I don't fully grasp the process. Will the stir bar rotate the liquid or suspension? Then, due to viscosity at the walls, the bar will rotate faster than the liquid, and mix it, which defeats the purpose of centrifugation. Or does a bar used in a previous step stay in the liquid during centrifugation?
Then, being probably denser than the liquid, the bar will quickly drift to the wall and make the rotor unbalanced.
At the curved wall, the bar may sit on its ends or at its central bulge, both producing about the same stress. Let's take an AlNiCo bar of d=6mm L=30mm ρ=7200kg/m3 m=6.1×10-3kg, neglecting the polymer coating.
14.7km/s2 make a distributed load of two F=45N over two L=15mm half-lengths. In a bar with end supports, the bending moment is Γ=FX/2=0.34N×m. For an R=3mm cylinder, Γ=σ(π/4)R3 tells a stress σ=16MPa.
Thyssen-Krupp's datasheet (appended here under, rename it .zip, unzip it, see their page 13) announces 250MPa flexural strength. I expect exactly this parameter to vary horribly among the manufacturers. It's a sintered powder whose interesting properties are magnetic, not mechanical.
I feel the safety factor 16 a bit uncomfortable for an unsound material. It is brittle and behaves more like a good ceramic than like a healthy alloy.
So if you make the centrifugation once in your lab and can mitigate bad consequences, try your luck. But if the process step shall be repeated thousand times at a customer, I would experiment it many times before, or amend it.
If the walls are ferromagnetic, the bar will stick to them, but the force shouldn't break an AlNiCo. The bar may give steel walls a permanent magnetization that interferes with subsequent use.
The moving magnet will induce some voltages in the surroundings, including sensors. They may even brake the bar, which would then slip and stir your liquid.
If your bar is of rare earth instead, which you recognize by its much stronger sticking force to steel, everything gets much worse. Computed safety factor 6, pretty small, and the stuff is very fragile. Sticking to curved steel walls may very well break SmCo, which then snaps together and splats liquid out. Better avoid rare earth bars.