Skip to content

Fix dimensional inconsistency in maximum-strain failure criterion #8

Description

@vsdsantos

Problem

max_strain_criteria compares strains directly against stress allowables. In src/failure_criterias.jl:17-29, the checks use m.Xt, m.Xc, m.Yt, m.Yc, and m.S12 as strain limits. These quantities have units of stress, so the criterion is dimensionally inconsistent and will generally classify realistic strains as safe regardless of the material stiffness.

Engineering consequence

Maximum-strain predictions are incorrect, especially when comparing materials with different E1, E2, and G12. The implementation can also silently produce meaningless results because Julia does not enforce physical units.

Proposed improvement

Compute allowable strains from the elastic constants and strengths, e.g. ε1t = Xt/E1, ε1c = Xc/E1, ε2t = Yt/E2, ε2c = Yc/E2, and γ12 = S12/G12. Apply the sign-dependent limits to the local strain components. Add documentation clarifying engineering-shear conventions and regression tests for uniaxial and combined loading.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions