You are here
Help to correct my problem for delamination with cohesive element
Hello,
I am Pradeep Master's student in Mechanical Engineering. I am trying to model an Abaqus model for a thin composite plate with cohesive element layer in between. For convinence I first tried to modle a cantiliver beam with cohesive layer in between the layer. I have modeled the program using Abaqus 6.9-2. The program is running and I am even getting the general contact-traction damage initiation to work, as per my limitaion to the knowledge, but I am not getting the cohesive elements to erode. Can anyone help in this topic. I am also attaching the input file for the program I have done. If anybody have already done something related to this please do send me the inputfile. Regarding my input file since I cannot attach the file extension with .inp I have converted the file to .doc and attached.Thank you
Delamination with cohesive element
Hello,
Is there anyone to help me with my above problem. I really need a help at this time, sooner the possible. Thank you.
Pradeep
Delamination with cohesive element
You have incompatible properties for your cohesive elements. Assuming,
that strength and energy release rate are physical properties of your
actual material (however the G looks too small), the stiffness should
be at least 1e15. You can use higher, but not less!!!
Hello Mr. Sergej ,
Hello Mr. Sergej ,
Thank you for the help. Do you mean to say that the G value for the cohesive element property is too small or it is for the SiC/SiC material. This were the material properties I got from one of the paper where they conducted a actual test for SiC/SiC composite.
Thank you.
Cohesive stiffness
Hello!
I don't know the material properties for SiC/SiC. May be 2 J/m^2 is ok. But you have to increase the element stiffness.
For
the bi-linear elements it is very easy to check if the cohesive
properties are good. Usually you have two physical values: the strength
"s" and energy release rate "G". Then, for bi-linear traction law
the element maximum opening "delta" is equal 2G/s. The area under you
curve should be equal G. And element stiffness you usually can choose
arbitrary, but the curve should have good shape. In the image the first
curve is Ok (usually even higher stiffness is used). Second figure is
soft element. In the last figure the element is too soft, you shouldn't
use such traction law. But in your file you have traction law like in
the third figure. For the exponential damage evolution law the
calculations are not so easy, but in general the situation is the same.
Hello Mr. Sergej,
Hello Mr. Sergej,
Thank you once again for the help . Do you mean to say that I must use linear traction law option instead of exponential traction law? Thank you
Cohesive stiffness
Hello!
No, you can use exponential law. But to
choose "good" element stiffness you in any case have to check if your
element is not too soft. And for this linear law is more simple. For
your strength and energy release rate the stiffness should be not less
than 1e15, otherwise you will get traction law like in Fig. 3.
cohesive stiffness
Hello, Mr. sergej,
I did as you have suggested to me to give higher values for the G to make the cohesive element stiffer and ran the program. I am still not getting the elements to erode instead they are being distorted. I cannot attach the input file here I will attach the input file separately. Thank you.
cohesive stiffness
Mr. Sergej,
I have attached the input file as cohesiveuniform-mesh4.doc in my initial problem statement but deleted the first input file. Thank you.
Cantilever beam
Hi!
I don't know what are you trying to model,
but it is not cantilever beam. For cantilever beam your model is
overconstrained (what is the meaning of XSYMM condition in your
model?). Second, it is better to use displacement loading, because in
force loading you may get unstable crack propagation. And third, why
don't you start with simple classical DCB test with 2D model and
isotropic material?
Cantilever beam
Hello Mr. Sergej,
Thank you for the help. I was working on plate problem and trying to do cantilever beam first but mistakely I assigned the XSYMM to the beam too. Anyway thank you my prgram seems to be working. I am getting the elements to erode.
Dear Sergei
Dear Sergei Tarasov,
thak you for sharing your knowledge. yould you please mention whether it is (" the stffness should be at least 1e15 ") the initial stiffness (Ko) of the traction separation cueve ?
What is the unit of this stiffnesss ( 1e15 ) you ve mentioned here?
Thanking You
SAJIB
Material properties
Hello,
If any one has done any FEA analysis for SiC/SiC composite with cohesive element layer in between, Please do sent me the material properties. Thank you.