MELTS Software Users

MELTS for Excel, rhyolite-MELTS (GUI) and alphaMELTS => The MELTS family of algorithms, and scientific discussion => Topic started by: Arya Udry on August 30, 2016, 09:29:04 AM

Title: Overestimation of spinel in rhyolite-MELTS
Post by: Arya Udry on August 30, 2016, 09:29:04 AM
Hello everyone,

I have been performing rhyolite-MELTS fractional crystallization calculations from typical tholeiitic basalts (at FMQ+1, 0.5 wt % H2O, and 4 kbars). Although my resulting melt compositions seem realistic (as well as the CIPW normative mineralogies), the fractionated phases are completely off. For some models, the fractionated spinel content ends up being 50% (in addition to cpx). I am aware that MELTS and pMELTS overestimate the liquidus T of spinel, but does rhyolite-MELTS also highly overestimate its stability?

Thank you very much

Arya
Title: Re: Overestimation of spinel in rhyolite-MELTS
Post by: Paula on October 24, 2016, 04:58:18 PM
Hi Arya,

Rhyolite-MELTS is basically a tweak of MELTS. So it will have the same problems with overstabilisation of spinel, if there is any Cr2O3 in the system. You could try removing Cr2O3 from the oxides (and adding it as a trace element, if you like). This will tend to underestimate the stability of spinel but may generate fractionated assemblages that are closer to those observed naturally.

Cheers,
Paula