While performing Rockwell hardness tests (HRC scale) on a steel rod, I noticed that the diamond indenter caused the surface of the rod to slightly bend downward during the test; however, this bending was less prominent when I raised the sample stage such that the rod firmly pressed against the indenter head; this is a cone-shaped chamber with a hole in the middle, from which the indenter goes in and out.
Through many tests, it has become evident that there is greater scatter when there is visible bending of the rod, but I haven't been able to identify the relationship between the degree of bending and measured hardness.
EDIT: I am already aware that bar deflection will scatter results. What I am trying to understand is how it does so. Looking at it from a logical standpoint, if the bar bends away from the indenter, this will trick the hardness tester into thinking the indenter penetrated a deeper distance, registering the measured hardness as systematically lower; however, this is not necessarily the case.
Thank you!