Re: bernd

["Bernd Ehret" <bernd.ehret@xxxxxxxxxxxxx>]

thank you !!

answer with file tomorrow

 

Bernd

 

 

----- Original Message -----

From: "David L Parkhurst" <dlpark@xxxxxxxx>

To: "Bernd Ehret" <bernd.ehret@xxxxxxxxxxxxx>

Sent: Wednesday, September 04, 2002 3:45 PM

In-Reply-To:

Subject: Re: bernd

> 
> Very large mixing fractions (100000) will probably cause 
numerical
> problems; fractions < 10 or even 100 should be ok provided 
the solutions
> start with about 1 kg water.
> 
> 
> 
> Which fractions I have to use if some Deka or even millions of 
cubic
> meter of
> water are to be mixed ?
> 
> You will 
have to use representative volumes, summing to  1 is preferable.
> 
> > With the results from the  MIX steps mentioned above I 
couldn't estimate
> which kind
> of fractions give the right 
answer.
> 
> For pure mixing, results should be the same with the 
same proportions. I
> would have to see a complete input file, hopefully 
not too complicated, to
> tell you what is going on.
> 
> 
David
> 
> David Parkhurst (dlpark@xxxxxxxx)
> U.S. 
Geological Survey
> Box 25046, MS 413
> Denver Federal 
Center
> Denver, CO 80225
> 
> Project web page: https://wwwbrr.cr.usgs.gov/projects/GWC_coupled
> 
> 
> 
> 
                                                                                                                      
> 
                      
"Bernd 
Ehret"                                                                                   
> 
                      
<bernd.ehret@xx-         
To:      "David L Parkhurst" <dlpark@xxxxxxxx

>                        
>                       Cottbus.De>              cc:                                                                   
>                                                In-Reply-To: Subject: Re: bernd                                                    
>                       09/04/02 04:51                                                                                 
>                       AM                                                                                             
>                                                                                                                      
>                                                                                                                      
>
>
>
> From many tests with MIX I got different results for the same mixing
> fractions
> of the input solutions, for example:
>
> MIX 1
> # 0 0.1
> # 1 0.15
> # 2 0.25
> # 3 0.5    -> norm 1
>
> # 0 100000
> # 1 150000
> # 2 250000
> # 3 500000 -> no convergence
>
> 0 7.2
> 1 10.8
> 2 18.0
> 3 36.0     -> no convergence
>
> # 0 0.072
> # 1 0.108
> # 2 0.180
> # 3 0.360 -> different results as with first fractions
> SAVE solution 1001
> END
>
> MIX 2
> # 0 0.1
> # 1 0.2
> # 2 0.3
> # 1001 0.4
>
> # 0 100000
> # 1 200000
> # 2 300000
> # 1001 400000
>
> 0 7.2
> 1 14.4
> 2 21.6
> 1001 28.8
>
> # 0 0.072
> # 1 0.144
> # 2 0.216
> # 1001 0.288
> SAVE solution 1002
> END
>
> MIX 3
> # 0 0.1
> # 1 0.25
> # 2 0.35
> # 1002 0.3
>
> # 0 100000
> # 1 250000
> # 2 350000
> # 1002 300000
>
> 0 7.2
> 1 18.0
> 2 25.2
> 1002 21.6
>
> # 0 0.072
> # 1 0.180
> # 2 0.252
> # 1002 0.216
> SAVE solution 1003
> END
>
> My conclusions are:
> -the mixing fractions may not be to large, that is not >> 1
> -norm to 1.0 gives other results than without norm
>
> The question is:
> Which fractions I have to use if some Deka or even millions of cubic meter
> of
> water are to be mixed ?
> With the results from the  MIX steps mentioned above I couldn't estimate
> which kind
> of fractions give the right answer.
>
> Bernd
>
>
>
>
>