Objective  As an indispensable process in the production and processing of medium density fiberboard (MDF), abrasive belt sanding directly affects the processing efficiency and surface quality. At present, the research foundation on the sanding of wood materials is relatively weak, and most of them use metal grinding and solid wood as the research object. There are very few researches on the abrasive belt sanding of wood-based panels. In actual production, the selection and replacement of abrasive belts mostly rely on the experience of workers. The research can provide scientific basis and theoretical support for the selection and replacement of abrasive belts in actual production, and promote the development of wood processing technology and abrasive belt manufacturing technology in the direction of high efficiency and intelligence.

  Method  This study used MDF as the experimental material to test the sanding efficiency and surface roughness during abrasive belt sanding. Combining with the three-dimensional topography, the study analyzed the effects of abrasive belt wear on the material removal rate and surface roughness, and made predictions on the life of the abrasive belt.

  Result  In the initial stage of MDF sanding, the abrasive grains were sharp and the material removal rate was high, but as the abrasive grains breaking and falling off, the material removal rate decreased faster. The material removal rate in the middle period of sanding tended to be dynamic and stable. In the later period of sanding, the material removal rate decreased again as the abrasive grains were further passivated and partially flaked off. As the number of sanding increased, the abrasive grains gradually became passivated and the diameter of the tip increased during the continuous wear process, leaving a widening sanding mark on the surface of the specimen, and the S_a and S_dr of surface roughness of the specimen showed a downward trend. The roughness parameter values in the early and late stages of sanding were relatively scattered, which in the middle stage were relatively concentrated and had good convergence. While the S_a and S_ku of the surface roughness of the abrasive belt was relatively concentrated in the early and late sanding stage, and the value in the middle stage was relatively scattered. In the case of the 120-mesh abrasive belt used in the test, when the material removal rate was used as the evaluation index, the cumulative sanding length of 23 096 m was the end of service life of the abrasive belt. When the S_a of surface roughness was used as evaluation index, the cumulative sanding length of 18 375 m was the end of the service life of abrasive belt.

  Conclusion  Abrasive belt wear to failure can be roughly divided into three stages. The sanding efficiency is directly proportional to the sharpness of the abrasive particles involved in the sanding and the number of effective abrasive particles. Abrasive belt wear has a greater impact on the surface roughness of sanding. When the purpose of the sanding process is different, the evaluation results of the abrasive belt life are different according to different inspection indicators. The life of the abrasive belt can be reasonably predicted by the changing trend of the material removal rate and the surface roughness.