Glass fiber-reinforced plastic mortar (GRPM) pipes are widely used in municipal and industrial engineering. However, their socket interfaces are prone to leakage failure under external loads. Existing studies lack targeted test methods for large-diameter GRPM interfaces. In this study, we design a four-point bending test device to study interface mechanical properties and water tightness and reveal its failure mechanism.

Based on simply supported beam stress theory, the test simplifies the GRPM pipe with a socket interface into a simply supported beam model and applies symmetrical vertical load to simulate the actual engineering load effect on the pipe interface. The test device comprises three core subsystems: support, loading, and measurement systems. The support system adopts a 500 mm-high steel base and arc-shaped pipe supports processed by a 50 mm thick steel plate, which can effectively transmit load and constrain the horizontal displacement of the pipe body while allowing the interface to rotate freely in the plane. The loading system uses a 500 kN three-channel electro-hydraulic servo testing machine matched with a self-developed arc-shaped loading beam, which realizes pure bending stress in the interface area and avoids stress concentration damage to the pipe wall. The measurement system adopts a DH3816N static strain testing system equipped with six displacement sensors, which are symmetrically arranged at the pipe ends, 600 mm away from the pipe ends and directly above the interface to collect real-time vertical displacement data. The test specimen is a DN800 GRPM pipe with a socket interface, with a total length of 3.0 m after assembly. The test process includes device installation, specimen alignment, water injection, 5 kN preloading debugging, and formal loading at a constant rate of 0.05 mm/s. The test terminates when rapid interface leakage occurs or the maximum vertical load reaches 300 kN, monitoring interface vertical displacement, relative rotation angle, axial pull-out amount, and water leakage state synchronously.

The following are the test results: (1) The socket interface maintains good water tightness under 0–64 kN load. The initial leakage occurred at 64 kN, marking the functional failure of water tightness, with an interface vertical displacement of 16.9 mm and a relative rotation angle of 1.90°. As the load increases to 90, 160, and 220 kN, the leakage rate and volume gradually increase, resulting in complete water tightness failure. (2) The interface vertical displacement increases rapidly before reaching 80 kN and slows down considerably from 80 to 300 kN. No structural damage occurs at 300 kN, and the interface still retains good mechanical bearing capacity. (3) The relative rotation angle increases linearly before leakage and grows slowly afterward, reaching 3.90° at the end of loading, which is only 2.0° higher than that at initial leakage. (4) Crisp sounds are heard at 120, 190, 250, and 280 kN due to sudden socket and spigot axial pull-out. The maximum axial pull-out displacements at the interface bottom and middle are 57 and 47 mm, respectively, accounting for 15% and 12.4% of the socket length.

Although conventional GRPM socket interfaces have excellent structural reliability, premature water tightness failure at 64 kN is a critical design flaw, mainly due to interface multidirectional deformation exceeding the rubber ring’s effective sealing range. Optimizing sealing ring height and socket length and adding spigot-thickened sections can effectively improve water tightness. The proposed test method provides a reliable basis for pipeline design optimization and can serve as an experimental teaching platform for civil engineering graduates.