The Validity Range of Low Fidelity Structural Membrane Models

Abstract

Three classes of structural models can be considered for the characterization of thin extensible pressurized membranes. A low-fidelity, linear model assumes that all of the membrane’s resistance to a transverse pressure is provided from geometric stress-stiffening, and that movement along the membrane is purely out-of-plane. A medium-fidelity model can include geometric nonlinearity (finite strains, non-conservative pressure loads); while the highest level of membrane modeling assumes material nonlinearity: specifically, hyperelasticity. Common modeling applications such as performance prediction, failure prevention, and structural optimization all require repeated function evaluations. As computational cost is proportional to model fidelity, what is the validity range of the two lower fidelity models discussed above? The presence of a large pre-tension within the membrane is known to lessen the role of nonlinear elastic effects: the range of linear transverse deformation increases with pre-tension. Conversely, very low pre-tensions require the use of a nonlinear model, as the linear model becomes unbounded. This work studies the Hencky–Campbell problem for model validation purposes: hydrostatic inflation of a thin flat circular membrane, clamped along its boundary, with an arbitrary initial tension. A full-field, non-contact visual image correlation system is used to estimate the material properties of the rubber membrane, measure the state of pre-tension in the circular sheet, and document the displacement and strain fields as a function of applied pressure. The resulting data set is then compared directly with numerical simulations, in order to estimate the location of the surface of data points wherein a particular low fidelity model (either linear or geometrically nonlinear) loses its predictive capability.