Design and analysis of 2D one-way splitter waveguide based on topological photonics.

We present a new high-efficiency splitter waveguide design based on photonic topological insulators. The system's robust edge states allow electromagnetic waves to propagate in the 2D waveguide without backscattering, resulting in almost 100% transmission in the outputs. We also study resonating modes in the structure and show that introducing specific defects can create such modes. We consider four domains with rods of varying magneto-optical properties to provide edge modes in the system. By eliminating rows and columns of rods, we calculate the transmission at the outputs, revealing resonating modes in the middle of the structure with spatial symmetry. Our calculations indicate that the most promising resonating mode occurs when two rods and two columns are eliminated, with a quality factor Q = 1.02 × 106 at frequency f = 8.23 GHz and almost zero transmission at this frequency to the outputs. We further confirm our results using the transmission line resonator model as a semi-analytical model, which agrees well with our findings.

Bayesian Optimization of Booster Disinfection Scheduling in Water Distribution Networks.

Chlorine remains the most widely used disinfectant in drinking water treatment and distribution systems worldwide. To maintain a minimum residual throughout the distribution network, chlorine dosage needs to be regulated by optimizing the locations of chlorine boosters and their scheduling (i.e., chlorine injection rates). Such optimization can be computationally expensive since it requires numerous evaluations of water quality (WQ) simulation models. In recent years, Bayesian optimization (BO) has garnered considerable attention due to its efficiency in optimizing black-box functions in a wide range of applications. This study presents the first attempt to implement BO for the optimization of WQ in water distribution networks. The developed python-based framework couples BO with EPANET-MSX to optimize the scheduling of chlorine sources, while ensuring the delivery of water that satisfies water quality standards. Using Gaussian process regression to build the BO surrogate model, a comprehensive analysis was conducted to evaluate the performance of different BO methods. To that end, systematic testing of different acquisition functions, including the probability of improvement, expected improvement, upper confidence bound, and entropy search, in conjunction with different covariance kernels, including Matérn, squared-exponential, gamma-exponential, and rational quadratic, was conducted. Additionally, a thorough sensitivity analysis was performed to understand the influence of different BO parameters, including the number of initial points, covariance kernel length scale, and the level of exploration vs exploitation. The results revealed substantial variability in the performance of different BO methods and showed that the choice of the acquisition function has a more profound influence on the performance of BO than the covariance kernel.

Surrogate optimization of a lattice foot orthotic.

BACKGROUND: Additive manufacturing enables to print patient-specific Foot Orthotics (FOs). In FOs featuring lattice structures, the variation of the cell's dimensions provides a locally variable stiffness to meet the therapeutic needs of each patient. In an optimization problem, however, using explicit Finite Element (FE) simulation of lattice FOs with converged 3D elements is computationally prohibitive. This paper presents a framework to efficiently optimize the cell's dimensions of a honeycomb lattice FO for flat foot condition. METHODS: We built a surrogate based on shell elements whose mechanical properties were computed by the numerical homogenization technique. The model was submitted to a static pressure distribution of a flat foot and it predicted the displacement field for a given set of geometrical parameters of the honeycomb FO. This FE simulation was considered as a black-box and a derivative-free optimization solver was employed. The cost function was defined based on the difference between the predicted displacement by the model against a therapeutic target displacement. RESULTS: Using the homogenized model as a surrogate significantly accelerated the stiffness optimization of the lattice FO. The homogenized model could predict the displacement field 78 times faster than the explicit model. When 2000 evaluations were required in an optimization problem, the computational time was reduced from 34 days to 10 hours using the homogenized model rather than explicit model. Moreover, in the homogenized model, there was no need to re-create and re-mesh the insole's geometry in each iteration of the optimization. It was only required to update the effective properties. CONCLUSION: The presented homogenized model can be used as a surrogate within an optimization framework to customize cell's dimensions of honeycomb lattice FO in a computationally efficient manner.

Computationally efficient model to predict the deformations of a cellular foot orthotic.

BACKGROUND: Foot orthotics (FOs) are frequently prescribed to provide comfortable walking for patients. Finite element (FE) simulation and 3D printing pave the way to analyse, optimize and fabricate functionally graded lattice FOs where the local stiffness can vary to meet the therapeutic needs of each individual patient. Explicit FE modelling of lattice FOs with converged 3D solid elements is computationally prohibitive. This paper presents a more computationally efficient FE model of cellular FOs. METHOD: The presented FE model features shell elements whose mechanical properties were computed from the numerical homogenization technique. To verify the results, the predictions of the homogenized models were compared to the explicit model's predictions when the FO was under a static pressure distribution of a foot. To validate the results, the predictions were also compared with experimental measurements when the FO was under a vertical displacement at the medial longitudinal arch. RESULTS: The verification procedure showed that the homogenized model was 46 times faster than the explicit model, while their relative difference was less than 8% to predict the local minimum of out-of-plane displacement. The validation procedure showed that both models predicted the same contact force with a relative difference of less than 1%. The predicted force-displacement curves were also within a 90% confidence interval of the experimental measurements having a relative difference smaller than 10%. In this case, using the homogenized model reduced the computational time from 22 h to 22 min. CONCLUSION: The presented homogenized model can be therefore employed to speed up the FE simulation to predict the deformations of the cellular FOs.