Hi Gio,

I searched my notes and manuals about this and came up with nothing. In
all the examples I've seen, the first time is always zero, and I have
been in habit of setting up my power functions this way for a long
time. I don't remember what FEFLOW assumes for a situation like you
describe, but my guess is that it will either use the value at time 730
for the value at time 0 or will set the value to 0 at time 0 and
interpolate between time 0 and time 730.

There is no direct way of "turning off" (deactivating) a boundary
condition (BC) within a single feflow model, unless you use the
SwitchBC module which you'd have to get from WASY. In some cases a BC
node can be deactivated using BC constraints, but this will not always
work (especially as the model complexity or number of such constraints
increases). You can also deactivate a BC by using two feflow models,
one that simulates a portion of time without the BC, and one set up with
it. The first model would be used to initialize the second. Run a
model from zero to 730 without the BC, use the result to initialize a
second model that has the time varying BC.

Pete


Giovanni Firmani wrote:
>
> Pete,
> I have a terrible doubt now. If I define a power function (for a head
> boundary condition for example) file like the following one
>
> !
> ! t [d] fvalue
> !---------------------
> 730 334
> 1460 334
>
> does it mean that the head boundary condition is off between 0 and
> 730? Running the model I noticed that is not like that but if I do not
> want the boundary condition ON I have to use a fake head and a
> constrain that is able to disable it between 0 and 730.
>
> Giovanni
>
> On Sun, Feb 8, 2009 at 12:59 PM, Pete Sinton <psinton@...
> <mailto:psinton%40aquageo.us>> wrote:
> > Hi Gio,
> >
> > Yes, that works. I find it convoluted and cumbersome, so I resort to
> > bypassing feflow's interface. I know I will get what I want.
> >
> > Glad to hear you solved the problem
> >
> > Cheers
> >
> > --
> >
> > Pete Sinton
> >
> > Principal Hydrogeologist
> >
> > AquaGeo, Ltd.
> >
> > 11560 Penny Rd
> >
> > Conifer CO 80433
> >
> > Cell 720.841.5526
> >
> > Phone 303.838.7863
> >
> > psinton@... <mailto:psinton%40aquageo.us>
> >
> > www.aquageo.us
> >
> > This e-mail together with any attachments is confidential and may be the
> > subject of legal privilege. If you are not the intended recipient,
> you are
> > not permitted to print, copy, disclose or use the content or
> attachments in
> > any way. AquaGeo, Ltd. respectfully requests that you reply to this
> e-mail
> > and delete this message. AquaGeo, Ltd. accepts no responsibility for
> changes
> > made to this e-mail or any attachments after transmission from
> AquaGeo, Ltd.
> > Thank you.
> >
> > Thanks Pete,
> >
> > As workaround I manage to generate some input files: one for the
> > "power functions" and one for the "time variant ID". For the
> > generation of the file, I have temporary assigned the boundary
> > conditions as the time-constant and the I exported them as ASCII file
> > that I modified and imported again to simulate time-variant boundary
> > conditions.
> > However, I noticed that importing an external file using the command
> > "database" is quite tricky. Following the procedure described in this
> > forum:
> >
> > http://feflow.info/forum/index.php/topic,446.msg1332.html#msg1332
> <http://feflow.info/forum/index.php/topic,446.msg1332.html#msg1332>
> >
> > I used the commands: Assign - database - Import Time-varying IDs -
> > Head BC/Seepage - Automatic regionalization.
> >
> > But I had to use 'Inverse
> > Distance Weighting' with only one neighboring point as the
> > regionalization method for getting the exact numbers I wanted for each
> > node.
> >
> > In theory the command "interpolate" the data. To avoid the
> > interpolation I used "