iMechanica - delamination - Comments

Low Velocity Impact on Laminated Composite

praveen.comp — Thu, 07 Jul 2011 03:45:52 -0400

Hi Ramdas Sir,

I am doing my M.Tech project and I'm stuck in modelling the impact.

Could you please help me out?

Since it is a low velocity impact, I want to model delamination, but while modeling composite layup, all i see is a planar surface. When i tried to model it as solid layup, then i couldn't assing the region to the plies.

Even if my composite layup is right, but then i'm not able to simulate the impact.

I gave velocity constraints to the impactor and predefined velocity in da direction normal to the plate toward the plate. But still it is not affecting the plate.

I want to model the impactor as a rigid body.

Please help me out, sir...

I would be highly obliged..

Can you send

Hitesh1811 — Thu, 17 Mar 2011 09:57:30 -0400

Hi Shi,

I am interesting in computational mechanics. Can you send me your code?I want to saw that how this build and how implement this code in another software.

can your program is useful

svsrsomaraju — Wed, 09 Feb 2011 06:21:33 -0500

can your program is useful in ansys for generating interface elements.Since i am facing problem while modelling cohesive zone.

Hi Shi, Could you send me

vinh phu nguyen — Thu, 06 Jan 2011 09:59:23 -0500

Hi Shi,

Could you send me your email? I couldn;t find it in the website you told me.

Phu

Hi Shi, Although my code

vinh phu nguyen — Thu, 06 Jan 2011 09:56:52 -0500

Hi Shi,

Although my code currently does not support Abaqus input, it has been implemented in such a way that new input format can be incorporated easily.

Ok, I am sending you that piece of code.

Phu

Phu, I have a related

Jianxu Shi — Wed, 05 Jan 2011 18:42:21 -0500

Phu,

I have a related question to Abaqus: can you code work on Abaqus input file: aka, input=abaqus inp file without interface elements, output=new abaqus input file with interface elements block inserted in a specified solid region.

PS: I am interested in obtaining a copy of your program. we can talk about future collaboration. For more info about myself, see gem-consultant.com (quite dated but you can email me for updated info).

Best,

J.Shi

Hi Hu, My code implements

vinh phu nguyen — Sun, 02 Jan 2011 19:50:49 -0500

Hi Hu,

My code implements the standard interface elements within the framework of cohesive zone models. I do not know the Abaqus implementation so I cannot describe the difference.

Happy new year,

Phu

Hi Phu, Abaqus provide a

D Y Hu — Thu, 30 Dec 2010 02:35:54 -0500

Hi Phu,

Abaqus provide a kind of element name cohesive, which can simulate interface between solid and solid. Can you describe your program how to solve the crack progression. I am also interest in the differences between your program and abaqus cohesive element.

Thanks!

Hi Lianhua, My interface

vinh phu nguyen — Thu, 23 Dec 2010 20:57:30 -0500

Hi Lianhua,

My interface element model is able to deal with solid-solid interfacial failure. My interface elements generator howver just reads a FE mesh and insert interface elements where necesarry. It does not know about solid or shell elements.

I leave your question on failure model for solid/shell interface to experts on this field. Sorry for not being so helpful.

Phu

Hi Phu, I am sorry I did

Lianhua Ma — Thu, 23 Dec 2010 09:46:44 -0500

Hi Phu, I am sorry I did not give more details about my question. I’m trying
to make it clear in this post.

In the commercial FE software(ANSYS or ABAQUS), as we know, most composite models (matrix
and inclusion) are generally constructed using(2D or 3D) solid elements for all parts. In such cases, we need to generate interface elements as you mentioned when modeling material failure behavior. I think this can be easily implemented in ABAQUS.

Now let’s consider a special kind of composite, carbon nanotube(CNT)-reinforced
composite, to simplify we can treat CNT as a shell model(which has already developed
in some literatures). In this case, the matrix and CNT can be modeled respectively
by solid and shell elements in ABAQUS. I am currently interested in the interfacial
failure issue of this model. Using your program, could you generate corresponding
interface elements for this model? Could you give more information about the
special cohesive interface failure simulation?

Thanks

Lianhua

Hi Lianhua, Could you be

vinh phu nguyen — Thu, 23 Dec 2010 04:00:02 -0500

Hi Lianhua,

Could you be more specific please? I din't get your point, sorry.

Phu

Hi Nguyen, the code you

Lianhua Ma — Wed, 22 Dec 2010 20:58:42 -0500

Hi Nguyen, the code you mention may be convenient to generate interface meshes in open source FE software. I think it is also easy to generate such meshes in ANSYS or ABAQUS.

I have one question for the simulation of material interfacial failure.

Is it possible to model the failure of the interface between the solid and shell model by zero thickness interface element?

Thanks

Lianhua

Hi Arun I'm somewhat

Reuben Govender — Fri, 29 Jan 2010 09:32:39 -0500

Hi Arun

I'm somewhat sceptical about "pure shear" experiments, especially for any laminar or bonded materials. My reasoning is that methods such as the Iosipescu test, or asymmetrical 4 pt shear, do not create pure shear except for an infinitessimally small region of the specimen. If one moves away from the position of pure shear by a small amount, the material is subject to a complex stress state. As I alluded to in my earlier post, lap shear is not "pure shear". Indeed if the line of shear does not align perfectly with the line of application of force, there will be some tensile or compressive stress imposed. Also as the test progresses, the asymmetric application of the forces to each leg cause bending, which superimposes on the shear.

I have been successful in running models of ENF tests (which are nominally pure mode II fracture) in Abaqus. I hesitate to claim successful modelling of shear, as while the global responses (such as force-displacement histories) of the model compare well to experiment, there are still some details on an element scale level that I am not happy about. Also just getting the simulations with cohesive elements to run to completion is a non-trivial exercise. Quasi-static simulations of cohesive elements do not converge very quickly, and often require some tweaking of solver parameters to achieve convergence.

Pure Mode II shear experiment

Arun Krishnan — Thu, 28 Jan 2010 13:39:03 -0500

Hi Reuben,

I came across this post while searching for modeling pure shear using cohesive elements. I am using an Iosipescu specimen to determine shear and fracture properties of bonded polymers. By obtaining fracture properties at the crack tip, I can show that a cracked Iosipescu specimen is under pure Mode II loading.

I am also trying to model the shear failure using cohesive elements, but I have not been too successful. Have you got any answers to your questions on pure shear modeling?

Arun

Hi Venkat, Most of our

Arun Krishnan — Wed, 18 Nov 2009 12:23:22 -0500

Hi Venkat,

Most of our data are on same and bi-material adhesive joints. We use polymers like polycarbonate, PMMA and Homalite in conjunction with metals like aluminum, steel etc. bonded together with adhesives. So our data is on polymer-metal combinations and not on composites.

You should be able to find tensile and shear data for certain popularly used composites in literature. If it is a specially manufactured one, it might benefit you to ask the manufacturer; they might have tested for these properties. Otherwise for you model, use a material with similar properties for which you can find values. Try validation with some experimental work. I generally use this website called engineeringvillage.org for finding references. You could use this or Google scholar depending on availability in IIT KGP.

Good Luck,

Arun