In this paper we propose an indirect low-dimension design representation to enhance topology optimization capabilities. Established topology optimization methods, such as the Solid Isotropic Material with Penalization (SIMP) method, can solve large-scale topology optimization problems efficiently, but only for certain problem formulation types (e.g., those that are amenable to efficient sensitivity calculations). The aim of the study presented in this paper is to overcome some of these challenges by taking a complementary approach: achieving efficient solution via targeted design representation dimension reduction, enabling the tractable solution of a wider range of problems (e.g., those where sensitivities are expensive or unavailable). A new data-driven design representation is proposed that uses an augmented Variational Autoencoder (VAE) to encode 2D topologies into a lower-dimensional latent space, and to decode samples from this space back into 2D topologies. Optimization is then performed in the latent space as opposed to the image space. Established topology optimization methods are used here as a tool to generate a data set for training by changing problem conditions systematically. The data is generated using problem formulations that are solvable by SIMP, and are related to (but distinct from) the desired design problem. We further introduce augmentations to the VAE formulation to reduce unrealistic scattering of small material clusters during topology generation, while ensuring diversity of the generated topologies. We compare computational expense for solving a heat conduction design problem (with respect to the latent design variables) using different optimization algorithms. The new non-dominated points obtained via the VAE representation were found and compared with the known attainable set, indicating that use of this new design representation can simultaneously improve computational efficiency and solution quality.

No abstract

No abstract

No abstract

No abstract

Templating with self‐assembled colloidal crystals has enabled fabrication of porous materials with potential for sensing, heat transfer, and energy storage applications. Herein, electrical, thermal, and mechanical properties of templated electroplated metallic copper inverse opals (CIO) containing close‐packed 500 nm pores are evaluated via a four‐point probe, eddy current, lap shear measurements, thermal cycling, and finite‐element analysis (FEA). CIOs are found to have an electrical conductivity of ≈4 × 106 S m−1, about 15% of pure copper, about the expected value based on the measured electroplated copper conductivity and the CIO pore structure. The good electrical and thermal properties of this structure, coupled with its connected porosity, make it attractive for two‐phase cooling applications. However, the thermal reliability of this structure raises questions about its integrity. It is found that this structure fails around 150 cycles when cycled between −40 and 200 °C at 10 °C min−1. Scanning electron micrograph (SEM) images of failed samples show that the failure plane lies within the CIO and generally at the thinnest interconnects in the structure, which agrees with FEA of the CIO structure, which shows stress concentrations at thin regions of the interconnects in the structure.

The heat transfer and fluid flow performance of a hybrid jet plus multipass microchannel heat sink in two-phase operation is evaluated for the cooling of a single large area, 3.61 cm2, heat source. The two-layer branching microchannel heat sink is evaluated using HFE-7100 as the coolant at three inlet volumetric flow rates of 150, 300, and 450 ml/min. The boiling performance is highest for the flow rate of 450 ml/min with the maximum heat flux value of 174 W/cm2. Critical heat flux (CHF) was observed at two of the tested flow rates, 150 and 300 ml/min, before reaching the maximum operating temperature for the serpentine heater. At 450 ml/min, the heater reached the maximum allowable temperature prior to observing CHF. The maximum pressure drop for the heat sink is 34.1 kPa at a heat flux of 164 W/cm2. Further, the peak heat transfer coefficient value of the heat sink is 28,700 W/m2 K at a heat flux value of 174 W/cm2 and a flow rate of 450 ml/min. Finally, a validated correlation of the single device cooler is presented that predicts heat transfer performance and can be utilized in the design of multidevice coolers.

A new method for optimizing the layout of device-routing systems is presented. Gradient-based topology optimization techniques are used to simultaneously optimize both device locations and routing paths of device interconnects. In addition to geometric considerations, this method supports optimization based on system behavior by including physics-based objectives and constraints. Multiple physics domains are modeled using lumped parameter and finite element models. A geometric projection for devices of arbitrary polygonal shape is developed along with sensitivity analysis. Two thermal-fluid systems are optimized to demonstrate the use of this method.