Modelagem numérica para predição da perda de pré-tensão em juntas aparafusadas

Abstract

Bolted joints are widely employed in mechanical and structural systems due to characteristics such as versatility, ease of assembly, and the ability to withstand complex loading conditions. However, under dynamic loading, these joints may experience a progressive loss of preload, a phenomenon known as self-loosening, which compromises structural integrity and may result in catastrophic failure. The present work conducts an experimental and numerical investigation of preload loss in bolted joints subjected to transverse dynamic excitation. An experimental apparatus was developed, consisting of two flange-type steel plates connected by a stud manufactured from ASTM A320 L7 steel, instrumented with electrical resistance strain gauges positioned inside a longitudinal groove machined along the bolt axis, enabling continuous and direct monitoring of preload evolution. Dynamic loading was applied using a servo-hydraulic testing machine operating under displacement control, with three distinct configurations investigated, varying the stud nominal diameter and the excitation frequency. In parallel, a finite element model was developed, incorporating contact interactions, frictional effects, preload application, and cyclic transverse loading conditions. A friction coefficient correction methodology based on the effective contact area between thread flanks is proposed to compensate for the geometric idealization of the model, in which simultaneous contact on both thread flanks overestimates the frictional resistance observed in the tests. The corrected model successfully reproduces the progressive preload reduction for all three investigated cases, showing good agreement with the experimental results. These findings highlight the decisive role of the thread interface friction coefficient in the self-loosening mechanism and demonstrate the potential of the finite element method as a complementary tool to experimental testing, enabling parametric investigations with lower operational costs and greater analytical flexibility.