Regarding storage of History variables

[ad10]

Dear all

I am formulating a User element in which I required to store previous step strains

Though I am assigning the history variable hr(nh1) for the previous and hr(nh2) for the current step. But, during simulation values are not getting stored!!  any suggestions about what probably going wrong?

do i = 1,3
                  en(i) = hr(nh1+nn)  ! getting previously stored strains
 end do
And updating the history with,

 do i = 1,3
                  hr(nh2+nn)     = ee(i)
end do

For each step, I am getting en as a zero vector.

Thanks in advance!

[JStorm]

Do you want to have the values of the last time step (equilibrium step) or from the last Newton iteration?

You are right with your procedure if you need the values of the last time step.
Maybe you have missed to inform FEAP about the required number of history components?
It is explained in pmanual85.pdf in chapter 5.7.1.

If you want to store your variable in each iteration, then look for nh3 in the manual.

[Prof. R.L. Taylor]

Set nh1 = 3 in isw=1 part of element

Should be hr(nh1+nn-1) = ee(nn). Etc

[ad10]

Thanks for replying back.

I have formulated an element for Linear Viscoelastic material,

The strange issue I am facing here,
My element works very perfectly for single-arm but the solution unconverges (UN) as I introduce another arm(Increases the viscoelastic terms nv)

I am thinking the issue is with storing the qi(1:ntm,2) values because I am getting results for qi(1:ntm,1).

I am storing the elastic strains and shear strains in ISW 3 or 6 only. I have added my element here. Can you please comment?

Thanks in advance

[Prof. S. Govindjee]

nh1 should be set to the number of (scalar) history variables times the number of element gauss points.

So for example if you are storing a single symmetric tensor per arm of the VE model, and you had 4 Gauss points in the element, then nh1 = 6 * 4 * number_arms_in_VE_model.

[ad10]

Thanks a lot Professor!!

I able to formulate one. I am making mistake in sizing of nh1.

[ad10]

Dear all,

I was able to use nh1: previous step variables at step (tn) and nh2: current step variables (tn+1). I want to know, can I get data of all previous values (time data) of variables available for calculating a value of the variable for the current step. Is that possible to do?

[Prof. S. Govindjee]

We only store t_n and t_n+1.  If you would like to have t_{n-1}, t_{n-2}, etc., then you will need to store it yourself.  Two obvious options are to write a macro to write the data to disk or to write a macro to store the data in memory.

[Prof. R.L. Taylor]

Storing ALL previous values is not a good idea as it would involve too much data storage for a problem with many elements.   We always try to do solutions from t_n to t_n+1.