Date of Award

12-2015

Level of Access Assigned by Author

Campus-Only Thesis

Degree Name

Doctor of Philosophy (PhD)

Department

Chemical Engineering

Advisor

Douglas W. Bousfield

Second Committee Member

Mehdi Tajvidi

Third Committee Member

Douglas J. Gardner

Additional Committee Members

Jerker Jader

Albert Co

Abstract

Cellulose nanofibers (CNF) are an exciting new renewable material produced from wood fibers. Even at low solids content, CNF-water suspensions have a complex rheology that includes extreme shear-thinning as well as viscoelastic properties and a yield stress. In the rheology of CNF suspensions, the measurement method may influence the results due to wall-slippage, but it is unclear how the behavior near walls influences the measurement method and what process equipment can manipulate this material.

Parallel-plate and vane geometries were utilized to compare yielding and flow of CNF suspensions obtained by steady-state shear and oscillatory rheological measurements. Four different methods were compared as techniques to obtain a yield stress. The results are compared to pressure driven flow in a tube. Cone and plate geometries were found to lead to sample ejection at low shear rates: floc-floc interactions can explain this ejection. The suspensions violated the Cox-Merz rule in a significant manner as a sign of containing weak gel structures and the formation of a water-rich layer near the solid boundaries. For suspensions lower than 3% solids, the yield stress measured with different procedures were within 20% of each other, but for high solids suspensions, differences among the methods could be as large as 100%; the water-rich layer formation likely is the cause of these results. Oscillatory methods are suggested as a technique to obtain yield stress values. The pressure driven flow results were consistent with the power-law line fitted to the parallel-plate geometry data from steady shear.

The capability of the extrusion process was investigated for pumping CNF suspensions through different dies. The extrusion process resulted in acceptable pumping rates which was in good agreement with the mathematical model. However, attributable to the extreme shear-thinning behavior of CNF, the pressure counter-flow dominates the drag flow along the screw channel and does not allow the material to be forced through the die. The extruder was replaced by a progressive cavity pump which showed great capability for CNF casting. A mathematical model was used to predict the pump curves for different materials. The model satisfied the experimental data of discharge pressure vs. flow rate for different Newtonian fluids in the presence and absence of turbulent flow across the leakage channels between the pump cavities. The model worked well for CNF suspensions with lower yield stress as well.

CNF was added to paperboard coatings and the change in mechanical stiffness was measured. Coating slurries were made under different formulations for latex-CNF contents. It appears by keeping both latex and CNF contents high enough, a two dimensional network can be formed that increases the stiffness of the sample. Two different layers of coatings were applied on the paper sheets and the changes after each layer were investigated. However, due to flocculation of CNF under shear flowing through the gap between the coating rod and the substrate, chunks of CNF sat on the coated surface and the fibers did not flow into a uniform coating layer. As a result, CNF presence in the coatings slightly modified stiffness of the substrate (~20%). Pure CNF containing CMC was also applied on paperboard. An increase of 14% in bending stiffness was observed for coated samples compared to the substrate. A finite element model was used to examine the effect of change in rheology of coatings by adding CNF on blade coating process. Owing to the increase in shear viscosity of CNF-containing coatings, the blade force to maintain a certain coat weight increased linearly with the increase in the infinite shear viscosity of the slurries.

Share