Constitutive modelling of compression and stress relaxation in pine pellets

Highlights: • Mechanical and rheological parameters of six different pine particle sizes. • Constitutive model for simulating compression and stress relaxation. • Comparison of existing constitutive models. • Novel pseudo spring porosity for modelling non-linearity. • Linear relation between spring modulus and energy required for compression. -- Abstract: The increasing pellet production and a demand for making high quality biofuel pellets call for tools that can facilitate producers to meet these requirements and help understanding the effect feedstock and process parameters. In this study, mechanical and rheological properties of pine pellets made of different particle sizes and compression speeds were studied via pelleting tests and numerical simulations. Single pelleting tests were performed with six different particle size samples, ranging between 0.25 and 2.8 mm, and pelleted at compression speeds of 1, 5, and 10 mm min⁻¹. The experimental results of specific compression and extrusion energy showed a positively linear correlation between particle size and energy consumption. The highest pellet durability was observed for pellets produced from small and mixed particle sizes. Eight different constitutive models were evaluated on their ability to simulate compression and stress relaxation, and their level of complexity. A non-linear Maxwell representation of the Standard Linear Solid (SLS) model was setup and fitted to the experimental compression data. The model coefficient of spring 1 composes the asymptotic stress level of the relaxed pellet, and the coefficient of spring 2 was found to be positively correlated with particle size. The viscosity of the dashpot is also found to be positively correlated with particle size, likewise it depends on the compression speed, where higher compression speed resulted in lower viscosities. The results of the study elucidate new insight into mechanical behavior of biomass particle compression, and the resultant simulations have utility for predicting the pressure requirements to produce pellets.