ERROR on msg file and analysis cannot continue 1

[kaninchenofmersey]

Hi all,
I am doing analysis on axial pressure on the 2D hemisphere by using modified riks method.

After 50 increment there is still no yield thus I increase to 80 increment.

However the analysis cannot continue as I encounter errors

THE STRAIN INCREMENT HAS EXCEEDED FIFTY TIMES THE STRAIN TO CAUSE FIRST YIELD AT 1 POINTS.

Here is the input file. Could anyone enlighten me on this error and help me to continue running the analysis?

 

[corus]

I had a look at the model by importing it into 6.7. I can't see a step not can I see any loads. I don't think you have a hemisphere by the way as you have axisyemmetric geometry. Your shape, if it were devloped 360 degrees, would be a kind of dome with the centre pushed down into a point. Move the geometry on to the x=0 line and remove the left hand half so that you have half a hemisphere. Imagine that line swept round and you will have a hemispherical shell.

corus

 

[kaninchenofmersey]

Hi Corus,

Does it mean that I only need 2 nodes which is node 1 and 999 for the center (at the origin). I have to model hemisphere based on ratio radius(R) over thickness(t) =100
thickness is 1 and H which is the distance from apex of the hemisphere to the center is 0.01R=1

 

[kaninchenofmersey]

The H should be 0.1R not 0.01 R.

 

[kaninchenofmersey]

more info...........both right and left edges of the hemisphere should be fixed to the wall.
Distributed load will apply from above to the surface of the hemisphere.

Hope anyone can help.

thanks in advance.

 

[bassmanjax]

Hmmm, I ran it as a STATIC job (no Riks) and it finished okay. Are you trying to capture the snap-through?

Also, as Corus said, the geometry is not a hemisphere. You appear to have modelled a full cross section of a hemisphere rather than half the (revolved) section. An axisymmetric model is the 2D represtation of a 3D model revolved about the y-axis.

Regards

Martin Stokes CEng MIMechE

 

[corus]

I'm not sure how a hemisphere can have right and left edges? Do you mean that around the circumference and at the centre point the henmisphere is fully restrained?

You will obviously need more than 2 nodes in order to model a hemisphere, otherwise it would be just a cone shape. If you're referring to the geometry definition in CAE then 2 points will be sufficient, with an arc drawn between them.

I think you really need to look at examples of axisyemmtric geometry as I don't think you understand the concept. There may be some in the Abaqus documentation.

corus

 

[kaninchenofmersey]

this is the drawing of what I should model anyway u guys have done a good job in enlighten me about modelling a 2D axysymmetric model.Bassmanjax is right about this. It should be revolved at y-axis.

The dome must be fixed at node 1( ive included the coordinate as well from calculation) and at the opposite.

Distributed load DLOAD will apply from above.

What chapter I need to find to read about axysymmetric modelling?

thx

 

[bassmanjax]

Have a look in "Getting Started with ABAQUS: Interactive Edition", section 3.1.

Regards

Martin Stokes CEng MIMechE

 

[kaninchenofmersey]

"Have a look in "Getting Started with ABAQUS: Interactive Edition", section 3.1."


Any tips to write the coding? As far as it concerns the section 3.1 only apply when drawing the part.

 

[bassmanjax]

Section 3.1 covers the different types of elements that are available in ABAQUS. Section 3.1.2 illustrates how an axisymmetric element is defined. Do a search for 'axisymmetric' in the docs - plenty of light reading there...!

There are a few rules with axisymmetric elements in ABAQUS;

- The axis of rotation is the 2 direction (y axis).
- You cannot have negative coordinates in the 1 direction (x axis).

Regards

Martin Stokes CEng MIMechE

 

[kaninchenofmersey]

section 3.1 here?

http://panther4.nus.edu.sg:2080/v6.4/books/gsa/def.htm

or here?

http://panther4.nus.edu.sg:2080/v6.4/books/key/def.htm

can you provide the link please?

thx

 

[bassmanjax]

Those links don't work for me... The docs I am referring to are for v6.7. Looks like that particular section doesn't refer to the same subject between the versions.

I don't have access to the v6.4 docs, but like I said above, do a search for 'axisymmetric' in the docs and you will find plenty of relevant material.

Regards

Martin Stokes CEng MIMechE