Matte–slag interaction simulation in the flash smelting settler using coupled CFD-DEM

Jani-Petteri Jylhä

Matte–slag interaction simulation in the flash smelting settler using coupled CFD-DEM

Julkaisun otsikon käännös: Kiven ja kuonan vuorovaikutuksen simulointi liekkisulatusuunin alauunissa kytketyllä CFD-DEM-menetelmällä

Jani-Petteri Jylhä

Tutkimustuotos: Doctoral Thesis › Collection of Articles

Abstrakti

Liekkisulatus on merkittävä kuparinvalmistusmenetelmä. Uunin sisällä kuonakerroksen läpi laskeutuvia kuparikivipisaroita on kuitenkin mahdotonta tarkastella korkean lämmön ja uunin suljetun rakenteen vuoksi. Tästä syystä uunissa tapahtuvia ilmiöitä on tutkittu numeerisen virtausdynamiikan (computational fluid dynamics, CFD) avulla. Tässä tutkimuksessa kytkettyä CFD-DEM (discrete element method, DEM) menetelmää käytettiin simuloimaan sulia kuparikivipisaroita liekkisulatusuunin alauunissa. Perinteisesti DEM on ollut kiinteiden partikkelien mallinnusmenetelmä. Sulia pisaroita saatiin kuitenkin tutkittua simuloimalla niitä pehmeinä palloina, joiden kanssa käytettiin tässä työssä kehitettyjä alimalleja pisaroiden yhdistymisen ja reaktiokinetiikan vaikutusten kuvaamiseksi. Koska korkealämpötilaprosessia oli mahdotonta havainnoida suoraan, menetelmän validoimiseksi kehitettiin fyysinen malli käyttäen öljyä ja muovipalloja. Pallot syötettiin öljyyn, jonka läpi laskeutuessa ne kuvattiin. Vastaava malli myös simuloitiin, johon kuvattuja tuloksia voitiin verrata. Sekä kokeessa että simulaatiossa näkyi kanavoituvaa virtausta. Virtauksessa pallojen luoma vastus loi imun, joka veti viereisiä palloja lähemmäs, jolloin syötetyt pallot muodostivat kanavan öljyn läpi. Muodostunut kanava nopeutti pallojen laskeutumista merkittävästi. Öljyn pohjalla kuitenkin muodostunut virtaus kääntyi kohti laitoja ja siitä ylöspäin, mikä kykenee kuljettamaan hitaammin laskeutuvia pisaroita ja siten ne voivat jäädä öljykerrokseen. Jokaisessa CFD-DEM:llä lasketussa simulaatiossa havaittiin samanlainen kanavoitumisilmiö, joka sai kivipisarat laskeutumaan kuonan läpi merkittävästi nopeammin kuin Stokesin lailla lasketun laskeutumisnopeuden mukaan. Yhdistymis- ja reaktiokinetiikkamallit nopeuttivat laskeutumista entisestään, sillä yhdistyminen kasvattaa pisaroiden kokoa ja reaktiokinetiikka aiheuttaa pisaran tiheyden kasvua. Tiheys nousee, kun raudan ja rikin poistuessa pisaran kuparipitoisuus kasvaa. Kanavoituminen myös lisäsi pisaroiden yhdistymismäärää tuomalla pisarat lähemmäs toisiaan, jolloin ne törmäilivät ja myöskin yhdistyivät enemmän.

Julkaisun otsikon käännös	Kiven ja kuonan vuorovaikutuksen simulointi liekkisulatusuunin alauunissa kytketyllä CFD-DEM-menetelmällä
Alkuperäiskieli	Englanti
Pätevyys	Tohtorintutkinto
Myöntävä instituutio	Aalto-yliopisto
Valvoja/neuvonantaja	Jokilaakso, Ari, Vastuuprofessori Jokilaakso, Ari, Ohjaaja
Kustantaja	Aalto University
Painoksen ISBN	978-952-64-1680-9
Sähköinen ISBN	978-952-64-1681-6
Tila	Julkaistu - 2024
OKM-julkaisutyyppi	G5 Artikkeliväitöskirja

Tutkimusalat

laskennalliset menetelmät
kupari
pisarat

Pääsy asiakirjaan

http://urn.fi/URN:ISBN:978-952-64-1681-6

Siteeraa tätä

@phdthesis{fe7973110be74b34a02a7f9f9a4e6554,

title = "Matte–slag interaction simulation in the flash smelting settler using coupled CFD-DEM",

abstract = "The flash smelting process is a major copper production method. However, a closed furnace with high temperatures makes physical observations of matte settling through the slag physically impossible and, thus, computer simulations are needed to study settling. Previously, settling simulations have been made using computational fluid dynamics (CFD). In this project, a new method of using coupled CFD-DEM (discrete element method) was developed and used to study copper matte droplets in a flash smelting settler. Traditionally, CFD-DEM is a fluid and solid particle simulation method; however, by simulating soft spheres and creating user-defined submodels for coalescence and reaction kinetics, liquid matte droplets settling through a slag layer could be studied. Due to the impossibility of observing a high-temperature process in situ, a room temperature physical model was needed to validate the chosen simulation method. Plastic spheres were fed into an oil bath, in which their settling was filmed and compared to simulation results of the same system. Both the experiments and the simulation produced a channeling flow, where drag from the settling spheres pulls the flow of droplets into a narrow channel, accelerating the settling velocity significantly. However, at the bottom of the oil, the flow turns sideways and then up, forming a vortex ring around the channel, which may entrain slowly settling objects. Several matte–slag simulations were made with CFD-DEM, utilizing either the coalescence model or the combined coalescence and reaction kinetics model. Channeling flow, similar to the sphere–oil experiment and simulation, occurred with every matte–slag simulation. The flow caused the matte droplets to settle through the slag significantly faster than predicted by Stoke's law. The coalescence and reaction kinetics models further accelerated the matte droplets as coalescence increased the size of the droplets while reaction kinetics caused copper to concentrate in the droplets as iron and sulfur were removed, which increased droplet density. The coalescence rate was increased by the channeling flow as it caused the droplets to enter a narrower channel, causing an increased number of collisions and, thus, coalescence.",

keywords = "computational methods, copper, droplets, laskennalliset menetelm{\"a}t, kupari, pisarat, computational methods, copper, droplets",

author = "Jani-Petteri Jylh{\"a}",

year = "2024",

language = "English",

isbn = "978-952-64-1680-9",

series = "Aalto University publication series DOCTORAL THESES",

publisher = "Aalto University",

number = "36",

address = "Finland",

school = "Aalto University",

}

TY - BOOK

T1 - Matte–slag interaction simulation in the flash smelting settler using coupled CFD-DEM

AU - Jylhä, Jani-Petteri

PY - 2024

Y1 - 2024

N2 - The flash smelting process is a major copper production method. However, a closed furnace with high temperatures makes physical observations of matte settling through the slag physically impossible and, thus, computer simulations are needed to study settling. Previously, settling simulations have been made using computational fluid dynamics (CFD). In this project, a new method of using coupled CFD-DEM (discrete element method) was developed and used to study copper matte droplets in a flash smelting settler. Traditionally, CFD-DEM is a fluid and solid particle simulation method; however, by simulating soft spheres and creating user-defined submodels for coalescence and reaction kinetics, liquid matte droplets settling through a slag layer could be studied. Due to the impossibility of observing a high-temperature process in situ, a room temperature physical model was needed to validate the chosen simulation method. Plastic spheres were fed into an oil bath, in which their settling was filmed and compared to simulation results of the same system. Both the experiments and the simulation produced a channeling flow, where drag from the settling spheres pulls the flow of droplets into a narrow channel, accelerating the settling velocity significantly. However, at the bottom of the oil, the flow turns sideways and then up, forming a vortex ring around the channel, which may entrain slowly settling objects. Several matte–slag simulations were made with CFD-DEM, utilizing either the coalescence model or the combined coalescence and reaction kinetics model. Channeling flow, similar to the sphere–oil experiment and simulation, occurred with every matte–slag simulation. The flow caused the matte droplets to settle through the slag significantly faster than predicted by Stoke's law. The coalescence and reaction kinetics models further accelerated the matte droplets as coalescence increased the size of the droplets while reaction kinetics caused copper to concentrate in the droplets as iron and sulfur were removed, which increased droplet density. The coalescence rate was increased by the channeling flow as it caused the droplets to enter a narrower channel, causing an increased number of collisions and, thus, coalescence.

AB - The flash smelting process is a major copper production method. However, a closed furnace with high temperatures makes physical observations of matte settling through the slag physically impossible and, thus, computer simulations are needed to study settling. Previously, settling simulations have been made using computational fluid dynamics (CFD). In this project, a new method of using coupled CFD-DEM (discrete element method) was developed and used to study copper matte droplets in a flash smelting settler. Traditionally, CFD-DEM is a fluid and solid particle simulation method; however, by simulating soft spheres and creating user-defined submodels for coalescence and reaction kinetics, liquid matte droplets settling through a slag layer could be studied. Due to the impossibility of observing a high-temperature process in situ, a room temperature physical model was needed to validate the chosen simulation method. Plastic spheres were fed into an oil bath, in which their settling was filmed and compared to simulation results of the same system. Both the experiments and the simulation produced a channeling flow, where drag from the settling spheres pulls the flow of droplets into a narrow channel, accelerating the settling velocity significantly. However, at the bottom of the oil, the flow turns sideways and then up, forming a vortex ring around the channel, which may entrain slowly settling objects. Several matte–slag simulations were made with CFD-DEM, utilizing either the coalescence model or the combined coalescence and reaction kinetics model. Channeling flow, similar to the sphere–oil experiment and simulation, occurred with every matte–slag simulation. The flow caused the matte droplets to settle through the slag significantly faster than predicted by Stoke's law. The coalescence and reaction kinetics models further accelerated the matte droplets as coalescence increased the size of the droplets while reaction kinetics caused copper to concentrate in the droplets as iron and sulfur were removed, which increased droplet density. The coalescence rate was increased by the channeling flow as it caused the droplets to enter a narrower channel, causing an increased number of collisions and, thus, coalescence.

KW - computational methods

KW - copper

KW - droplets

KW - laskennalliset menetelmät

KW - kupari

KW - pisarat

KW - computational methods

KW - copper

KW - droplets

M3 - Doctoral Thesis

SN - 978-952-64-1680-9

T3 - Aalto University publication series DOCTORAL THESES

PB - Aalto University

ER -

Matte–slag interaction simulation in the flash smelting settler using coupled CFD-DEM

Abstrakti

Tutkimusalat

Pääsy asiakirjaan

Sormenjälki

Siteeraa tätä