Explaining the large variability in empirical relationships between magnetic pore fabrics and pore space properties

The magnetic pore fabric method was proposed in the 1990s as a fast and efficient way to characterize the preferred shape and alignment of pores in sedimentary rocks. Since then, magnetic pore fabrics have been interpreted based on empirical relationships, stating for example that the orientation of magnetic pore fabric relates to pore alignment, and that degree and shape of the magnetic anisotropy correlate with pore aspect ratio and pore shape. Unfortunately, these empirical relationships vary largely between studies, making it challenging to interpret the data quantitatively. Here, we are presenting numerical models and experiments that help explain the variability in published relationships. Magnetic pore fabrics are determined by first impregnating the samples with a suitable ferrofluid, followed by magnetic anisotropy measurements in various fields and at multiple frequencies. These data are compared to known pore properties in synthetic samples, or X-ray-computed-tomography-based models of the pore space in natural samples. Our data show how the magnetic properties of various ferrofluids vary with measurement conditions, in particular measurement frequency. Ultimately, the effective susceptibility of the ferrofluid at measurement conditions largely affects measured magnetic anisotropy, and thus the observed empirical relationships. These findings allow to re-interpret existing empirical relationships and their variability, and also provide recommendations for future magnetic pore fabric studies.