Tip Penetration Resistance Under a Coaxial-Sleeve Stress-Shielding Boundary: Dimensionless and Mechanism-Corrected Models for Probing Stored Wheat

Deep in situ sensing is essential for grain storage safety, yet penetration resistance restricts sensing devices in deep grain layers. Everting growth robots reduce body-side friction through tip eversion and are therefore well suited to low-disturbance probing in grain bulks. However, real eversion couples soft-body deployment, particle rearrangement and boundary contact, so tip resistance cannot be measured independently with conventional full-contact models. This study combined dimensional analysis, discrete element simulations and coaxial double-sleeve equivalent penetration tests to establish a dimensionless baseline model (DBM) and a mechanism-corrected penetration model (MCPM). The double-sleeve apparatus acts as an equivalent rigid test boundary; it does not reproduce membrane eversion or flexible deployment. Within the stable penetration interval (100–300 mm), the measured tip resistance reached only 8.0–12.5% of the unshielded DBM baseline, a model-relative estimate obtained under different boundary conditions. The geometry exponent changed from −0.5642 in the DBM to +0.2983 in the MCPM, reflecting an empirical correction fitted to three retained geometries. Prediction failed for the lowest aspect ratio (λ=0.64), and an aspect ratio of 1.2–1.6 is suggested only as a preliminary range pending validation. The proposed equivalent test and corrected model provide preliminary references for tip-load estimation and probe optimisation of everting growth robots.