Search
Close this search box.
Notifications
Clear all

About UEL problem!

4 Posts
2 Users
0 Reactions
478 Views
Posts: 4
Topic starter
(@hadishosseini)
New Member
Joined: 12 years ago

About UEL problem!

Topic Tags
3 Replies
Posts: 3998
(@jorgen)
Member
Joined: 5 years ago

How different are your predictions from Abaqus?
How did you evaluate the UEL?

-Jorgen

Topic Tags
3 Replies
Posts: 4
Topic starter
(@hadishosseini)
New Member
Joined: 12 years ago

I apply same boundary condition on the model with just one element.

Now, the stress and strain calculated by UEL is different with that calculated with abaqus directly. My uel is just written for C3D8 element, which is also provided in ABAQUS.

I have checked the B matrix and D matrix, and they are all correct. So I think I made some mistake in defining RHS or AMATRX.

The following is my INPUT file.

*Heading

** Job name: Cube_UEL Model name: Model-1

** Generated by: Abaqus/CAE Version 6.8-1

*Preprint, echo=NO, model=NO, history=NO, contact=NO

**

** PARTS

**

*Part, name=Cube

*Node

1, 0., 0., 0.

2, 1., 0., 0.

3, 1., 1., 0.

4, 0., 1., 0.

5, 0., 0., 1.

6, 1., 0., 1.

7, 1., 1., 1.

8, 0., 1., 1.

*USER ELEMENT, NODES=8, TYPE=U1, COORDINATES=3,I PROPERTIES=0,PROPERTIES=2,VARIABLES=96

1,2,3

*Element, type=U1

1, 1, 2, 3, 4, 5, 6, 7, 8

*Nset, nset=ALL, internal, generate

1, 8, 1

*Elset, elset=ALL, internal

1,

** Section: Cube

**Solid Section, elset=ALL, material=steel

*UEL PROPERTY, ELSET=ALL

210000., 0.3

*End Part

**

**

** ASSEMBLY

**

*Assembly, name=Assembly

**

*Instance, name=Cube-1, part=Cube

*End Instance

**

*Elset, elset=ELEMENT, instance=CUBE-1

1,

*Nset, nset=X0, instance=CUBE-1

1, 4, 5, 8

*Nset, nset=Xn, instance=CUBE-1

2, 3, 6, 7

*Nset, nset=Y0, instance=CUBE-1

1, 2, 5, 6

*Nset, nset=Yn, instance=CUBE-1

3, 4, 7, 8

*Nset, nset=Z0, instance=CUBE-1, generate

1, 4, 1

*Nset, nset=Zn, instance=CUBE-1, generate

5, 8, 1

*End Assembly

**

** BOUNDARY CONDITIONS

**

** Name: BC-1 Type: Displacement/Rotation

*Boundary

X0, 1, 1

** Name: BC-2 Type: Displacement/Rotation

*Boundary

Y0, 2, 2

** Name: BC-3 Type: Displacement/Rotation

*Boundary

Z0, 3, 3

** ----------------------------------------------------------------

**

** STEP: Step-1

**

*Step, name=Step-1

*Static

0.001, 1., 1e-05, 1.

**Step, name=Step-1, perturbation

**Static

**

** BOUNDARY CONDITIONS

**

** Name: BC-4 Type: Displacement/Rotation

*Boundary

Xn, 1, 1, 0.1

**

** OUTPUT REQUESTS

**

*Restart, write, frequency=0

**

** FIELD OUTPUT: F-Output-2

**

**Output, field

**Element Output

**SDV,

**

** FIELD OUTPUT: F-Output-1

**

*Output, field, variable=PRESELECT

**EL PRINT,ELSET=RFSET

**SDV,

**EL File,ELSET=RFSET

**SDV

**

** HISTORY OUTPUT: H-Output-2

**

*Output, history

**Element Output, elset=RFSET

**SDV,

**

** HISTORY OUTPUT: H-Output-1

**

*Node Output, nset=XN

U1, U2, U3

*Node Output, nset=X0

U1, U2, U3

*End Step

Reply
Posts: 4
Topic starter
(@hadishosseini)
New Member
Joined: 12 years ago

I wrote the DU(delta displacement) and U(displacement of nodes) in message file.

I know, at the beginning of the increment, DU should be zero. But, then strain and stress are also zero. When calculated the RHS = int(transpose(B)*stress)*DV, it also should be zero.

I am confused, why at the end of the increment, U = inverse(AMATRX)*RHS is not zero?

I think I made some mistake on defining RHS.

Could you give me some suggestions?

Thank you.

Reply
Share: