FRACTURE MECHANICS

[phanhung]

Hi

i am working with DEBOND two surfaces, my file get this kind of massage: "THE SLAVE-MASTER PAIR ASSEMBLY_DB2 ASSEMBLY_DB1 SPECIFIED ON *DEBOND HAS NOT BEEN DEFINED. THIS MAY BE DUE TO NONEXISTENT SLAVE OR MASTER SURFACE, OR DUE TO ERRORS WHEN PROCESSING THE *CONTACT PAIR KEYWORD". Althought i have already defined it in CONTACT PAIR and again in INITIAL CONDITIONS. Please help me how to solve this problem.

Thanks

Phanhung

 

[phanhung]

Hi

i am using
*DEBOND
*FRACTURE CRITERION, TYPE=COD
(and *FRACTURE CRITERION, TYPE= CRACK LENGTH)

But two surfaces did not seperate (no cracking propagation), please tell me which parameters should be consider to deal with two obove cases.

thanks

Phanhung

 

[xerf]

I am not sure exactly what kind of problem you encounter.

If the bonded surfaces do not separate, it means the COD criterion is not fulfilled given the current loading.

The parameters you should consider are explained in the documentation, basically: for the TYPE=COD, you must supply pairs of COD vs. crack length whereas for TYPE=CRACK LENGTH you must supply the total time vs. crack length.

 

[phanhung]

Thanks for your information
i wrote my propram as below:

1. debond the first node
*Debond, Slave=DB1, Master=DB2,
Time increment=0.005, Frequency=1, Output=BOTH
0.0, 1.0
0.005, 0.0
*Fracture Criterion, Type=COD, Tolerance=0.1,
Distance=1.0850, Symmetry
2.65, 0.86

2. debond the second node
*Debond, Slave=DB1, Master=DB2,
Time increment=0.001, Frequency=1, Output=BOTH
0.0, 1.0
0.005, 0.0
*Fracture Criterion, Type=COD, Tolerance=0.1,
Distance=1.09, Symmetry
2.89, 1.21


i have tried many values of distance and COD pairs, but the result still the same. which others parameter can effect to this problem (bonded surfaces do not seperate).

 

[xerf]

Are you trying to debond the surface node by node by yourself ?? (You do not need to put more than one set of DEBOND/FRACTURE CRITERION per crack/per step)

-the Distance parameter for the FRACTURE CRITERION refers to the position with respect to the current crack-tip.
As the surfaces debond, the crack-tip moves and the program will evaluate the propagation criterion with respect to the new crack-tip position automatically.

-also, in order to check the functionality of your model you could try to put a small value for the COD criterion.

 

[phanhung]

I tried another smaller values as your suggestion, but the result still the same.

if i put DISTANCE value reach > 3.16 (or around this value) then it get error messege (too many attempt..) but the value is smaller, debonded surfaces were still not seperated.

Anything else effect to this problem such as MATERIAL, BOUDARY CONDITION, INCREMENT ...??

i do know how to send for you my model, but please help me to take a look on the debond file
*Heading
*include, input=fracturecrack_model.inp
**------------------step1, load apply-------------------------------
*Step, name=crackanalysis, nlgeom
*Static
0.001, 1.0
*Boundary, type=VELOCITY
velocitynode, 1, 1, 0.2
velocitynode, 2, 2
*el print, frequency=0
*node print, frequency=0
*contact print, Slave=DB1, Master=DB2, frequency=10
DBT, DBSF, DBS
*Contact file, Slave=DB1, Master=DB2, frequency=10
DBT, DBSF, DBS
*print, contact=yes
*End Step
**-------------------step1, debond--------------------------
*Step, nlgeom, inc=1
*Static
0.005, 0.005, 0.00001
*Boundary, type=VELOCITY
velocitynode, 1, 1, 0.2
velocitynode, 2, 2
*Debond, Slave=DB1, Master=DB2,
Time increment=0.005, Frequency=1, Output=BOTH
0.0, 1.0
0.005, 0.0
*Fracture Criterion, Type=COD, Tolerance=0.1,
Distance=1.0850, Symmetry
2.65, 0.86
*controls, parameters=field, field=displacement
, 0.1
*Contact file, Slave=DB1, Master=DB2
DBT, DBSF, DBS
*contact print, Slave=DB1, Master=DB2, frequency=10
DBT, DBSF, DBS
*print, contact=yes
*EL Print, Frequency=0
*Node Print, Frequency=0
*End step

 

[xerf]

You should try to use smaller time increments for the steps.

 

[phanhung]

Thanks for your help, my model can be said ok, now.For another question,
1. i want to plot some curves like curves in the EXAMPLE PROBLEMS MANUAL - 1.4.4 CRACK GROWTH IN A THREE POINT BEND SPECIMENT. In data XY PLOT, we have many choices to plot, but i can not find which part will be chosen to plot these curves.
2. Do these curves describe for whole model or just for one special node??

Thanks