[ABAQUS] Energy balance v6.4-v6.7 1

[doubleje]

I have an elastic crack propagation analysis using slave-master method with crack length time-dependent criterion. It works fine unless I can’t find the balance energy between internal energy and work external like the following (ABAQUS Theory Manual/Energy Balance 1.5.5)

If we note :

[ul]
[li]ALLWK : External work for the whole model[/li]
[li]ALLIE : Internal energy for whole model[/li]
[li]ALLSE : Strain energy for whole model[/li]
[/ul]

And here is my results :

I cannot identify what is the problem with the energy balance ABAQUS.
I appreciate it if you can give me any help.

Thanks!

 

[xerf]

1. The first principle of thermodynamics (energy conservation) is in terms of time rates.

2. The energy balance using the total energetic quantities computed by ABAQUS (as history variables) is given by:

ETOTAL= ALLIE (total internal energy)
+ ALLKE (kinematic energy)
+ ALLVD (viscous dissipation energy)
+ ALLFD (frictional dissipation energy)
– ALLWK (external work)
=constant

where:
ALLIE=ALLSE + ALLPD + ALLCD + ALLAE,

ALLSE=recoverable strain energy
ALLPD=plastic dissipation energy
ALLCD=energy dissipated by creep, viscoelasticity and swelling
ALLAE=“artificial” strain energy.

Depending on the type of the problem being solved, some of the quantities above are zero.

 

[doubleje]

Thank xref for your answer, I understand and agree with you for the energy balance. For the problem elastic crack propagation,
all quantities ALLKE, ALLVD, ALLVD, ALLSE , ALLPD , ALLCD , ALLAE are zero. (Fig. 1)

Fig.1 : Energy quantities zero

I obtain:

ALLIE=ALLSE

this relation is verified (Fig. 1)

Fig. 2: Total internal energy

Then I can rewrite the energy balance like this:

ETOTAL = ALLIE (total internal energy) – ALLWK (external work) = constant.

An other way, I have to find:

Delta( ALLIE) = Delta(ALLWK).

the tests pass results don't match for ABAQUS v. 6.4

Fig. 3: Energy balance

Here is my model ABAQUS (input *.INP) for whom has the same question for a crack propagation problem.

Click here to download the model

 

[xerf]

Again, the energy balance (dE_total/dt=0) is in terms of rates. Your time increments are quite large and response seems quite non-linear.

Try to reduce the size of the maximum time increment.

I got almost overlapping curves for ALLWK and ALLIE using tmax=0.05.

(Also, I would turn NLGEOM=ON, but probably you know better if this is suitable.)

 

[dldhdz]

I've tried abaqus 661 with a small tmax, it looks ok, i mean etotal=allie-allwk