Disclaimer: I have never used a 1/32" ball endmill in HDPE either, however I’ll say that whenever I have had issues when not making well-formed chips while cutting plastics (especially HDPE), the answer as always been “feed faster” for me.
30ipm at 10.000 on what I assume is a two-flute endmill grants a chipload of 0.0015".
For 1/16" I would use a 0.002’‘–0.003’’ chipload as a starting point, so on a 1/32" that 0.0015" value makes sense, still you could try to use the feedrate override during a test cut to see if bumping it by increments of 10% improves things. I would do this test while also reducing depth of cut by 50%, for the sake of keeping tool deflection to a minimum and not take any chances with such a fragile tool.
FWIW, here’s a thread where I used a 1/16" endmill in C3D’s two-color HDPE, which turned out super clean:
I’m with @Lowbrowroyalty : if you can buy your way out of this problem, consider getting another endmill.
If I had to do it and a 1/32" was mandatory, I would pick something like this :