Quantcast

Ballistic Impact

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Ballistic Impact

Nick Dutton

Hi All,

   I am trying to determine the nodal displacements of a cylindrical projectile after impacting a hard target in Abaqus 6.8-1.  I am thinking I will need to use a UMAT (which I have never done) for an equation I am using that better predicts stress at high strain rates than the Johnson-Cook model.  Currently I am trying to get this to run just once so I can see the output file before I start optimizing my code.

  Can someone please give me advice on the following?  What I am doing wrong, how and where do I implement the UMAT (I know C++ and Python), etc..

***
  I am modeling an axisymmetric deformable projectile, and using an axisymmetric analytical rigid wire for the impact surface.

***
  For the material, I have under "material behaviors":

Deformation Plasticity:
  I have values for young's modulus, poisson's ratio, and yield stress.  I am unsure what I need to put for exponent and the yield offset input boxes, or if I even need to define "deformation plasticity."

Plastic:
  I am currently trying to use Isotropic hardening (should I be using Johnson-Cook?).  I am uncertain as to what values need to go in the Yield Stress/Plastic Strain input boxes.  I have quasi-static compression test data for the material.  Should I be fill these input boxes with that data?  What Yield stress value corresponds to the 0 plastic strain?

  There is a suboption menu with "Rate Dependent" in which you can select Johnson-Cook hardening.  I currently do not have it selected.

Elastic:
  I have isotropic type selected and Young's modulus and Poisson's ratio in the input boxes.


***
  I defined a section for the projectile of type solid, homogeneous using the defined material and it's properties.

***
  Under Assembly I used an Independent instance type. When I created the parts, I made it so that one end of the projectile (the impact end) was at the same location as the anvil (rigid wire).

***
  I currently have 2 steps:
Initial:
  I defined a reference point on the analytical rigid wire at it's center, and put an ENCASTRE BC there. Do I need to create a contact surface?

Step-1:
  I chose dynmamic explicit with all default options.  What should I change to obtain better results?  Since the Initial step wouldn't let me define an axial velocity I did it on this step.  I chose the entire projectile part to give this velocity to.  Should I give the velocity to each node on the projectile mesh?  If so, how would I do this?

***
  For the mesh, I am currently using a Tri element shape with "Free" selected under Technique (should I select "Use mapped meshing where appropriate" under "Algorithm?").

  Under Element type I am currently using "Explicit" under "Element Library" and "Quadratic" under "Geometric Order", and all the default under "Element Controls". I also have selected "Axisymmetric Stress" for "Family".

I realize that there are a lot of questions.  Thank you so much for taking the time to read all of this.

Sincerely,
  Nick
   

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Ballistic Impact

Nick Dutton
I removed Deformation Plasticity and created a contact surface.  I am getting the following errors:

INCREMENT     1 STARTS. ATTEMPT NUMBER  3, TIME INCREMENT  6.250E-02

 ***WARNING: THERE ARE 2 UNCONNECTED REGIONS IN THE MODEL.
   CONTACT PAIR (ASSEMBLY_PROJECTILE-1_PROJ_END,ASSEMBLY_ANVIL-1_SURF-1) NODE
   PROJECTILE-1.34 IS OVERCLOSED BY 16.8125 WHICH IS TOO SEVERE.

 ***WARNING: THE STRAIN INCREMENT HAS EXCEEDED FIFTY TIMES THE STRAIN TO CAUSE
             FIRST YIELD AT 74 POINTS

 ***WARNING: THE STRAIN INCREMENT IS SO LARGE THAT THE PROGRAM WILL NOT ATTEMPT
             THE PLASTICITY CALCULATION AT 74 POINTS
 

 ***NOTE: MATERIAL CALCULATIONS FAILED TO CONVERGE OR WERE NOT ATTEMPTED AT ONE
          OR MORE POINTS. CONVERGENCE IS JUDGED UNLIKELY.
 

 ***NOTE: SEVERE CONTACT OVERCLOSURES EXIST. CONVERGENCE IS JUDGED UNLIKELY.

 ***WARNING: CONVERGENCE JUDGED UNLIKELY.  INCREMENT WILL BE ATTEMPTED AGAIN
             WITH A TIME INCREMENT OF 1.56250E-02


Can anyone give me some guidance?

Thanks!

--- In [hidden email], "Nick Dutton" <greenlran@...> wrote:

>
>
> Hi All,
>
>    I am trying to determine the nodal displacements of a cylindrical projectile after impacting a hard target in Abaqus 6.8-1.  I am thinking I will need to use a UMAT (which I have never done) for an equation I am using that better predicts stress at high strain rates than the Johnson-Cook model.  Currently I am trying to get this to run just once so I can see the output file before I start optimizing my code.
>
>   Can someone please give me advice on the following?  What I am doing wrong, how and where do I implement the UMAT (I know C++ and Python), etc..
>
> ***
>   I am modeling an axisymmetric deformable projectile, and using an axisymmetric analytical rigid wire for the impact surface.
>
> ***
>   For the material, I have under "material behaviors":
>
> Deformation Plasticity:
>   I have values for young's modulus, poisson's ratio, and yield stress.  I am unsure what I need to put for exponent and the yield offset input boxes, or if I even need to define "deformation plasticity."
>
> Plastic:
>   I am currently trying to use Isotropic hardening (should I be using Johnson-Cook?).  I am uncertain as to what values need to go in the Yield Stress/Plastic Strain input boxes.  I have quasi-static compression test data for the material.  Should I be fill these input boxes with that data?  What Yield stress value corresponds to the 0 plastic strain?
>
>   There is a suboption menu with "Rate Dependent" in which you can select Johnson-Cook hardening.  I currently do not have it selected.
>
> Elastic:
>   I have isotropic type selected and Young's modulus and Poisson's ratio in the input boxes.
>
>
> ***
>   I defined a section for the projectile of type solid, homogeneous using the defined material and it's properties.
>
> ***
>   Under Assembly I used an Independent instance type. When I created the parts, I made it so that one end of the projectile (the impact end) was at the same location as the anvil (rigid wire).
>
> ***
>   I currently have 2 steps:
> Initial:
>   I defined a reference point on the analytical rigid wire at it's center, and put an ENCASTRE BC there. Do I need to create a contact surface?
>
> Step-1:
>   I chose dynmamic explicit with all default options.  What should I change to obtain better results?  Since the Initial step wouldn't let me define an axial velocity I did it on this step.  I chose the entire projectile part to give this velocity to.  Should I give the velocity to each node on the projectile mesh?  If so, how would I do this?
>
> ***
>   For the mesh, I am currently using a Tri element shape with "Free" selected under Technique (should I select "Use mapped meshing where appropriate" under "Algorithm?").
>
>   Under Element type I am currently using "Explicit" under "Element Library" and "Quadratic" under "Geometric Order", and all the default under "Element Controls". I also have selected "Axisymmetric Stress" for "Family".
>
> I realize that there are a lot of questions.  Thank you so much for taking the time to read all of this.
>
> Sincerely,
>   Nick
>


Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Re: Ballistic Impact

Uday komaragiri
Material: JC will be better for high impact analysis.. but for the
sake of your first attempt, Isotropic hardening plasticity is just
fine. You can specify different hardening behavior for different
strain rates.

Steps: Choosing the defaults for Explicit is OK. However, select the
right TIME PERIOD for your analysis. It dafaults to 1 otherwise.

BCs: You need to specify initial velocity in the initial step, NOT in
the explicit step. If you are working in CAE, go to initial conditions
and specify velocity of the projectile in the initial step. You can
select the whole region to specify the initial velocity. The
velocities in the Explicit step will then be COMPUTED by the code.

Contact: Since you are working with Axisymmetric parts, you cannot use
'general contact', so you need to define the contact pairs by
selecting the surfaces that will contact (i..e outer surfaces of
projectile and the impact surface).

Finally, there is an example problem in the manual on projectile
impact.. I think it includes 'material damage' as well... you might
want to take a look..

Regards,
Uday




On Sun, Jul 12, 2009 at 4:29 PM, Nick Dutton<[hidden email]> wrote:

> I removed Deformation Plasticity and created a contact surface.  I am getting the following errors:
>
> INCREMENT     1 STARTS. ATTEMPT NUMBER  3, TIME INCREMENT  6.250E-02
>
>  ***WARNING: THERE ARE 2 UNCONNECTED REGIONS IN THE MODEL.
>   CONTACT PAIR (ASSEMBLY_PROJECTILE-1_PROJ_END,ASSEMBLY_ANVIL-1_SURF-1) NODE
>   PROJECTILE-1.34 IS OVERCLOSED BY 16.8125 WHICH IS TOO SEVERE.
>
>  ***WARNING: THE STRAIN INCREMENT HAS EXCEEDED FIFTY TIMES THE STRAIN TO CAUSE
>             FIRST YIELD AT 74 POINTS
>
>  ***WARNING: THE STRAIN INCREMENT IS SO LARGE THAT THE PROGRAM WILL NOT ATTEMPT
>             THE PLASTICITY CALCULATION AT 74 POINTS
>
>
>  ***NOTE: MATERIAL CALCULATIONS FAILED TO CONVERGE OR WERE NOT ATTEMPTED AT ONE
>          OR MORE POINTS. CONVERGENCE IS JUDGED UNLIKELY.
>
>
>  ***NOTE: SEVERE CONTACT OVERCLOSURES EXIST. CONVERGENCE IS JUDGED UNLIKELY.
>
>  ***WARNING: CONVERGENCE JUDGED UNLIKELY.  INCREMENT WILL BE ATTEMPTED AGAIN
>             WITH A TIME INCREMENT OF 1.56250E-02
>
>
> Can anyone give me some guidance?
>
> Thanks!
>
> --- In [hidden email], "Nick Dutton" <greenlran@...> wrote:
>>
>>
>> Hi All,
>>
>>    I am trying to determine the nodal displacements of a cylindrical projectile after impacting a hard target in Abaqus 6.8-1.  I am thinking I will need to use a UMAT (which I have never done) for an equation I am using that better predicts stress at high strain rates than the Johnson-Cook model.  Currently I am trying to get this to run just once so I can see the output file before I start optimizing my code.
>>
>>   Can someone please give me advice on the following?  What I am doing wrong, how and where do I implement the UMAT (I know C++ and Python), etc..
>>
>> ***
>>   I am modeling an axisymmetric deformable projectile, and using an axisymmetric analytical rigid wire for the impact surface.
>>
>> ***
>>   For the material, I have under "material behaviors":
>>
>> Deformation Plasticity:
>>   I have values for young's modulus, poisson's ratio, and yield stress.  I am unsure what I need to put for exponent and the yield offset input boxes, or if I even need to define "deformation plasticity."
>>
>> Plastic:
>>   I am currently trying to use Isotropic hardening (should I be using Johnson-Cook?).  I am uncertain as to what values need to go in the Yield Stress/Plastic Strain input boxes.  I have quasi-static compression test data for the material.  Should I be fill these input boxes with that data?  What Yield stress value corresponds to the 0 plastic strain?
>>
>>   There is a suboption menu with "Rate Dependent" in which you can select Johnson-Cook hardening.  I currently do not have it selected.
>>
>> Elastic:
>>   I have isotropic type selected and Young's modulus and Poisson's ratio in the input boxes.
>>
>>
>> ***
>>   I defined a section for the projectile of type solid, homogeneous using the defined material and it's properties.
>>
>> ***
>>   Under Assembly I used an Independent instance type. When I created the parts, I made it so that one end of the projectile (the impact end) was at the same location as the anvil (rigid wire).
>>
>> ***
>>   I currently have 2 steps:
>> Initial:
>>   I defined a reference point on the analytical rigid wire at it's center, and put an ENCASTRE BC there. Do I need to create a contact surface?
>>
>> Step-1:
>>   I chose dynmamic explicit with all default options.  What should I change to obtain better results?  Since the Initial step wouldn't let me define an axial velocity I did it on this step.  I chose the entire projectile part to give this velocity to.  Should I give the velocity to each node on the projectile mesh?  If so, how would I do this?
>>
>> ***
>>   For the mesh, I am currently using a Tri element shape with "Free" selected under Technique (should I select "Use mapped meshing where appropriate" under "Algorithm?").
>>
>>   Under Element type I am currently using "Explicit" under "Element Library" and "Quadratic" under "Geometric Order", and all the default under "Element Controls". I also have selected "Axisymmetric Stress" for "Family".
>>
>> I realize that there are a lot of questions.  Thank you so much for taking the time to read all of this.
>>
>> Sincerely,
>>   Nick
>>
>
>
>
>
> ------------------------------------
>
> Community email addresses:
>  Post message: [hidden email]
>  Subscribe:    [hidden email]
>  Unsubscribe:  [hidden email]
>  List owner:   [hidden email]
>
> Shortcut URL to this page:
>  http://groups.yahoo.com/group/AbaqusYahoo! Groups Links
>
>
>
>
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Re: Ballistic Impact

Dave Lindeman
In reply to this post by Nick Dutton
Two possibilities:

1)  Your surface normals are reversed.

2)  Significant penetration is being encountered during a single time
step.  If, in fact, you are simulating ballistic impact, then your
impact speeds are relatively large, and you should probably consider
using Explicit.  If, for some reason, you need to use Standard, you'll
need to use a much smaller time step, and/or allow ABAQUS to cut back to
a much smaller time step.

Regards,

Dave

-------------------------
Dave Lindeman
Lead Research Specialist
3M Company
3M Center 235-3F-08
St. Paul, MN 55144
651-733-6383


Nick Dutton wrote:

>
>
>
> I removed Deformation Plasticity and created a contact surface. I am
> getting the following errors:
>
> INCREMENT 1 STARTS. ATTEMPT NUMBER 3, TIME INCREMENT 6.250E-02
>
> ***WARNING: THERE ARE 2 UNCONNECTED REGIONS IN THE MODEL.
> CONTACT PAIR (ASSEMBLY_PROJECTILE-1_PROJ_END,ASSEMBLY_ANVIL-1_SURF-1) NODE
> PROJECTILE-1.34 IS OVERCLOSED BY 16.8125 WHICH IS TOO SEVERE.
>
> ***WARNING: THE STRAIN INCREMENT HAS EXCEEDED FIFTY TIMES THE STRAIN TO
> CAUSE
> FIRST YIELD AT 74 POINTS
>
> ***WARNING: THE STRAIN INCREMENT IS SO LARGE THAT THE PROGRAM WILL NOT
> ATTEMPT
> THE PLASTICITY CALCULATION AT 74 POINTS
>
>
> ***NOTE: MATERIAL CALCULATIONS FAILED TO CONVERGE OR WERE NOT ATTEMPTED
> AT ONE
> OR MORE POINTS. CONVERGENCE IS JUDGED UNLIKELY.
>
>
> ***NOTE: SEVERE CONTACT OVERCLOSURES EXIST. CONVERGENCE IS JUDGED UNLIKELY.
>
> ***WARNING: CONVERGENCE JUDGED UNLIKELY. INCREMENT WILL BE ATTEMPTED AGAIN
> WITH A TIME INCREMENT OF 1.56250E-02
>
> Can anyone give me some guidance?
>
> Thanks!
>
> --- In [hidden email] <mailto:Abaqus%40yahoogroups.com>, "Nick
> Dutton" <greenlran@...> wrote:
>  >
>  >
>  > Hi All,
>  >
>  > I am trying to determine the nodal displacements of a cylindrical
> projectile after impacting a hard target in Abaqus 6.8-1. I am thinking
> I will need to use a UMAT (which I have never done) for an equation I am
> using that better predicts stress at high strain rates than the
> Johnson-Cook model. Currently I am trying to get this to run just once
> so I can see the output file before I start optimizing my code.
>  >
>  > Can someone please give me advice on the following? What I am doing
> wrong, how and where do I implement the UMAT (I know C++ and Python), etc..
>  >
>  > ***
>  > I am modeling an axisymmetric deformable projectile, and using an
> axisymmetric analytical rigid wire for the impact surface.
>  >
>  > ***
>  > For the material, I have under "material behaviors":
>  >
>  > Deformation Plasticity:
>  > I have values for young's modulus, poisson's ratio, and yield stress.
> I am unsure what I need to put for exponent and the yield offset input
> boxes, or if I even need to define "deformation plasticity."
>  >
>  > Plastic:
>  > I am currently trying to use Isotropic hardening (should I be using
> Johnson-Cook?). I am uncertain as to what values need to go in the Yield
> Stress/Plastic Strain input boxes. I have quasi-static compression test
> data for the material. Should I be fill these input boxes with that
> data? What Yield stress value corresponds to the 0 plastic strain?
>  >
>  > There is a suboption menu with "Rate Dependent" in which you can
> select Johnson-Cook hardening. I currently do not have it selected.
>  >
>  > Elastic:
>  > I have isotropic type selected and Young's modulus and Poisson's
> ratio in the input boxes.
>  >
>  >
>  > ***
>  > I defined a section for the projectile of type solid, homogeneous
> using the defined material and it's properties.
>  >
>  > ***
>  > Under Assembly I used an Independent instance type. When I created
> the parts, I made it so that one end of the projectile (the impact end)
> was at the same location as the anvil (rigid wire).
>  >
>  > ***
>  > I currently have 2 steps:
>  > Initial:
>  > I defined a reference point on the analytical rigid wire at it's
> center, and put an ENCASTRE BC there. Do I need to create a contact surface?
>  >
>  > Step-1:
>  > I chose dynmamic explicit with all default options. What should I
> change to obtain better results? Since the Initial step wouldn't let me
> define an axial velocity I did it on this step. I chose the entire
> projectile part to give this velocity to. Should I give the velocity to
> each node on the projectile mesh? If so, how would I do this?
>  >
>  > ***
>  > For the mesh, I am currently using a Tri element shape with "Free"
> selected under Technique (should I select "Use mapped meshing where
> appropriate" under "Algorithm?").
>  >
>  > Under Element type I am currently using "Explicit" under "Element
> Library" and "Quadratic" under "Geometric Order", and all the default
> under "Element Controls". I also have selected "Axisymmetric Stress" for
> "Family".
>  >
>  > I realize that there are a lot of questions. Thank you so much for
> taking the time to read all of this.
>  >
>  > Sincerely,
>  > Nick
>  >
>
>
Loading...