Hi All - In another thread there has been discussion about backlash of belts and since Frankie uses belts and I have been wanting to do this particular model for some time I did it this morning. The aim is to determine how many teeth distribute the load on the belt. So I modelled a straight bit to keep it simple. The PU belt is the same size as an AT5 belt but the result will be similar on all tooth shapes. The Al "rack" has a stiffness of 70,000MPa the PU is 6.0Mpa and the steel "belt" is 11,700MPa. This is the equivalent stiffness of the 2.5x1.5mm area of the cords that run through it, calculated from the quoted specific stiffness of an AT5 belt of 17605N/mm width. The cords bundle is a 7x3 with an outside dia of 0.9mm which means there is approx 1 cord per mm width of belt. My 10mm belts have 10 cords and the 16mm I have has 17 cords. I have used friction co-efficient of 0.2 at the contact surfaces. In the images you can see that the first tooth does most of the work, the second a bit the third not much then past that they are not doing anything. So if you assume that 4 teeth are doing the load transfer you conservative. this model uses zero clearance teeth as I couldn't find the actual clearance of the teeth but this would not change the trend. Peter