Intuitionally part thickness uniformity is often attributed to sink marks, but most researchers have concluded that such conception is very vague or subjective!
Instead it is part thickness distribution with respect to injection point (gate) towards the extreme point on the melt flow path i.e, last fill point. A good injection molding part thickness should always decrease (not uniform) along the flow path and that thickness decrease magnitude is the prime factor responsible for sink marks.
It also needs to know the condition of the part geometry that you are trying to mold. Does it get sink over the top of a rib, Part feature or are you getting sink across a flat surface with no back side features? If the condition is over a rib or a back side feature, it will need to know what your rib to wall stock ratio is and the material you trying to mold. If there is no back side feature it will still need the material you are molding and the material thickness as well as a copy of the process you molding with.
With this part being ABS you want to have a rib to wall stock ratio under 40%. Anything over that will have a tendency to show sink on the class A due to mass build up behind it. If you are under the 40% rib to wall stock ratio then we will need to see your process sheet and a part & runner lay out in the tool.
Sudden thickness variations directly contribute sink marks due to differential cooling and insufficient packing. So the issue can be best tackled by a cautiously controlled thickness distribution in product design.
Most of the sink marks are cause of poor design or poor process control on the machine. The reason for the sink mark is difference of temperature between the outer shell of the part and the inner core (usually the inner core is hotter). so if you eliminate massive heat in the inner core (like reducing wall thickness of ribs or bosses) you will have a fairly good solution but you need also to control the process (cooling time, packing etc….), processing also independently contribute to sink marks.., Basically injection (processing) pressure and temperature are primarily polymer and then mold and machine dependent. i.e., injection melt state variables., So part form / features have very less to do with setting injection melt state.
However a good control to prevent sinks can be achieved by increasing packing intensification rate, this technique can considerably compensate thickness distribution variance also, which is quite often.