Plasma spray for depositing thermal barrier coatings features large distributions of particle states that result in significant variations in coating quality. These variations arise from distributions of particle sizes, large spatial gradients of plasma thermal-fluid fields, and temporal variations of the arc and jet. This paper describes a simplified approach for studying how particle state distributions are influenced by torch conditions and powder distributions, and the implications for deposition rate monitoring and control. The approach combines a simplified jet model with a more detailed particle model. The important fluid-thermal spatial gradients in the plasma jet are captured using a three zone model: a core region, modeled by growth of a turbulent shear layer around a laminar core, a transition region and a similarity region. Plasma-particle momentum and thermal interactions, particle phase transitions, internal particle temperature gradients, and collapse of in-flight hollow particles have been modeled using a multi-lumped particle model. Effects of distributions of particle size, particle morphology, injection velocity, and carrier gas flow were studied for YSZ spray in an Ar-He plasma. The results provide guidance on sensor design and operation and on approaches for plume location control.

This paper investigates the need and requirements for developing an improved sensing scheme for plasma spray, which would aid in compensating for the large variations in deposition rate. Experimental evaluation of a variety of current sensors reveals poor correlation with deposition rate. Important issues that explain these results include reconsidering what particle state characteristics should be measured and determining what sensing characteristics are required to achieve better correlation with deposited mass. A new particle flux sensor is proposed that meets these challenges is evaluated.