Fill out the form to download

Required field
Required field
Not a valid email address
Required field

Documentation

3D Triaxial Load Primary Creep (NAFEMS Test 11)

This triaxial load primary creep validation case belongs to solid mechanics. This test case aims to validate the following parameters:

  • Creep material behavior.
  • Standard and reduced integration elements.
  • Automatic time stepping.

The simulation results of SimScale were compared to the analytical results derived from [NAFEMS_R27]\(^1\).

Geometry

The geometry consists of a cube with edge length \(l\) = 0.1 \(m\).

cube geometry triaxial load primary creep validation
Figure 1: Cube geometry for the present validation project.

The coordinates for the points in the cube geometry are as tabulated below:

ABCDEFGH
x00.10.1000.10.10
y000.10.1000.10.1
z00000.10.10.10.1
Table 1: Cube dimensions in meters.

Analysis Type and Mesh

Tool Type: Code Aster

Analysis Type: Nonlinear static

Mesh and Element Types: The meshes for cases A and B are second-order hexahedral one-element meshes. They were created locally and imported to SimScale. For cases C and D, the standard algorithm was used to generate a second-order tetrahedral mesh.

CaseMesh TypeNumber of NodesElement Type
(A)2nd order hexahedral20Standard
(B)2nd order hexahedral20Reduced integration
(C)2nd order standard235Standard
(D)2nd order standard235Reduced integration
Table 2: Mesh characteristics.

Find below the mesh used for cases C and D. It’s a standard mesh with second-order tetrahedral cells.

cube second order standard mesh
Figure 2: Second-order standard mesh used for cases C and D.

Simulation Setup

Material:

  • Steel (linear elastic)
    • \(E\) = 200 \(GPa\)
    • \(\nu\) = 0.3
    • \(\rho\) = 7870 \(kg/m³\)
    • Creep formulation: Time hardening
      • A = 2.6041667e-46 \(1/s\)
      • N = 5
      • M = -0.5

Boundary Conditions:

  • Constraints
    • \(d_x\) = 0 on face ADHE;
    • \(d_y\) = 0 on face ABFE;
    • \(d_z\) = 0 on face ABCD.
  • Surface loads
    • \(t_x\) = 300 \(MPa\) on face BCGF;
    • \(t_y\) = 200 \(MPa\) on face CDHG;
    • \(t_z\) = 100 \(MPa\) on face EFGH.

Advanced Automatic Time Stepping:

For cases A through D, the following advanced automatic time stepping settings were defined under simulation control:

  • Retime event: field change;
  • Target field component: internal variable V1 (accumulated unelastic strain);
  • Threshold value: 0.0001;
  • Time step calculation type: mixed;
  • Field change target value: 0.00008.

Reference Solution

The equations used to solve the problem are derived in [NAFEMS_R27]\(^1\). As SimScale uses SI units, the reference solution was adopted to a time unit of seconds instead of hours.

$$\epsilon_{xx}^c = – \epsilon_{zz}^c = \frac{0.004218}{60} \sqrt{t} \tag{1}$$

$$\epsilon_{eff}^c = \frac{0.004871}{60} \sqrt{t} \tag{2}$$

$$\epsilon_{yy}^c = 0.0 \tag{3}$$

Result Comparison

Find below a comparison between SimScale’s results and the analytical solution presented in [NAFEMS_R27]\(^1\) for the average creep strain \(\epsilon_{xx}^c\) of the cube. The creep time is 3.6e6 \(s\) (equivalent to 1000 hours).

Case[NAFEMS_R27]SimScaleError (%)
(A)0.1333800.133123-0.192
(B)0.1333800.133123-0.192
(C)0.1333800.133107-0.205
(D)0.1333800.133107-0.205
Table 3: Comparison of SimScale’s results against an analytical solution.

In Figure 3, we can see how \(\epsilon_{xx}^c\), \(\epsilon_{yy}^c\), and \(\epsilon_{zz}^c\) are evolving for case D.

average creep strain plot
Figure 3: Average creep strain plot for case D.

\(\epsilon_{yy}^c\) and \(\epsilon_{zz}^c\) also show very good agreement with the analytical solution, having an error of 0% and -0.205%, respectively.

Last updated: July 21st, 2020

Contents
Data Privacy