[gtg258f]

Dear all,

I have a solver with an outer time loop, and inner loop with sub-iterations. In order to check the convergence of the outer and inner iterations, I wanted to access my field values at the
- previous time step: p.oldTime()
- previous inner iteration: p.prevIter()

While prev.Iter() seems to work fine, I am encountering issues with oldTime() and was wondering if I was doing something wrong.

I am storing values with storePrevIter(), should I use something else for oldTime()?

Another option is that there may be a memory issue with oldTime:
when one adds print statements before modifying a field value, then oldTime() is stored properly whereas it is not without the print statement (which seems fishy)


Playing around with the problem, I wrote a very simple code (attached) with the following algorithm, and a set-up ready to run (though the example code really can run on any case)

DP=1
for( time )
    p.storePrevIter()
    //comment 1: print DP.oldTime()
    for( innerIteration<2 )
    {
        DP.storePrevIter()
        DP+=DP
    }
    //comment 2: print p.oldTime()
    p = DP

without any print statement, the values at the end of the first time steps are:
p[0] oldTime, prevIter, current = 4 1 4
DP[0] oldTime, prevIter, current = 4 2 4

with the print statements in comment 1 and 2, the values are properly stored and read as:
p[0] oldTime, prevIter, current = 1 1 4
DP[0] oldTime, prevIter, current = 1 2 4


Any help on that problem would be great,
Diane

[mbookin]

I do not know why this happen, but of you call p.oldTime() before you update your p then p.oldTime() will have the right value. You dont even need to print, just call p.oldTime(); I think pseudocode can illustrate this:

P=1;
for( time )
    {
    P+=P
    Info<<P<<"_"<<P.oldTime()<<endl;
    }

Will give

2_2   
3_3
..

And

P=1;
for( time )
    {
    P.oldTime();
    P+=P
    Info<<P<<"_"<<P.oldTime()<<endl;
    }

Will return

2_1
3_2
..