A 3D regression surface for the room temperature tightness gasket data reduction and bolt load design

The purpose of the present work is to propose a new approach for modelling the tightness behaviour of the gaskets used in bolted flange joints. This new approach consists of developing a mathematical model for a three-dimensional (3D) representation of the gasket tightness performance. Rather than considering a 2D graph for characterizing the complete gasket behaviour, a third axis is added to the Sg vs. Tp plot to dissociate the unloading cycles from initial gasket tightening. This leads to the definition of a surface that is represented by a simple polynomial equation that contains six coefficients that are determined by a simple regression calculation. In the first part of the paper, the new approach is tested through a database of 406 room temperature tightness (ROTT) tests performed on different gasket styles. Then, a statistical analysis of the predictions made with the new model demonstrates its ability to predict gasket leak rates much more accurately than it was previously possible with the gasket constants derived from the ROTT Draft 9 or 10 methods. It is also demonstrated that the new approach can be used to model successfully complex gasket behaviours such as the tightness hardening phenomenon. Secondly, the effect of gas pressure change on the leak rate is analysed. It turns out that at high gasket stresses, the actual scheme of the ROTT tests may not allow enough time for reaching a stabilized leak rate value because of a transitory time effect in the porous structure of gaskets. Then, in order to evaluate the characteristic waiting period (dwell time) before a stabilized flow rate is achieved following a change in the gas pressure level, a simple phenomenological analysis of the flow is performed.