The BarrierFunctionMaterial material may have some problem

[bdliangxy]

I'm using Phase field module to solve the classical Allen-Cahn Equation:

$$ \frac{\partial \xi}{\partial t} = -L_{\sigma} \left( g'(\xi) - \kappa \nabla^{2} \xi \right) $$

It should keep the interface standing. I separately use BarrierFunctionMaterial(input2.txt) and hand-writing DerivativeParsedMaterial(input2_correct.txt) to represent $g(\xi)$. But the one use BarrierFunctionMaterial have a interface gradually widening,shown in the first picture. Another one use hand-writing $g(\xi)$ have a constant interface, shown in the second picture.
Maybe I use BarrierFunctionMaterial wrong or others. I'm her to ask for you opinioin.
input2.txt
input2_correct.txt


Both pictures show the result have been simulated for 1000s.

The code in input files math the 2022 version moose, hope you don't care.

[GiudGiud]

It should keep the interface standing.

why would it not diffuse? I'm not familar with the physics

Maybe I use BarrierFunctionMaterial wrong or others

Can you output the material properties when using this material? (using outputs = exodus) and see if they look as expected.

The code in input files math the 2022 version moose, hope you don't care.

you're missing 4 years of bug fixing
What's the motivation?