User login


You are here

Surface-based cohesive behavior in abaqus 6.9-1

icetree's picture


   Can anyone explain the difference between  Surface-based cohesive behavior and cohesive element?And to the  Surface-based cohesive behavior,how to define the every parameters in it?if possible,please upload an example with the  Surface-based cohesive behavior method.

     please take  some valuable suggestions。



rajnish0007's picture

Hi As, I have used cohesive surfaces for debonding problem, I can say that the behavior of the cohesive surfaces is same as the cohesive elements, also you can model frictional behavior of debonded surfaces as in addition to cohesive forces. The application part is much simple if you are using abaqus 6.9., in that you can have cohesive surface interaction in contact of interaction module. In addition to this you have to define the damage parameters as you define for cohesive elements. For the further details you can refer the example of Alfano for debonding.  


Dr. Yaqoob Yasin and Rajnish Sharma

I have been trying to create an explicit model with a sticky interface using the cohesive surface behavior available in 6.9 but I keep running into the same error message.

"At least one non-blank surface must be defined for contact interactions referencing cohesive behavior under *CONTACT PROPERTY ASSIGNMENT (see the surface interaction named INTPROP)"

I even get this error when I import the input file "gcont_cohesive_options.inp" from the verification manual and select anything but default for the global property assignment under Interactions.

Any idea what is meant by a "non-blank surface"?

Thanks for your time,

tuhinsinha.25's picture


I am not sure if you have fixed the problem but I encountered the same thing in my general contact + cohesion model and here is how I rectified it.

1) Make two interaction properties, one containing the cohesion and damage behavior and other the rest of the contact behavior (normal, tangential, etc.)

2) Define general contact with global property assignments using the non-cohesive interaction behavior.

3) Create an individual property assignment where you select the surface pairs and use the cohesive interaction model here.

Continue your analysis in the regular way. It worked for me.




Thank you

Aaron, did you ever find the solution to your problem?

 I am currently having the exact problem that you described, even for the verification input file.


Kind regards, Dan 

Hi Rajnish,

I'm using surface-based cohesive behavior to predict delamination in compoistes. In my model, two surfeces with idential meshes contact with each other, but I found the delamination areas given by these two surfaces were different. Do you have any idea on why this occurs? thanks!



Hi Rajnish,

I am interested to learn about friction based simulation between Concrete and FRP. I am modeling a FRP strengthened RC beam. FRP is in tension face.

I have simulated the behavior using surface based simulation and got good results with delamination. As there is considerations (Knn) in surface based model for constraining normal behavior between FRP and concrete  there is no need for constraining the FRP and concerte in normal direction. But what should i do in case of Friction based analysis instead of cohesive analysis?

I have used the following keywords:

*CONTACT PAIR,INTERACTION=Friction,adjust=0.5,small sliding,TYPE=node TO SURFACE
*surface interaction,name=friction
*friction,TAUMAX=4.11,ELASTIC SLIP=0.005
*SURFACE BEHAVIOR,pressure-overclosure=HARD

but I didnt apply any constrain between FRP top and concrete bottom. Do i need to constrain ??

using friction based simulation i dont find any effectiveness of FRP. The load carrying capacity of beam should be greater than control beam. but i dont get it using the friction based simulation. can you help me at this poin?? I am stuck here for a long time.

Thanking You

Hello sir @ rajnish, I have modelled interface debonding using cohesive elements and cohesive surface behavior. Results from cohesive elements model are reliable. But when I use cohesive surface behavior, results are highly deviating. I have defined cohesive surface interaction properties as below. Knn=170MPa, Kss=Ktt= 84MPa. Damage intiation = Maxs criterion, Normal = 3MPa, shear 1= shear2 = 1.32MPa. Damage evolution is given using energy and linear softening behavior. Do I need to define anything more in this interaction property?? I am not understanding where I am commiting mistake. Your suggestions would be really helpful sir.




Hi Rajnish,

for modeling of bond-slip in a RC member I have used interaction model with cohesive behavior/Explicit. Therefore I have assign it in to surface to surface, but I got the message that I can not use these type of contact. Could you please help in this matter and tell me why I cannot do it?



rajnish0007's picture

Hello Hossein

 you can use the following key words to define the cohesive surface interaction:

*Surface Interaction, name=IntProp-1
Behavior, pressure-overclosure=HARD
*Cohesive Behavior
Initiation, criterion=QUADS
5., 2.5, 2.5
*Damage Evolution, type=ENERGY, mixed mode
behavior=POWER LAW, power=2.
*Damage Stabilization

*Surface Interaction, name=IntProp-2

** Interaction: Int-1
*Contact Inclusions
msurface , ssurface
*Contact property
, , IntProp-2
msurface , ssurface, IntProp-1
formulation, type=PURE MASTER-SLAVE
msurface , ssurface, MASTER


hello rajnish

when I use the cohesive elements to simulate the debonding of composite slab. I did the normal debond experiments and shear debond experiments, in order to get the Knn, Ktt, Kss, and other damage parameters along the normal and tangential directions.

I used the traction separation to mode cohesive section in abaqus.

But, I don't understand what the "mixed mode behavior" means, do you have some documents about this.

thanks very much.

hello rajnish

when I use the cohesive elements to simulate the debonding of composite slab. I did the normal debond experiments and shear debond experiments, in order to get the Knn, Ktt, Kss, and other damage parameters along the normal and tangential directions.

I used the traction separation to mode cohesive section in abaqus.

But, I don't understand what the "mixed mode behavior" means, do you have some documents about this.

thanks very much.

I have a question regarding surface-based cohesive behavior , I hope you don't mind me asking.
I am modeling the interface between components of a wellbore using surface-based cohesive behavior. I must assume that all the components of the wellbore are initially in contact and fully bonded in the first place. I am surprised by the results , the elements near contact went plastic but CSDMG is still zero . I think it means the debonding hasn't happened. How can it happen that plasticity happened which means cracks created but debonding is not happening.
I saw in the manual two lines which make me doubt the whole process of using cohesive behaviour in my model.
In the section 37.1.10 Surface-based cohesive behavior in Abaqus Analysis User's Guide under Interaction of traction-separation behaviour with compressive and friction behaviour, I saw these lines :
the cohesive behavior contributes to the contact normal stress only when a slave node is “open” (i.e., not in contact). In the case of “sticky” cohesive behavior—where the two surfaces are not initially in contact—cohesive effects are activated in the increment after the slave node status changes from open to closed.
What does it mean? Does it mean that the nodes should not be in contact at all in the first step? because all my slave nodes are in contact with master nodes . surfaces are fully bonded in the initial step.
Does it mean that de-bonding in the normal direction to the contact won't happen if I use cohesive behaviour? which is the direction I am hoping the debonding happens :( 


 I am working on cohesive zone model using abaqus inbuilt cohesive element. I am doing 3 element test for validating the traction separation law. cohesive element is sandwitched  between two 3d elements. Bottom element is fixed and displacement boundary condition is applied to the top element. I have some questions. Plz gime me youe valuable suggessitions

1. How to select the cohesive material parameters (stiffness, damage criterion, evolution law)

2. I am plotting Rf versus U graph but only linear part of t-delta curve is coming. I am not getting the softening part. why?

3. Is there any modeling issues or problem with cohesive material properties?

4. After simulation what should be the plot in abaqus to match global force deformation curve obtained from experiments?

5. If we unload the sample the traction separation curve should come to origin but how to do that?


Waiting for your positive reply 

rajnish0007's picture

 Hello chintan

 As per my best knowledge In abaqus element manual
"cohesive element" is described, there they have discussed the traction
separation and softening behavior. you can follow the same for element


Dr. Yaqoob Yasin and Rajnish sharma

I am trying to use cohesive material to simulate cutting.

When I am applying the cutting tool in the speciman ,  Instead of cutting it is just going in without creating any damage to the speciman.

Can you help me out with that?

Hi chintan,

Ktt is a constant relating the traction to the separation of the top and bottom surface of the interface. The unit of Ktt is N/m^3. It is not as same as stress's unit. Turon et al. gave a detailed discussion on how to determine the stiffness, Ktt etc.

see Turon A, Davila CG, Camanho PP,
Costa J. An engieering solution for mesh size effects in the simulation of
delamination using cohesive zone models. Engineering fracture mechanics


I am wondering if you know the default stiffness value in ABAQUS surface based cohesive behavior? Many thanks!


i wanto model composite slab, which is concrete and steel deck, the interface between the concrete and steel deck will be model using surface based cohesive, however when i try to run there is a problem in the interaction part, do anyone have simple CAE file the use surface based cohesive or any tutorial on how to model using surface based cohesive.

Please  help.

I'm newborn in ABAQUS. I want to make a static analysis of RC frame with the
bondslip effect after apply cyclic loading to the frame.

At the moment,I tied the reinforcement bars with the concrete. And manage to
plot hysteress curve of the frame. what I'm aspect was degradation of stiffness
and perhaps I can see pinching behaviour in the hsteresis curve which is not
happen to the current plot.
Detail of my model;
1.2D beam model (beam element for beams and column, truss element for
reinforcement bars)
2.Concrete damage plasticity-concrete model
3. Have inlude damage material properties to concrete and steel.

Hope to get help from you all.


Hi there! I am really new to Abaqus and I need to model a composite laminated crash cone under impact to a flat solid wall (analytical surface is used). I modelled a cone with a layers of conventional shell elements with hashin failure criteria and surface-based cohesive behaviour as interaction between layers representing the interface of composite laminates. The model runs successfully but once I checked the energy balance, the damage dissipation energy from the cohesive, it turned out to be NEGATIVE (I think this is wrong).
I tried to model a simple geometry model with cohesive elements and the damage energy from cohesive element is POSITIVE (which seems to be correct).

As far as I know the energy balance: internal energy = strain energy + artificial energy + damage dissipation energy + penalty contact energy +contraint penalty energy + plastic energy

where Total energy = internal energy + viscous energy + kinetic energy + frictional energy + external work = constant
I set up another history field output request to capture the energy of hashin damage to compare with the all damage energy of model and the result is that damage from surface-based cohesive is NEGATIVE. This means the more damage to the cohesive, the less internal energy in the cone which means that less energy absorbed from the impact. <-- totally wrong...

Due to the complex geometry of the model, cohesive element does not seem to the one I can apply to my model.
Could anyone explain or give me a solution why the damage energy from surface-based cohesive is NEGATIVE? Is it wrong?

Here is the interaction property of the surface-based cohesive

*Surface Interaction, name="interface prop"
*Cohesive Behavior, eligibility=ORIGINAL CONTACTS
*Damage Initiation, criterion=QUADS
*Damage Evolution, type=ENERGY, softening=EXPONENTIAL, mixed mode behavior=POWER LAW, power=2.
*Contact Damping, definition=CRITICAL DAMPING FRACTION

hi every body
i try to do the simulation of impact on glare structure (glare :glass aluminium reinforced),and to study the growth of  delamination between the aluminium layer and the glass fiber, i have done a cohesive layer with 0.0001 mm thikness,my problem tat i can't find the properties of the cohesive zone(for damage initiation and damage evolution),if some one can help
thanks a lot

Hi I want to model FRP debonding based on Lu bilinear model.

I read in an article that I can model it in ABAQUS by cohesive element, now I have some questions:

1- In definition of elastic parameter should I use Traction type or Coupled Traction?

2-To define debonding initiation should I use MAXS DAMAGE orQuads Damage?

3- How should I show to software that in the normal direction debonding not occur due to pressure?

4- To show completion of debonding in 3 directions (1 normal and to tangential direction) I want to define maximum energy in every direction, but by clicking on the Suboption key I see that there is only one energy that can be defined. how can I define energy in 3 directions?

5-What kind of element I should use for cohesive element? 3D stress or Cohesive?

6- What is the meaning of this error and how can I fix it ?


(I try static,Riks step with Maximum number of increment, initial, minimum and maximum, 10000,0.01,1E-010 ,1E+036,respectively to fix it but after 40 min the error was repeated!)

Hi Everyone

 I am currently working on the project of  structural health monitoring of sandwich composite beam and I am not able to find how to model the disbond model of sandwich composite beam. I just know that cohesive element and cohesive behaviour is helpful to make the debonding of skin of sandwich beam from core but I dont know how to make model and how to insert cohesive elements for debonding. please help me if anyone have an idea about that



Consult your lawyer before signing the contract.

Subscribe to Comments for "Surface-based cohesive behavior in abaqus 6.9-1"

Recent comments

More comments


Subscribe to Syndicate