If the fluid contains ferrite particles they can accumulate on the coupling magnet and abrade the housing. Any type of particle can accumulate and cause wear if there isn't much circulation in that area. Mag drive pumps are not tolerant of solids

What resin is used for the housing? It does appear stained or discolored. That may or may not indicate some deterioration of the resins (plastics are permeable and stain easily).

On an unrelated note, make sure you toss those split lock washers in the garbage. The gap let's corrosive liquids into the threads and causes seized fasteners. That one is also failing in a common mode for split washers.

You need to capture and analyze the debris in your system.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

Are there two separate containment shells? If so it's hard to see the exact scratch in the first picture.

There definitely is material accumulating around the outer magnet. Most adapter brackets will have multiple threaded ports to tighten the outer magnet to the motor shaft as well as to provide some type of leak detection connection. I would make sure there are threaded plugs in those to prevent dust ingress. You could also look at some type of grease/sealant between bracket/wet end and bracket/motor although that's probably overkill for all but the dustiest of environments.

The cut containment shell definitely looks like material accumulating on the ID of the outer magnet. It's also worth checking to see if there are any dents in that surface. If a mechanic put a tool too close to the magnets and dented that magnet cover could also explain that cutting but that would typically occur minutes (if not seconds) after start-up.

@TheHandModel
It was the same containment shell, the bottom picture was when pressure test (operating pressure) was done on the containment shell. The crack formed shortly after.

I suspect that the dust come from condensed nitric acid that leaked through the containment shell during normal operation.

@jsasada

If nitric acid is leaking through the containment shell normally, scratches seem like a secondary issue. Are there any signs of damage on the inside of the containment shell that would indicate a possible leakpath to the composite material? 40% nitric is below the fuming concentration and quite far away from the boiling point so any gas phase material that could easily permeate through the fluoropolymers should be extremely low.