[victorkemp]

C. MITC4+: thermal moment for coupled laminates

• Bug: MITC4's thermal load has [Bb]' SHELL_BALP (TBAR-TREF), but SHELL_BALP was always 0 for a PCOMP. It used ALPVEC(:,4), which is never set for a ply.
  • A stiff/soft strip with the same CTE in both plies curled (TZ = -0.30) and stretched 4x, instead of expanding freely.
  • A bimetal got the wrong curvature.
• Fixes:
  • PCOMP: SHELL_BALP = sum PLY_B alpha_ply.
  • PSHELL MID4: SHELL_BALP = [B]{alpha4}.
  • The thermal moment is now subtracted in the MID4 force recovery.
• Tests: free strips with DT = 100:
  • a steel/aluminum bimetal (CLT, equal to Timoshenko's formula to 1e-16)
  • a [0/0/90/90] carbon laminate with orthotropic CTE
  • stiff/soft plies with uniform CTE
  • each as PCOMP and as PSHELL MID4
  • checks: tip displacements, PCOMP ply stresses, and zero N/M on the PSHELL strips, all exact to f06 precision.
• Scope: MIN4 and TRIA3 don't use SHELL_BALP and are unchanged; they have the same thermal gap on dev.