Michael,

Correct. The configuration directive "ALL_CASES_CORRODED" simply means that the section modulus and cross-sectional area used to compute the bending and axial stresses are corroded.

Why do the piping codes recommend that corrosion is handled in this way?

First, corrosion physically is a helter-skelter affair where local areas and spots of the pipe thin. Corrosion is not a uniform boring of the interior of the pipe. The problem is you don't know where these corroded spots are. To reduce the pipe weight by the corrosion allowance would be non-conservative because you would be reducing the load.

Second, to my knowledge, there have been no SIF tests on corroded pipe. Even if there were such tests, your pipe would corrode differently than the specimens used in the tests. So SIFs and K-factors are based on uncorroded pipe.

Third, due to the unknowns of the corroded areas (location and depth), the flexibility analysis is performed with the uncorroded cross section. This means the solution of [K]{x} = {f}, which is {x} is the same whether you include corrosion or not. The only difference is in the final stresses. All other results (displacements, forces, moments, and restraint loads) will be the same. This is because [K] and {f} are (assumed to be) not affected by corrosion