Abstract: Fiber reinforced epoxy resin insulation pull rod is a key component for mechanical transmission and electrical insulation in gas insulated switchgear(GIS) circuit breakers. However, residual stress will be generated inside due to curing shrinkage and other factors, which can easily cause damage to the fiber-resin interface under the superposition of breaking loads, leading to mechanical failure or insulation breakdown. Currently, there is a lack of effective non-destructive detection technology, resulting in operating with defects. A residual stress detection technology based on the acoustic-elastic effect is proposed. A dual axis synchronous sliding table and a rotating unit are introduced to establish an automatic detection platform for insulation pull rods. The influence of residual stress on the ultimate load of insulation pull rods is verified through real type experiments. The results show that the automatic detection platform can automatically detect and image the residual stress of the insulation rod of the ultra-high voltage GIS circuit breaker. The risk of mechanical failure induced by residual stress of the insulation rod is classified by residual stress normalization. Finally, the experimental results verify that the amplitude and concentration position of residual stress of the insulation pull rod are consistent with its ultimate load and failure position. The research is of great significance for improving the manufacturing process and enhancing the detection capability of insulation pull rods.