Adhesive modelling

Hello,

returning to a subject I had left for a long time, I sincerely believe that there hasn't been so far a well explained/documented example or tutorial on how to effectively perform an analysis involving glue separation on RADIOSS (whichever or the newest versions). Brick elements will be used to simulate the glue. My latest small research has guided me to the use of material law 59, using an appropriate property such as FAIL_CONNECT. These two can be linked, OK with that. The question is: Do you also need to define a property like P43_CONNECT? Could you please explain? Lastly, how to assign your elements to the desired property when no 'assign' option exists in properties menu (only 'create' and 'update'). The same when I right click the property in the model browser: the 'assign' option is inactive.

Please excuse any ignorance here. This is my first attempt with RADIOSS and I sincerely hope it will not be the last. I have read some very off putting comments on the subject (RADIOSS & separation/delamination).

Thank you.

Find more posts tagged with

English

Radioss

Accepted answers

All comments

Altair Forum User

Hi Mvass,

It's interesting to know that you are exploring RADIOSS. As you mentioned, the material law (/MAT/CONNECT) and failure model (FAIL_CONNECT) are linked.

You have to define P43_CONNECT property and this property may only be used with this material law. You can assign it directly to the components by right clicking on the components in the model browser and assign.

We have an example Example 48 - Solid Spotweld in our Help Menu which uses these parameters. For your reference the deck files are attached along.

Unable to find an attachment - read this blog

mvass

Thank you for taking the time to answer to my question. Yes, I am aware of the example mentioned in your reply and I am currently having a look on it. Should I come across with any problems (which is highly possible), I'll let you know.

thank you.

mvass

Hello again,

before starting the creation of a simple model, I reviewed some of the basic theory for modelling such problems. What I have found on previous notes of mine and briefly presented below, is difficult (to me at least) to link it to the material laws and properties mentioned before. Without trying to be judgemental, or having the intention to cover a huge topic in a few lines, I would simply like to know if any of the following can be entered in RADIOSS or if there is something alternative. I am asking this because so far I cannot obtain a solid answer.

So, as far as I know the following are some of the most well documented and used methods to perform a de-gluing study:

1) Virtual Crack Closure Technique (VCCT): Here, the length of the crack is known as well as the crack tip. Failure occurs when the energy release rate G (can be mode I,II,III or a combination of those) is greater than the fracture toughness (Gc) of the material. VCCT can be used in conjunction with a failure criterion say max. hoop stress and once the failure criterion is meet, separation between the two bodies in contact occurs.

2) Cohesive zone modelling (CZM): In this case, no prior knowledge of the crack size and placement is needed. A layer (can be zero thickness) of interface elements (i.e. cohesive) is 'inserted' between the bodies in contact. CZ elements are defined (their material law if you like) by their traction on the top and bottom faces (t,s) with respect to relative displacement (nu) between these faces. The relation between traction and displacement can be bilinear, exponential, linear-exponential or user defined. The user is required to input a critical displacement (nu_critical), which when reached, the cohesive material starts to fail ('open').

3) Definition of a stress criterion: In this case the glued elements will fail (released) when a stress criterion of the type: (normal stress/normal stress allowable)^m + (shear stress/shear stress allowable)^m > 1. User supplies the allowables and the exponents (most cases the value of two (2) is entered, but this is not always correct...). When this criterion is meet, glue breaking occurs.

Based on the above, is there an entry in MAT law or CONNECT_FAIL related to any of the above? Something alternative?

Thank you in advance for your reply and please excuse me if I have forgotten something, or the brief information given above contain any mistakes or omissions. I do not consider myself an expert.

Altair Forum User

Hi Mvass,

Need a bit more information. Are you using composite materials for this analysis?.

mvass

Hello again.

Yes, I am thinking to use composite materials for a simple example, although what I have described in my previous email about the methods used, can be applied on metallic parts as well (for example two metallic parts joined together with glue, or weld).

For your convenience, assume two plates of composite material 'glued' together with epoxy. Any mat properties will do.

Thank you.

Altair Forum User

Hi Mvass,

Have you checked LAW59 with /FAIL/CONNECT?

mvass

@ Prakash Pagadala:

yes sir, as seen on my first post I have seen /MAT/LAW59, as well as /FAIL/CONNECT. As mentioned before, I cannot relate any of the LAW59 'flags' with the information I know about failure modelling and briefly described in a previous post of mine. Are there any entries in this MAT/LAW59 or in FAIL/CONNECT for providing information about critical opening length? cohesive energy? etc. Is there something alternative to those for the RADIOSS setup? That's what I would like to know.

Thank you.

Altair Forum User

Mvass,

Usually how /FAIL/CONNECT works is when the element internal energy reaches a threshold limit or a criteria limit (I think this is what you are talking about ryt, the cohesive energy) the solid element fails.

mvass

Mvass,

Usually how /FAIL/CONNECT works is when the element internal energy reaches a threshold limit or a criteria limit (I think this is what you are talking about ryt, the cohesive energy) the solid element fails.

The FAIL/CONNECT property, contains three flags for internal energy. Their units in SI are kg/surface area^2. Strain energy release rate (which is 'similar') is measured in energy per unit area squared. Is that kg energy? It does not look like that...

Altair Forum User

Hi Mvass,

I think the unit system is correct. As the description says, Energy/unit surface area which means Energy units (Kg*m^2/S^2)/m^2. This lands to Kg/s^2.

mvass

Hi Mvass,

I think the unit system is correct. As the description says, Energy/unit surface area which means Energy units (Kg*m^2/S^2)/m^2. This lands to Kg/s^2.

Sorry, I' haven't checked this thread for some time. Yes, you're right, I should have revised my physics 'units' class, before answering. So, that should be a good point to start.

Thank you.