Alternative colloidal binders having a high solid content to colloidal silica or alumina for refractory castables
ABDELOUHAB S. 1, LANG C. 1, DELMOTTE C. 1, VANDENEEDE V. 1, THIESEN M. 2, IBARRA L. 2, KRAUSE O. 2, BROCHEN E. 3, DANNERT C. 3
1 BCRC, Mons, Belgium; 2 Koblenz University of Applied Sciences, Höhr-Grenzhausen, Germany; 3 Forschungsgemeinschaft Feuerfest e. V. (FGF) at the European Centre for Refractories (ECREF), Höhr-Grenzhausen, Germany
Historically, the improvement of the high temperature properties of refractory castables was firstly achieved by a reduction of the share of aluminate cement (CAC) binder in the composition, moving hence from castables with medium or regular cement content towards low cement content or even ultra-low cement content. However, the presence of CaO remaining in the mix is still detrimental to some high temperature properties especially in corrosive environment. Moreover, CAC bonded castables must be dried carefully up to 600 °C to avoid premature cracks or even an explosion of a part of the lining or preshaped products.
Therefore, to overcome these issues, cement-free castables were developed by using alternative and innovative binders such as hydratable alumina (HA), colloidal silica (CS) and colloidal alumina (CA).
Although refractory monolithics bonded with HA perform better in corrosive environments than CAC ones and present also satisfactory thermomechanical properties, the drying step of such materials is time and energy consuming to avoid cracking/explosion of the lining which could be induced by the evacuation of the mixing water during the drying step.
The use of colloidal suspensions as binders for refractory castables was then considered. The literature review on colloidal bonded castables reports either the use of commercial colloidal suspensions or the use of lab colloidal suspensions. For the castables bonded with commercial colloidal binder, the most famous is CS following by CA. These ones have a high solid content from around 30 wt.% up to 60 wt.%. However, for the CS bonded castables, the silica content can still be a drawback for some steel applications. Indeed, the presence of amorphous silica in the composition of these refractory concretes leads to the apparition of liquid/viscous phases at high temperature and therefore induces a reduction in their thermomechanical resistance. It will also be an issue for the new steel process using direct reduction iron ore by hydrogen in forming metallic silicon. In this context, the commercial CA or lab colloidal suspension silica free as binders are considered with more reliability.
For the castables bonded with lab colloidal binders, these ones have a low solid content, around 5 wt.% up to 15 wt.%, and were produced by sol-gel route or coprecipitation. The castables performances studies bonded with these lab colloidal binders were focused on thermomechanical/thermochemical properties. These studies indicate that castables bonded with lab boehmite, spinel or mullite binders can present higher mechanical resistance at high temperature than castables bonded with CS or CA. These performances, especially green mechanical properties, could be further improved by increasing the solid content in the starting binders.
The aim of this study is to achieve a performance enhancement of castables bonded with colloidal sols, through the development of lab boehmite, spinel or mullite colloidal suspensions, by using the milling route. This one enables a higher solid content in the sols than the use of sol-gel route, typically the same solid content or higher than commercial colloidal silica suspensions.