This study presents a novel physical liver phantom that replicates the mechanical and visual properties of healthy, fatty, and fibrotic liver tissues to support verification, validation, and uncertainty quantification in digital twin frameworks for liver surgery. Simulating these three common liver conditions improves phantom fidelity for surgical training and model validation. We evaluated multiple materials (agar, gelatin, and konnyaku) for phantom fabrication and conducted a twelve-question survey with 25 surgeons to optimize formulations based on tactile realism and visual appearance. Standard surgical interactions (compression, pinching, pulling and cutting) were performed on both phantoms and fresh porcine liver models using the da Vinci® Xi surgical system. Stereo endoscopic video recordings enabled a comparative analysis of visual fidelity and tissue deformation under surgical manipulation. Mechanical properties (compressive and tensile Young’s modulus) were quantitatively measured under mechanical testing and compared with real tissue measurements in literature. The resulting phantom, validated through surgeon feedback and mechanical testing, provides a tangible platform for physical model validation and surgeon-in-the-loop simulations in digital twin applications. This work advances digital twin methodologies by offering a realistic testbed for model calibration and verification, bridging computational models and real surgical scenarios.