The Effect of Admixtures on the Buildability of 3D Printed Alkali-Activated Materials with Glass Wool

Abstract

Research about alkali-activated materials or geopolymers has increased in the past decade as they are regarded as a potential replacement to cement-based materials. This is due to the fact that alkali-activated materials possess several advantages over cement-based materials, and the most important one is that they are more sustainable to the environment, as they have less CO2 emissions, less energy consumption during the production process, and the use of byproducts as a binder.
The main aim of this work is to develop a mixture that has superior buildability properties with an open time of two hours or more. In order to achieve this, the effect of admixtures, specifically retarders and viscosity modifying admixtures had to be studied. The use of admixtures in geopolymer has been regarded as a complex subject because multiple factors influence their effect on the mixture, like: binders used and their percentages, alkaline activator used, and liquid to binder ratio.
In this work, the effect of 3 retarders: sucrose, sodium chloride, and sodium borate on the setting time, flowability, and yield stress development was studied. Moreover, the effect of those retarders was studied with the addition of viscosity modifying admixtures: nano-clay (attapulgite), xanthan gum, sodium carboxymethyl starch, and sodium carboxymethyl cellulose on the same properties.
The binders and the alkaline activator used were fixed through out the entire work. the binder composed of 60% fly ash, 20% blast furnace slag, and 20% glass wool. The alkaline activator consists of sodium hydroxide and sodium silicate with a SiO2/Na2O ratio of 1.5.
The addition of sucrose didn’t lead to any delay in the setting time. While both sodium chloride and sodium borate delayed the setting time, increased the flowability, and improved the extrudability of the mixtures. Out of the viscosity modifying admixtures, attapulgite had the most positive effect as it improved the thixotropic behaviour of all the mixes and improved the yield stress. Sodium carboxymethyl starch and cellulose almost had the same effect on the all the mixtures as they both improved extrudability and the yield stress in a similar trend. Xanthan gum was considered a poor admixture to be added to this mixture as it led to significant dryness which led to loss in fluidity which affected both flowability and extrudability.
The most suitable design was the B3lbA0.5 mixture which has 3% borax and 0.5% attapulgite and a liquid to binder ratio of 0.45. It had an initial setting time of more than 8 hours, flowability was maintained for 2.5 hours and it remained extrudable for 2 hours. Its yield stress ranged from 1360.87 Pa to 1977.78
Pa. After 28 days, it had a flexural and compressive strength of 5.72 MPa and 30.42 MPa respectively. Moreover, the flexural and compressive strength of 3D printed prisms was tested in the vertical and horizontal direction on the 7th day. The results showed that the flexural strength of the 3D printed prism tested in the vertical direction was the highest compared to the cast prism and 3D printed prism tested in the horizontal direction by 0.29% and 0.43% respectively. While the compressive strength of 3D printed cube tested in the horizontal direction was higher than both the cast cube and the 3D printed cube tested in the vertical direction by 0.24% and 0.32% respectively. The strength results indicate that
the mixture could be used in structural applications. The tensile bond strength was 0.69 MPa and the failure occurred near the glue, meaning that the interface is stronger than the material. In 3D printed concrete, usually the interface between the layers is the weakest point, however due to the addition of attapulgite the interface was strengthened. This explains why the compressive strength in the horizontal direction was higher than the vertical one as the interface was no longer the weak link in the specimen. The mixture’s buildability wasn’t optimal as the bottom layer had a slight big deformation and it had only
ten layers of printed material, but that is reasonable since the open time is 2 hours. This means that for big spans this mixture can be useful as it can be printed for a long period of time and its yield stress will develop.
In order for the mixture to be printable, it had to have an initial setting time higher than three hours, a spread diameter between 150-200mm and a yield stress above 1400 Pa. The above mentioned results
fall within the range of values that were needed to have a printable mixture. However, after printing the mixture it was concluded that a spread diameter of 150-180mm and a yield stress above 1600 Pa would be more suitable for both printability and buildability, because when the spread diameter was above
180mm and the yield stress was lower than 1600 Pa, the mixture’s viscosity impacted the buildability.