3. Modeling A#

3.1. Characteristics of modeling#

3D modeling is used.

The coefficient \({K}_{D}\) is chosen depending on the temperature and on \({\sigma }_{0}\).

T (°C)

0 MPa

100 MPa

200

200 MPa

900

14,355

14,855

15,355

1000

14.5

15

15.5

1025

14.5363

15.0363

15.5363

15.5363

1050

14.5725

15.0725

15.5725

Table 3.1-1 : K_D of modeling A

The discretization in time is quite fine:

(JUSQU_A = 2, NOMBRE = 10),

(JUSQU_A = 2., NOMBRE = 10),

(JUSQU_A = 20., NOMBRE = 10),

(JUSQU_A = 200., NOMBRE = 10),

(JUSQU_A = 2000., NOMBRE = 10),

(JUSQU_A = 20000., NOMBRE = 10),

(JUSQU_A = 200,000., NOMBRE = 10),

(JUSQU_A = 1000000., NOMBRE = 30),

(JUSQU_A = 1600000., NOMBRE = 30),

(JUSQU_A = 1700000., NOMBRE = 40),

(JUSQU_A = 1800000., NOMBRE = 40),

(JUSQU_A = 1900000., NOMBRE = 40),

(JUSQU_A = 2000000., NOMBRE = 40),

(JUSQU_A = 2100000., NOMBRE = 40),

(JUSQU_A = 2200000., NOMBRE = 40),

(JUSQU_A = 2300000., NOMBRE = 40),

(JUSQU_A = 2400000., NOMBRE = 40),

(JUSQU_A = 2500000., NOMBRE = 40),

3.2. Characteristics of the mesh#

It is chosen to represent the cylindrical test piece by a block in order to be able to perform a calculation on a single element. The modeling has 3 planes of symmetry

Number of knots: 8

Number of stitches: 1 (HEXA8)

_images/100002010000027E000002A9D2D338CDC42533CB.png

Figure 3.2-1: Modeling A mesh

3.3. Tested sizes and results#

Two calculations are performed, the first with an explicit integration algorithm (ALGO_INTE =” RUNGE_KUTTA “), the second with an implicit integration algorithm (ALGO_INTE =” NEWTON”).

Calculation in explicit resolution:

Evolution of the constraint, \({\mathrm{\sigma }}_{0}\), as a function of time. This value is tested at various times:

Instant

Reference Type

Reference

Tolerance (%)

20

“ANALYTIQUE”

252.76091

0.5

2000

“ANALYTIQUE”

164.261

0.5

200000

“ANALYTIQUE”

101.596

0.5

1000000

“ANALYTIQUE”

75.97849999999999997

0.5

1600000

“ANALYTIQUE”

55.54209999999999998

10.0

Table 3.3-1 : Results of modeling A

Evolution of the damage variable, \(D\) as a function of time. This value is tested at various times depending on the modeling:

Instant

Reference Type

Reference

Tolerance (%)

20

“ANALYTIQUE”

2.3168400000000001E-4

0.5

2000

“ANALYTIQUE”

2.77144E-3

0.5

200000

“ANALYTIQUE”

0.032255100000000002

0.5

1000000

“ANALYTIQUE”

0.110134

1.0

1600000

“ANALYTIQUE”

0.28131600000000001

10.0

Table 3.3-2 : Results of modeling A

Evolution of the viscoplastic isotropic work hardening variable, \(r\), as a function of time. This value is tested at various times:

Instant

Reference Type

Reference

Tolerance (%)

20

“ANALYTIQUE”

1.6445100000000001E-3

0.5

2000

“ANALYTIQUE”

2.2312600000000001E-3

0.5

200000

“ANALYTIQUE”

2.6251500000000001E-3

0.5

1000000

“ANALYTIQUE”

2.747799999999999999E-3

0.5

1600000

“ANALYTIQUE”

2.79276E-3

0.5

Table 3.3-3 : Results of modeling A

Evolution of the viscoplastic isotropic work hardening variable, \(p\), as a function of time. This value is tested at various times:

Instant

Reference Type

Reference

Tolerance (%)

20

“ANALYTIQUE”

1.644569999999999999E-3

0.5

2000

“ANALYTIQUE”

2.231879999999999999E-3

0.5

200000

“ANALYTIQUE”

2.6301200000000001E-3

0.5

1000000

“ANALYTIQUE”

2.7607899999999999E-3

0.5

1600000

“ANALYTIQUE”

2.814779999999999998E-3

0.5

Table 3.3-4 : Results of modeling A

Calculation in resolution implicit :

Evolution of the constraint, \({\sigma }_{0}\), as a function of time. This value is tested at various times:

Instant

Reference Type

Reference

Tolerance (%)

20

“ANALYTIQUE”

252.76091

2.5

2000

“ANALYTIQUE”

164.261

2.5

200000

“ANALYTIQUE”

101.596

2.5

Table 3.3-5: Results of modeling A

Evolution of the damage variable, \(D\) as a function of time. This value is tested at various times depending on the modeling:

Instant

Reference Type

Reference

Tolerance (%)

20

“ANALYTIQUE”

2.3168400000000001E-4

3.5

2000

“ANALYTIQUE”

2.77144E-3

4.0

200000

“ANALYTIQUE”

0.032255100000000002

7.0

Table 3.3-6: Results of modeling A

Evolution of the viscoplastic isotropic work hardening variable, \(r\), as a function of time. This value is tested at various times:

Instant

Reference Type

Reference

Tolerance (%)

20

“ANALYTIQUE”

1.6445100000000001E-3

2.5

2000

“ANALYTIQUE”

2.2312600000000001E-3

1.0

200000

“ANALYTIQUE”

2.6251500000000001E-3

1.0

Table 3.3-7: Results of modeling A

Evolution of the viscoplastic isotropic work hardening variable, \(p\), as a function of time. This value is tested at various times:

Instant

Reference Type

Reference

Tolerance (%)

20

“ANALYTIQUE”

1.644569999999999999E-3

2.5

2000

“ANALYTIQUE”

2.231879999999999999E-3

1.0

200000

“ANALYTIQUE”

2.6301200000000001E-3

0.5

Table 3.3-8: Results of modeling A

Note: This calculation does not converge beyond the instant 200000.