Reducing stiffness

[maajidchishti]

Dear all

I am running a model of composite joint to failure. I would like anyone to help me in reducing the stiffness of the model. The model is being very stiff compared to the real structure. I am aware of the fact that FEM models are usually stiffer than the real structure, however my model is too stiff compared to the real structure.

I am using viscosity to lead to convergence and I am using implicit analysis in my simulation. My model also has bolt pre-load, contact and friction in the simulation. I would like anyone to suggest ways to reduce the stiffness of the model. Please give ideas, how can I reduce the stiffness of the model. Thanking you in advance.

Kind regards

Maajid

[maajidchishti]

Further information on the model
I am using Linear solid elements C3D8 and C3D6. The model contains a composite plate joint to an aluminium plate with a titanium countersunk bolt. I have modeled all the components of the joint using solid elements. I have also implemented contact and friction in the model. the joint is fixed at one end and loaded in tension on the other end.
Your help will be much appreciated.
Regards
Maajid

--- In Abaqus@..., "maajidchishti" <maajidchishti@...> wrote:

>
> Dear all
>
> I am running a model of composite joint to failure. I would like anyone to help me in reducing the stiffness of the model. The model is being very stiff compared to the real structure. I am aware of the fact that FEM models are usually stiffer than the real structure, however my model is too stiff compared to the real structure.
>
> I am using viscosity to lead to convergence and I am using implicit analysis in my simulation. My model also has bolt pre-load, contact and friction in the simulation. I would like anyone to suggest ways to reduce the stiffness of the model. Please give ideas, how can I reduce the stiffness of the model. Thanking you in advance.
>
> Kind regards
>
> Maajid
>

[sridharan.venkataramanan]

 

Fully integrated linear hexa (C3D8) elements are definitely stiff in
nature. You should avoid them. Either opt for C3D8R or C3D8I.

Regards,

Sridharan

 

From: Abaqus@... [mailto:Abaqus@...] On Behalf
Of maajidchishti
Sent: Sunday, June 21, 2009 7:41 AM
To: Abaqus@...
Subject: [Abaqus] Re: Reducing stiffness

 






Further information on the model
I am using Linear solid elements C3D8 and C3D6. The model contains a
composite plate joint to an aluminium plate with a titanium countersunk
bolt. I have modeled all the components of the joint using solid
elements. I have also implemented contact and friction in the model. the
joint is fixed at one end and loaded in tension on the other end.
Your help will be much appreciated.
Regards
Maajid

--- In Abaqus@... <mailto:Abaqus%40yahoogroups.com> ,
"maajidchishti" <maajidchishti@...> wrote:
>
> Dear all
>
> I am running a model of composite joint to failure. I would like
anyone to help me in reducing the stiffness of the model. The model is
being very stiff compared to the real structure. I am aware of the fact
that FEM models are usually stiffer than the real structure, however my
model is too stiff compared to the real structure.
>
> I am using viscosity to lead to convergence and I am using implicit
analysis in my simulation. My model also has bolt pre-load, contact and
friction in the simulation. I would like anyone to suggest ways to
reduce the stiffness of the model. Please give ideas, how can I reduce
the stiffness of the model. Thanking you in advance.
>
> Kind regards
>
> Maajid
>





[Non-text portions of this message have been removed]

[srinivas srinivas-2]

haii maajid,

           In my knowledge C3D8 and C3D6 elements are the full integration elements. they are used only while doing the frequency analysis. normally fully integration elements exhibit shear locking so they are the stiff elements, so they should not be used. and reduced integration elements overcome this shear locking however they too have problem called hourglassing. but this can be resolved to the maximum extent using the present softwares. so if u are modelling a composite material using solid elements better to use the reduced integration elements with enhanced hourglass control.
        HOPE THIS HELPS U....


--- In Abaqus@..., "maajidchishti" <maajidchishti@...> wrote:

>
> Further information on the model
> I am using Linear solid elements C3D8 and C3D6. The model contains a composite plate joint to an aluminium plate with a titanium countersunk bolt. I have modeled all the components of the joint using solid elements. I have also implemented contact and friction in the model. the joint is fixed at one end and loaded in tension on the other end.
> Your help will be much appreciated.
> Regards
> Maajid
>
> --- In Abaqus@..., "maajidchishti" <maajidchishti@> wrote:
> >
> > Dear all
> >
> > I am running a model of composite joint to failure. I would like anyone to help me in reducing the stiffness of the model. The model is being very stiff compared to the real structure. I am aware of the fact that FEM models are usually stiffer than the real structure, however my model is too stiff compared to the real structure.
> >
> > I am using viscosity to lead to convergence and I am using implicit analysis in my simulation. My model also has bolt pre-load, contact and friction in the simulation. I would like anyone to suggest ways to reduce the stiffness of the model. Please give ideas, how can I reduce the stiffness of the model. Thanking you in advance.
> >
> > Kind regards
> >
> > Maajid
> >
>

[maajidchishti]

thanks Sirnivas

I will try to use reduced integration elements with hourglass control.  can hourglass be a problem during simple tensile loading analysis? Thank you
regards
Maajid

--- In Abaqus@..., "srinivas" <balaga_dfm@...> wrote:
>
> haii maajid,
>
>            In my knowledge C3D8 and C3D6 elements are the full integration elements. they are used only while doing the frequency analysis. normally fully integration elements exhibit shear locking so they are the stiff elements, so they should not be used. and reduced integration elements overcome this shear locking however they too have problem called hourglassing. but this can be resolved to the maximum extent using the present softwares. so if u are modelling a composite material using solid elements better to use the reduced integration elements with enhanced hourglass control.
>         HOPE THIS HELPS U....

[srinivas srinivas-2]

haii maajid,
       you can use with no doubt. simple tensile test has no problem.


--- In Abaqus@..., "maajidchishti" <maajidchishti@...> wrote:

>
> thanks Sirnivas
>
> I will try to use reduced integration elements with hourglass control.  can hourglass be a problem during simple tensile loading analysis? Thank you
> regards
> Maajid
>
> --- In Abaqus@..., "srinivas" <balaga_dfm@> wrote:
> >
> > haii maajid,
> >
> >            In my knowledge C3D8 and C3D6 elements are the full integration elements. they are used only while doing the frequency analysis. normally fully integration elements exhibit shear locking so they are the stiff elements, so they should not be used. and reduced integration elements overcome this shear locking however they too have problem called hourglassing. but this can be resolved to the maximum extent using the present softwares. so if u are modelling a composite material using solid elements better to use the reduced integration elements with enhanced hourglass control.
> >         HOPE THIS HELPS U....
>