# Residual Stress Calculation with Abaqus

Hello There,

I'd like to discuss here my result about residual stress. I calculated with Abaqus a thermal transient from 20 to 300°C +relaxation(unconstrained, keeping the same final T=300°C) on a planar model of a composite material (simplified).  The 2 circles are tungsten (W), the Matrix is copper (Cu). Stress (eigenspannung) should arise from mismatch of expansion coeffiecients (CTE_W~0.3*CTE_Cu), but it should also desappear after relaxation, since circles and matrix are free to expand and both modeled as pure elastic phases. Despite the model is perfect elastic and free to expand in x and y (only BOTTOM nodes are fixed in y), I get non-zero stress across the region.

Why? Does it make any physical sense?

I actually expected  zero stress everywhere after relaxation.

I tried both coupled temperature-displcement and uncoupled procedures in Abaqus. I used in the first case CPS3T elements and in the second one DC2D3 (Heat Transfer) + CPS3 elemnets.

Alex

*************

My Step looks like this:
*Boundary
BOTTOM, 2, 2

** STEP: Step-1  #### --> HEATING 20-300
**
*Step, name=Step-1, nlgeom=YES, inc=1000
*Coupled Temperature-Displacement, DELTMX=1
0.0001, 1, 1e-10, 0.002
*Boundary
ALLN, 11, 11, 300. ## ALLN is a node set which includes all nodes

** STEP: Step-2 ############## RELAX
**
*Step, name=Step-2, nlgeom=YES, inc=300
*Coupled Temperature-Displacement
0.01, 1, 1e-10, 0.05

Free Tags: