I am simulating a dropping object that falls under gravity force into a liquid domain. But when I add mass scaling, the gravity forces (applied via Gravity load in abaqus) do not seem to work and the object either doesn't fall or falls too slowly. Apparently when I apply a mass scaling with a factor of X, the gravity force scales down by a factor of 1/X. Is this correct and is there any way to deal with this problem without changing the gravitational acceleration?