Use this URL to cite or link to this record in EThOS:
Title: Energy based dissolution simulation using smoothed particle hydrodynamic sampling
Author: Jiang, Min
ISNI:       0000 0004 5993 0966
Awarding Body: Bournemouth University
Current Institution: Bournemouth University
Date of Award: 2016
Availability of Full Text:
Access from EThOS:
Access from Institution:
Fluid simulation plays an important role in Computer Graphics and has wide applications in film and games. The desire for an improved physically-based fluid simulation solver has grown hand in hand with the advances made in Computer Graphics. Interesting fluid behaviours emerge when solid objects are added to a simulation: when fluid and solid make contact, they do not only have a physical interaction (e.g., buoyancy), but also a chemical reaction (e.g., dissolution) under the right conditions. Dissolution is one of the most common natural phenomena which is an important visual effect in fluid simulation. However this phe- nomenon is difficult to simulate due to the complexity of the behaviour and there are only few techniques available. A novel unified particle-based method for approximating chemical dis- solution is introduced in this thesis which is fast, predictable and visually plausible. The dissolution algorithm is derived using chemical Collision Theory and integrated into a Smoothed Particle Hydrodynamics (SPH) framework. The Collision Theory of chemistry is used as an analogy to the dissolution process modelling. Dissolution occurs when solute submerges into solvent. Physical laws govern the local excitation of so- lute particles based on the relative motion with solvent particles. When the local excitation energy exceeds a user specified threshold (activation energy), the particle will be dislodged from the solid. Unlike previous methods, this dissolution model ensures that the dissolution result is in- dependent of solute sampling resolution. A mathematical relationship is also established between the activation energy, the interfacial surface area, and the total dissolution time — allowing for intuitive artistic con- trol over the global dissolution rate. Applications of this method are demonstrated using a number of practical examples, including antacid pills dissolving in water and hydraulic erosion of non-homogeneous ter- rains. Both solutes and solvents are represented by particles, and the dis- tribution of the solute particles greatly affects the plausibility of the dissolution simulation. An even but stochastic distribution of particles on both the surface and within the volume of the solute is essential for a good visual simulation of the dynamic process of dissolution. A new iterative particle-based sampling method derived from SPH is introduced in this thesis which can generate a range of blue noise pat- terns and is computationally efficient, controllable and has a variety of applications. This approach resolves many of the limitations of classic blue noise methods, such as the lack of controllability or varying the dis- tribution properties of the generated samples. Fast sampling is achieved in general dimensions for curves, surfaces and volumes. By varying a sin- gle parameter, the proposed method can generate a range of controllable blue noise samples with different distribution properties which are suit- able for various applications such as adaptive sampling and multi-class sampling. The SPH sampling approach is used for solute particle distribution which guarantees a predictable and smooth dissolution process thanks to the evenly distributed density and also gives the user control of the volume change during the phase transition. The proposed SPH sampling method achieves better visual effects compared with simple grid sampling and other blue noise sampling methods. Our energy based dissolution simulation with SPH sampled solute and solvent ensures that the dissolution behaviour is physically and chemi- cally plausible, while supporting features such as object separation and sharp feature rounding. The simulation is parallelized per particle on a GPU to enhance the performance.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available