News:

alphaMELTS 1.9 is available at the download and information site.
alphaMELTS 2 standalone & for MATLAB/Python are available on request
(see the Version 2 tab on the download site for more details).
For news of all MELTS software see the MELTS Facebook page.

Main Menu

I'm almost there.

Started by coblem, March 04, 2008, 08:28:03 AM

Previous topic - Next topic

coblem

Hi Paula.

Like Sara, I am also really new to MELTS, and have got half way through the adiabat_1ph tutorial before I got stuck.

I have installed Perl, and I am running through the Command Prompt on Windows.  I put the contents of the MELTS folder into my default command prompt folder, so I don't have to reset my defaults.

I am stuck on step 4 of the tutorial (choice 10) where I am to 'find the liquidus' in phase diagram mode.  I re-downloaded the files, and still can't find my error.  Instead of the correct output from the tutorial:

Quote
Your choice: 10

... etc...

Enter 1 for special upwards-only search for troublesome phases: 0
...Adding the solid phase feldspar to the assemblage.

... etc ...

...Quadratic convergence accepted, but non-optimal.
...Dropping phase feldspar from the assemblage.
Isograd solution at: P 500.000000, T 1493.349219
liquid: 100.350 g 48.51 1.01 17.58 0.89 0.03 7.56 9.07 12.41 2.64 0.03 0.08 0.20
Activity of H2O = 0.0179036
Return to menu now (1) or increment P and continue calculations (0)? 1



The output file I get, following the same steps, reads as follows.

Quote
Your choice: 10

... etc...

Enter 1 for special upwards-only search for troublesome phases: 0
...Adding the solid phase feldspar to the assemblage.
...Adding the solid phase olivine to the assemblage.
Initial Guess (not an isograd solution): P 500.000000, T 1473.150000
liquid: 82.239 g 49.00 1.23 15.85 1.08 0.04 8.49 8.71 12.39 2.83 0.04 0.10 0.24

Activity of H2O = 0.0266901

...etc...

...Dropping phase feldspar from the assemblage.
...Adding the solid phase feldspar to the assemblage.
...Dropping phase feldspar from the assemblage.
...Adding the solid phase feldspar to the assemblage.
...Dropping phase feldspar from the assemblage.
...Adding the solid phase feldspar to the assemblage.
...Dropping phase feldspar from the assemblage.
THE QUADRATIC MINIMIZATION ALGORITHM HAS FAILED TO CONVERGE.
...Quadratic convergence failure. Aborting.
Not all calculations performed!


I don't know what I did wrong, and am at a loss as to where to go from here.
Have you seen this before, or do you have any suggestions.

I have also attached a copy of the full command line input/output as a txt document if my above quote isn't clear.

Thanks very much.

Sincerely,

Matthew Coble

Paula


Hi Matthew,

You didn't do anything wrong.  I tried it and got the same error.  I'm looking into it at the moment and will let you know when a corrected version is available.  Sorry for the inconvenience.  In the meantime, try setting the temperature to 1221 oC with option 2 and then miss out the option 10 step altogether.  That should give you an idea how a simple calculation works until the liquidus function is fixed.

Paula

coblem

Hi Paula.

Thanks for the quick response.
After changing the temperature manually with option 2 (I used 1223 so it would be consistent with the posted results), and omitting option 10,  it worked fine.  My output file was identical to the posted answers. 


However, I am trying to run a simple isobaric fractional crystallization of a natural basalt sample from Steens Mountain.  When I try to run option 10, it would always freeze at the first attempt of '...Adding the solid phase olivine to the assemblage.'.  After reading the post about Isobaric Batch Melting by jcwhite, I tried removing the H2O buffer.  I got a little further (many more calculated steps), but then got into an endless loop of
...Adding the solid phase olivine to the assemblage.
...Dropping phase olivine from the assemblage.
...Adding the solid phase olivine to the assemblage.
...Dropping phase olivine from the assemblage.
...Adding the solid phase olivine to the assemblage.
...Dropping phase olivine from the assemblage.
...Adding the solid phase olivine to the assemblage.
...Dropping phase olivine from the assemblage.
                    ...etc, forever...


Does this have the anything to do with the above problem, or am I making a different mistake?
Would it be better to just guess a temperature in option 2 through trial and error, or will the results be different?

Thanks again for all your time and help.
Sincerely,

Matthew Coble


Paula

Hi Matthew,

I found the problem in adiabat_1ph.  It turned out to be a little more difficult to fix than I expected and not a simple reversal of previous changes.  I've been trying to make the isograd routine a little less temperamental recently anyway (such as making the special 'upwards' search for trouble phases available as a special 'downward' search for troublesome liquid).  So I've made some of those improvements a bit more general.  Long story short... the tutorial example works now. 

I'm trying to get the next version of adiabat_1ph up this week.  I'll update the tutorial post to match at the same time as there are some slight differences now (e.g. temperature is marginally higher but only by ~0.004 oC).

If you can hang on a couple of days you'll be able to run the Steens Mountain basalt calculation with the new version.  In the meantime, you can either use trial and error with a combination of option 2 and 3 or, since you know that the liquidus phase is probably olivine, you can run option 10 and look for the olivine-out reaction instead.  Either method should give similar enough results.

For example, if you were going to find the liquidus for the tutorial example by looking for the feldspar-out reaction you would see something like this:

Quote
Your choice: 10
Stable (0) or metastable (1) search? 0
Phase to track boundary of (by name, lower case): feldspar
How many coexisting phases to track? 1
Enter 1 if phase is in on high-temperature side: 0
Enter 1 for special monotonic search for troublesome phases: 0
Initial Guess (not an isograd solution): P 500.000000, T 1488.509375
liquid: 98.346 g 48.53 1.03 17.25 0.90 0.03 7.72 9.25 12.32 2.65 0.03 0.08 0.20
Activity of H2O = 0.0187985  Melt fraction = 0.980027
feldspar: 2.004301 g, composition K0.00Na0.18Ca0.82Al1.82Si2.18O8
...Dropping phase feldspar from the assemblage.
...Adding the solid phase feldspar to the assemblage.
...Quadratic convergence was acceptable at iterQuad = 47 (rNorm = 1.32697e-12).
...Quadratic convergence was acceptable at iterQuad = 35 (rNorm = 1.12141e-11).
...Dropping phase feldspar from the assemblage.
...Adding the solid phase feldspar to the assemblage.
...Quadratic convergence was acceptable at iterQuad = 84 (rNorm = 3.28951e-11).
...Dropping phase feldspar from the assemblage.
...Adding the solid phase feldspar to the assemblage.
...Quadratic convergence was acceptable at iterQuad = 60 (rNorm = 9.21743e-12).
...Quadratic convergence was acceptable at iterQuad = 98 (rNorm = 2.28834e-11).
...Dropping phase feldspar from the assemblage.
Isograd solution at: P 500.000000, T 1493.353125
liquid: 100.350 g 48.51 1.01 17.58 0.89 0.03 7.56 9.07 12.41 2.64 0.03 0.08 0.20
Activity of H2O = 0.0179036  Melt fraction = 1
Return to menu now (1) or increment P and continue calculations (0)? 1

Choose:

It will be slightly simpler for olivine as you won't be given the option of tracking more than one coexisting phase.

Thanks for your patience and for help in reporting the problem.  It's good to have it fixed just before the next release instead of afterwards.

Paula


coblem

Thanks so much, again, for the response. 
I'll try your suggestions for practice and wait to use the new version when it's available.

Matthew

!acompdummy

how is this with windows vista?

coblem


Paula

Thanks for the reply Matthew.  I also only have XP.  I've been told by a few people that everything works pretty much the same on Vista as on XP.  If anyone does have problems with Adiabat_1ph on Vista we'd be glad to hear about them...

I think !acompdummy's post was spam though.  The bumpers on my new car are just fine and I don't think any users unfortunate enough to be involved in a fender-bender will be looking here for repair options!  As the Vista question is a legitimate one and the web link is not offensive I'm leaving the post intact.  But in general spamming leads to an immediate ban.

Cheers,
Paula