A multi-scale study of the permeability of compressed nonwoven filter media

The deformation of filter media and its impact on their flow and filtration properties has become an aspect of increasing importance when predicting filter efficiency and lifetime. While the deformation of the media due to the flow during operation was the subject of many experimental and simulation studies, only a few works addressed the effects caused by deformations of nonwoven filter materials due to manufacturing. In particular, the change of material properties due to compression was the object of research. Media compression occurs during calendering, combination of media with supporting or spacer meshes and pleating. In each of these scenarios, it turned out that taking into account the changes in the material volume fraction of the nonwoven is essential for an appropriate computer-aided prediction of flow and filtration. The present contribution is devoted to the investigation the permeability of compressed nonwoven filter media, similar to a previous work that was motivated by a better understanding of flow-induced material compression. A filter material used in oil filtration was chosen for the investigation. For the experimental part of the study, several samples of the material were compressed by applying different mechanical loads. A sufficiently long waiting time ensured that there was no further relaxation of the nonwoven. For each load (including the uncompressed media), the media thickness was measured following the lines of ISO 9073 in order to obtain the corresponding compression level. The intrinsic permeability of the compressed media was derived from measurements of the air permeability according to ISO 9237. On the microscopic level of the fibers and pores of the media, computer simulations of flow using CT images taken from several samples of both compressed and uncompressed samples. From the pressure drop obtained in such a flow simulation, the corresponding permeability of the material can be deduced. In order to reduce the CT data required, computational structural mechanics was additionally used to simulate the compression of the original material. Repeating the flow simulations for the virtually compressed nonwoven, another set of permeability values is obtained. In doing so, filter media designers could predict the permeability of the material depending on the level of compression based on a reduced set of CT data. Once validated, the computer-aided prediction could be done for nonwoven structures generated by suitable software tools. This paper presents and discusses the results of the multi-scale study of compressed filter media.