Knowledge about the frictional characteristics of materials is required for equipment design, either to exploit the friction or to overcome it. This paper deals with studying the coefficients of sliding friction of non-treated and extrusion-exploded wheat straw and rice straw on mild steel and aluminum surfaces and the effects of several factors, including the moisture content (at five levels 35, 45, 55, 65, and 75%), normal pressure (at the following five levels: 2, 4, 6, 8, and 10 kPa), and length (at the following five levels: 30, 60, 90, 120, and 150 mm). Through the use of a statistical analysis method, we found that the moisture content, normal pressure, and interactions between them had statistically significant effects on the coefficient of sliding friction (p < 0.01). The coefficient of sliding friction was much larger than that of rice straw on both contact surfaces, and the coefficient values of both straw samples on the aluminum surface were notably higher than those on mild steel. A significantly higher value was found for the coefficient of sliding friction of extrusion-exploded straw compared with that of untreated straw on both surfaces. As the normal pressure increased from 2 to 10 kPa, the coefficient of sliding friction decreased first, and then increased sharply, finally fluctuating slowly for both kinds of straw and on both tested surfaces, with the maximum value was observed at a normal pressure of 6 kPa. When the length increased from 30 to 150 mm, a slightly linear upward trend was observed for the coefficient of sliding friction of untreated wheat straw, and the coefficient of sliding friction of untreated rice straw initially increased and then decreased.

A force-feeding device with a double-roller anisotropic was designed for the D200 single screw straw fiber extruder to keep the performance continuity of the system, which could improve the productivity of straw fiber. Four factors (the diameter of the auxiliary roll, the difference in linear speed of the two rolls, the gap between two rollers and the spindle speed as the test factors) were investigated to establish regression model to analyze the influence of the coupling of multiple factors on the test indexes. It was demonstrated that significant effects (p < 0.05) on the feeding rate and material loss were produced by four factors according to the experimental results. The sequence of influence in descending order on feeding rate was spindle speed, linear speed difference, gap and auxiliary roller diameter. The effects on materials loss in descending order were auxiliary roller diameter, line speed difference, spindle speed and gap. The optimal combination of parameters was obtained by the response surface, which were an auxiliary roller diameter of 230 mm, a spindle speed of 104.49 rpm, a line speed difference of 2840 mm/s and a gap between the two rolls of 14 mm. The average feeding rate was 2.3798 t/h, and the loss was 1.908 kg/h, and the errors were within 3.28%, which satisfied the feeding requirements for rice straw fiber production with high feeding efficiency and low raw material losses. This study provided a reference for the forced compression feeding and fibrillation process of rice straw.