Skip to content

Latest commit

 

History

History

Crack1

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Energy Release Rate (Global Energy Method)

Tested with CGX 2.19 / CCX 2.19, Gmsh 4.9.3

  • Linear static solution
  • Symmetric model
  • Parametric modeling and meshing with Gmsh
  • Command line parametrization with param.py
  • Evaluation of the total strain energy
File Contents
ct.fbd Full analysis control script for CGX
par.ct.geo Parametric geometry and meshing script for Gmsh
solve.inp CCX input
path.fbd CGX script for the path plot
path.plt Gnuplot script for the path plot
test.py python script to run the simulation

Reference Solution and Evaluation

The energy release rate is determined from two simulation runs with the present model. This is used to determine the shape factor Y, which is compared to an empirical estimate from the standard.

a in mm Strain energy of the quarter model in Nmm
20.5 0.05009
21.0 0.05309

Screenshot

Model Description

The model is a CT (compact tension) specimen with the simulation domain reduced to one quarter of the part due to symmetry.

The specimen is loaded by a force and the total strain energy is determined. The force is generated by a uniform pressure to the upper half of the pin hole.

From the difference of the strain energy for two different crack lengths the energy release rate can be calculated.

In order to determine the energy release rate, run the model twice with different settings of the crack length a.

For each run, record the crack length a and the strain energy from the file solve.dat.

Set the crack length:

> param.py par.ct.geo a=20.5

Run the analysis:

> cgx -b ct.fbd

Preprocessing

Run the mesh generation with optional specification of the crack length

> param.py par.ct.geo a=20.5
> gmsh ct.geo

The script ct.geo creates a quarter model of the ct specimen using the gmsh bottom-up procedure to produce the cross section shape and then extruding it into the thickness.

The geometry is driven by the parameters B (width of the specimen) and the crack length a. You might also wish to adjust the half width of the initial cut of the specimen t for cosmetic reasons.

The mesh density is controlled by the global maximum for the element size and by the element size at the three points in the crack plane.

The faces for symmetry constraints and load application are defined as so called physical groups in Gmsh. This has the advantage that properly named node sets are written to the output gmsh.inp.

The mesh then is read into CGX, the surface elements from Gmsh are deleted and the node set for the load is extended into a face set.

In the solver input file solve.inp, a uniform pressure is applied to set Load and appropriate constraints are applied to the sets Ysym (at y=0) and Zsym (at z=0). A single node at the origin is constrained in x-direction.

A linear static analysis is performed with material values for steel. The displacements and stresses are written to solve.frd and the total strain energy is written to solve.dat.

Postprocessing

The quarter model is expanded with the base region colored by the equivalent stress. A custom max value has been set for the color bar in order to prevent the non-blue colors being restricted to the vicinity of the crack tip.

Once the analysis is finished, the contents of solve.dat is displayed for convenient access.

Path Plot

Run

> cgx -b path.fbd

to create a path plot of the normal stress (normal to the crack plane) and of the crack opening (displacement in load direction) along the centerline of the specimen (x-axis). The path plots are created separately in CGX and then combined using a gnuplot script.

The negative stress values to the right of the crack tip are artifacts due to the extrapolation from the integration points to the nodes.