XANSYS Message: 20564 [Go back to message list]
[bookmark on del.icio.us]
No rating yet
Rate item:

Subject: Re: Impact Crack Propagation in ANSYS/LS-DYNA
Author: Jason Husband
Date: 2001-01-27 09:48:00

Hi Rick
DYNA is a well established, mature product with a wealth of experimental
verification. I think your problems are an issue of flipping the right
software switches. I have not used Taurus so can not help you there.
Here are some comments and things you can try.

ANSYS/LS-Dyna does not provide very good postprocessing of strains for
shell elements. You can access only the inner or outer strains with the
LAYER command . I am not sure how/if Powergraphics works with DYNA. Have
you stored stress and strain data for all the layers? How many
integration points through thickness are you using? How often are you
storing /post1 and /post26 results? What is the calculated time step
size compared to the contact step size? If you are accustomed to
stresses derived from fully integrated elements (i.e. implicit defaults)
then you may be horrified at the discontinuous stress field produced by
single point integration elements.

I hope one of these turns out ot be helpful.
Jason

Rick Fischer wrote:

> I'm looking at a little bit of everything: nodal, elemental,
> averaged, no average, Powergraphics on/off, etc. The model is a
> rigid ball dropped on a formed plastic sheet. Sorta complex
> geometry, so I made a simpler one. Both used Belytscho-Leviathan
> shells. Thinking something was happening in the shells, I made a
> cantileaver beam with bricks, got rid of the impact and tip loaded
> the thing, and so far the same results. Hourglass energies are below
> 5%. I'm thinking of trying to look at this in Taurus to see if this
> is a data translation issue, and looking at noise filtering. Any
> other suggestions?

> --- In xansys@y..., Jason Husband wrote:

>> Are you looking at nodal or element solutions.
>> Is this a shell element: I think all layers through thickness must

> reach

>> the failure criterion before element failure occurs.
>> Jason

>> rick.fischer@m... wrote:

>>> I am currently working on a problem where I am trying to use the

> failure option

>>> with the PLAW model. I am getting strange results. I have

> specified a failure

>>> strain of .10 and do not get failures until roughly double that.

> LSTC has

>>> confirmed that this failure value is the equivilent plastic

> strain.

>>> Plnsol,eppl,eqv shows strains way over the failure value. I'm a

> newbie Dyna

>>> user, so I could be doing something wrong, but so far I cant find

> it. My advice

>>> is be careful.

>>> Jason Husband on 01/24/2001 07:13:23 PM

>>> To: xansys
>>> cc: (bcc: Rick Fischer/MAIN/MC1)

>>> Subject: Re: [xansys] Impact Crack Propagation in ANSYS/LS-DYNA

>>> Don
>>> A couple of years ago I had an interesting experience simulating

> the

>>> catastrophic failure of a very large spinning air fan. I used LS-

> DYNA as

>>> a tool to gain insight into how and why the fan exploded. We had

> the fan

>>> pieces in our facility and a lots of physical evidence and

> mechanical

>>> testing to correlate the simulation results to. Technically crack
>>> propagation was the most difficult obstacle into understanding.

> This

>>> tragedy involved loss of multiple human lives and is still in

> litigation

>>> so I cannot divulge many details.

>>> Your problem could be fiendishly difficult, depending on the

> materials

>>> ability to undergo plastic deformation. If you prescribe a failure
>>> criterion with *mat_add_erosion you can make elements fail during

> the

>>> simulation.

>>> I obtained good results for my particular problem with element

> failure

>>> options. By 'good results' I mean the overall response of the

> structure

>>> (number and shape of broken pieces and damage to surrounding

> structure)

>>> and not the detailed fracture surfaces. If you want these details

> I feel

>>> you have your work cut out for you.

>>> Jason Husband
>>> QuEST

>>> Don Ohanehi wrote:

>>>> A colleague needs to run an impact analysis of a simple block

> and include

>>>> the simulation of crack propagation. In my experience with LS-

> DYNA and in

>>>> looking through the manuals, I did not see any evidence of

> capabilities for

>>>> conducting automatic re-meshing for simulating crack propagation.

>>>> I was wondering if any of you has had any experience in

> simulating crack

>>>> propagation in LS-DYNA or ANSYS/LS-DYNA. Does any of you know

> of any

>>>> software that will handle crack propagation for an impact

> problem?

>>>> Please feel free to post a response on the newsgroup, or

> communicate with

>>>> me directly. I will post a summary of responses that I receive

> directly,

>>>> if they seem of general interest to the newsgroup.

>>>> Thanks a million!

>>>> Don Ohanehi

>>>> Don Ohanehi, Ph. D.
>>>> Research Scientist

>>>> OFFICE: 540/231-3141
>>>> LAB : 540/231-7484

>>>> FAX : 540/231-9187
>>>> Virginia Tech
>>>> Engineering Science & Mechanics Dept.
>>>> 121 Patton Hall
>>>> Blacksburg, VA 24061-0219
>>>> E-mail : dohanehi@v...

Posts possibly associated with message #20564AuthorDateScore
20500Impact Crack Propagation in ANSYS/LS-DYNADon Ohanehi2001/01/24 
20503Re: Impact Crack Propagation in ANSYS/LS-DYNAosman buyukisik osman@2001/01/24 
20504Re: Impact Crack Propagation in ANSYS/LS-DYNAJason Husband2001/01/24 
20511Re: Impact Crack Propagation in ANSYS/LS-DYNABogdan Balasa2001/01/25 
20526Re: Impact Crack Propagation in ANSYS/LS-DYNARick Fischer2001/01/25 
20546Re: Impact Crack Propagation in ANSYS/LS-DYNAJason Husband2001/01/25 
20556Re: Impact Crack Propagation in ANSYS/LS-DYNAShen-Yeh Chen2001/01/26 
20563Re: Impact Crack Propagation in ANSYS/LS-DYNARick Fischer rickfischer@2001/01/27 
20564Re: Impact Crack Propagation in ANSYS/LS-DYNAJason Husband2001/01/27