method leads to premature brittle fracture in the numerical model; after introducing an inelastic-strain amplification coefficient of 1.2, the simulated softening branch agrees closely with the experimental result. Both DIC observations and numerical simulations show that HSCC damage initiates at multiple points and forms X-shaped conjugate microcrack bands during the early loading stage. Near peak load, the damage rapidly localizes and eventually develops into an inclined dominant shear crack. For rock, the Drucker-Prager shear damage model reproduces the experimentally observed shear-spalling failure mode and avoids the parameter-identification and convergence problems that may arise when rock data are forced into a CDP framework. The final simulated failure morphology is generally consistent with the experimental observations, demonstrating that the selected rock constitutive strategy can reasonably describe the macroscopic mechanical response of the material. [Conclusion] These findings suggest that reliable meso-scale simulation of RFC requires separate constitutive descriptions for low-strength HSCC and rock, rather than direct adoption of conventional concrete-oriented models. For C15 HSCC, recalibrating the descending-branch parameter, using the Najar energy method to describe damage evolution, and amplifying the inelastic strain can jointly capture the rapid post-peak stress drop and the actual energy dissipation process. For limestone, the Drucker-Prager model coupled with shear damage is more suitable than the conventional CDP model for capturing brittle compressive failure and maintaining numerical stability. The constitutive framework established in this study improves the physical fidelity of RFC constituent-material modeling and provides a practical parameter basis for future three-phase meso-scale simulations involving the rock skeleton, HSCC matrix, and interfacial transition zone. Future research should further quantify interfacial mechanical properties and cross-scale coupling so that the cooperative action between the rock skeleton and multiphase interfaces can be incorporated into RFC material design.