| dc.description.abstract | In recent years, Cold Spray technology has proven to be a promising method of powder deposition in surface coating and additive manufacturing applications. This process is done without melting the particles prior to deposition; therefore, a wide range of materials can be deposited onto a substrate through the powder’s kinetic energy. In contrast to traditional coating methods, in CS the particle temperature remains below its melting point. As a result, the adverse effects of melting and other temperature-related defects are avoided in CS. The kinetic energy needed for successful bonding is characterized by a powder’s critical velocity that is dependent on the properties of the powder material. The quality of the cold sprayed coatings depends strongly on the adhesion strength at the particle-substrate interface. Predicting the adhesion strength thus plays a crucial role in optimizing the process parameters of CS process to achieve desired surface coating qualities.
Utilizing numerical approaches, to study the occurrence of bonding in CS and observe the material jetting phenomena due to bonding, is more convenient than using empirical approaches. Experimental observation of particle bonding in CS is inherently very difficult due to the extremely small time and length scales at which particle bonding happens in CS. Numerical modeling has thus been an indispensable tool in the study of CS. Most numerical studies on CS have focused on using the traditional mesh-based FEM, which often face limitations when modelling the extreme plastic deformation occurring during particle impact. In comparison, meshless methods are proven to perform significantly better as they resolve the issue of mesh distortion in mesh-based methods Furthermore, majority of existing models are only able to model the impact and bonding processes and very few methods exist that can predict the bonding strength accurately.
In the present work, a computational method is proposed for modeling bonding of powder particles in cold spray capable of predicting the adhesion strength. The method relies on a bonding model developed in previous research [1, 2] which is based on the commonly held view that bonding occurs due to large plastic strains occurring at extreme rates. This is achieved by introducing a strain-like history variable named bonding parameter and two material constants, the critical surface adhesion energy, and the critical surface adhesion energy rate. In this thesis, the bonding model is complemented with a semi-empirical evolution law for adhesion strength on bonding boundaries. The strength evolution model interacts with the bonding evolution model and is coupled with the bonding parameter.
The model is implemented numerically within a material point method (MPM) in a way that effectively eliminates spurious mesh dependence and captures complex phenomena such as jetting. The adhesion strength model proposed in this study utilizes the direct bonding model results and fundamentals from our previous study [1, 2]. The adhesion strength model proposed here is then used to predict adhesion strength and study the case of single particle impacting a substrate.
In doing so, the distribution of adhesion strength will be shown through the contact region. The model parameters are also discussed, and it will be shown that how the change of these parameters will affect the adhesion strength. In previous studies the average adhesion strength values were achieved from experiments. The values for the average adhesion strength will also be highlighted in the present numerical study.
The simulations were performed for pure Aluminum (99.7%) particles impacting Al substrate. The chosen material is similar to previous research works and studies. Further into the thesis the adhesion strength model will be discussed, and corresponding results will be discussed afterwards. According to the results obtained, the average adhesion strength for a single pure Al powder (14 micro-meter diameter) impacting an Al substrate at 810 m/s (which is approximately the critical velocity) is predicted to be 42.16 MPa. The corresponding critical adhesion energy and energy rate used were 1500 𝐽/𝑚2 and 2 𝐽/𝑚2𝑠 (these will be discussed later in the thesis), which are also approximately near the calibrated values. Results are very encouraging and exhibit desirable agreement with known experimental data such as critical bonding velocity, adhesion strength, and deformed particle/substrate shape. | en |
