Benchmarking of adjoint sensitivity-based optimization techniques using a vehicle ride case study

Several studies on multibody dynamics optimization have been conducted. One important limitation of these studies is their computational e?ciency, especially when optimizing a complex system’s performance. The co-authors developed a very e?cient optimization technique based on an adjoint sensitivity analysis methodology. The scope of this article is to validate this technique by conducting a benchmark analysis against some of the most popular optimization methods, including gradient-based optimization using finite differences, design of experiment using optimal Latin hypercube, and design of experiment using full factorial design matrix. A vehicle system is used as a case study for optimizing its ride comfort.     

5‐Hydroxyl Hydantoins via One‐Pot Microwave‐Assisted Air Oxidation of Ugi Products

The novel and one‐pot synthesis of 5‐hydroxyl hydantoins from Ugi five‐component condensation procedures is reported. The transformation was promoted via a microwave heating method and provided 5‐hydroxyl hydantoins in workable yields, representing a direct approach to this synthetically challenging and biologically interesting pharmacophore.     

Magnetically induced Freedericksz transition and relaxation phenomena in nematic liquid crystals doped with azo-dyes

A static and dynamic investigation was performed on liquid crystal cells containing pure nematics and nematics doped with an azo-dye (Methyl Orange). It was found that the critical field for magnetic Freedericksz transition was decreased in samples containing the “trans” isomer and increased in those containing the “cis  ” isomer. Changes in the relaxation time _τA${\tau }_A$, _τB${\tau }_B$ intervening when switching on/off the magnetic field were also noticed. A theoretical model was elaborated to explain these phenomena.     

A class of supported membranes: formation of fluid phospholipid bilayers on photonic band gap colloidal crystals

We report the formation of a new class of supported membranes consisting of a fluid phospholipid bilayer coupled directly to a broadly tunable colloidal crystal with a well-defined photonic band gap. For nanoscale colloidal crystals exhibiting a band gap at the optical frequencies, substrate-induced vesicle fusion gives rise to a surface bilayer riding onto the crystal surface. The bilayer is two-dimensionally continuous, spanning multiple beads with lateral mobilities which reflect the coupling between the bilayer topography and the curvature of the supporting colloidal surface. In contrast, the spreading of vesicles on micrometer scale colloidal crystals results in the formation of bilayers wrapping individual colloidal beads. We show that simple UV photolithography of colloidal crystals produces binary patterns of crystal wettabilities, photonic stopbands, and corresponding patterns of lipid mono- and bilayer morphologies. We envisage that these approaches will be exploitable for the development of optical transduction assays and microarrays for many membrane-mediated processes, including transport and receptor-ligand interactions.     

Measurement and analysis of the dextran partition coefficient in sucrose crystallization

The effect of crystallization conditions on the dextran partition coefficient between impure syrup and sugar crystal has been investigated in a batch crystallizer. The crystallizer is operated isothermally at temperatures of 30, 40, and 50 °C, at constant relative supersaturations of 0.05, 0.07, and 0.09, and with mother liquor dextran concentrations of 1000 and 2000 ppm/Brix. The dextran content has been determined by the CSR method. A 1:1 mass ratio of high-fraction dextran (approximately 250,000 Da) and low-fraction dextran (60,000-90,000 Da) is used to represent a wide range of dextran contamination. It is seen that the dextran partition coefficient in sucrose crystallization increases with both increasing supersaturation and increasing crystallization temperature. However it appears that these are secondary effects, with the partition coefficient strongly correlating with crystal growth rate alone, despite the regressed data having large variations in temperature, mother liquor dextran content, and supersaturation. Dextran incorporation into the sugar crystal results from both dextran adsorption onto the crystal surface and mother liquor inclusions. The explanation for the variation in the dextran content in sugar crystal with respect to the growth rate is due to increased adsorption due to the higher surface roughness of crystals grown at high growth rates. Although the dextran concentration in the solution affects the dextran content in the crystal, it does not strongly affect the dextran partition coefficient.     

On the interplay between fast reaction and slow diffusion in the concrete carbonation process: a matched-asymptotics approach

A matched-asymptotics approach is proposed to show the occurrence of two distinct characteristic length scales in the carbonation process. The separation of these scales arises due to the strong competition between reaction and diffusion effects. We show that for sufficiently large times τ the width of the carbonated region is proportional to \(\sqrt{\tau}\), while the width of the reaction front is proportional to \(\tau^{\frac{p-1}{2(p+1)}}\) for carbonation-reaction rates with a power law structure like k[CO₂] ^p [Ca(OH)₂] ^q , where k>0 and p,q>1 and identify the proportionality coefficient asymptotically. We emphasize the occurrence of a water barrier in the reaction zone which may hinder the penetration of CO₂ by locally filling with water air parts of the pores. This non-linear effect may be one of the causes why a purely linear extrapolation of accelerated carbonation test results to natural carbonation settings is (even theoretically) not reasonable. Finally, we compare our asymptotic penetration law against measured penetration depths from Bune (Zum Karbonatisierungsbedingten Verlust der Dauerhaftigkeit von Außenbauteilen aus Stahlbeton, 1994). The novelty consists in the fact that the factor multiplying \(\sqrt{\tau}\) is now identified asymptotically by solving a non-linear system of ordinary differential equations, and hence, fitting arguments are not necessary to estimate its size. We offer an alternative to the (asymptotic) \(\sqrt{\tau}\) expression of the carbonation-front position obtained in Papadakis et al. (AIChE J. 35:1639, 1989).     

Experimental investigation into vortex structure and pressure drop across microcavities in 3D integrated electronics

Hydrodynamics in microcavities with cylindrical micropin fin arrays simulating a single layer of a water-cooled electronic chip stack is investigated experimentally. Both inline and staggered pin arrangements are investigated using pressure drop and microparticle image velocimetry (μPIV) measurements. The pressure drop across the cavity shows a flow transition at pin diameter–based Reynolds numbers (Re _d ) ~200. Instantaneous μPIV, performed using a pH-controlled high seeding density of tracer microspheres, helps visualize vortex structure unreported till date in microscale geometries. The post-transition flow field shows vortex shedding and flow impingement onto the pins explaining the pressure drop increase. The flow fluctuations start at the chip outlet and shift upstream with increasing Re _d . No fluctuations are observed for a cavity with pin height-to-diameter ratio h/d = 1 up to Re _d ~330; however, its pressure drop was higher than for a cavity with h/d = 2 due to pronounced influence of cavity walls.     

On the discontinuous Galerkin computation of N‐waves focusing

Sonic boom focusing phenomenon can be predicted using the solution to the nonlinear Tricomi equation which is a hybrid (hyperbolic‐elliptic) second‐order partial differential equation. In this paper, the hyperbolic conservation law form is derived, which is valid in the entire domain. In this manner, the presence of two regions where the equation behaves differently (hyperbolic in the upper and elliptic in the lower half‐plane) is avoided. On the upper boundary, a new mixed boundary condition for the acoustic pressure is employed. The discretization is carried out using a discontinuous Galerkin (DG) method combined with a Runge–Kutta total‐variation diminishing scheme. The results show the accuracy of DG methods to solve problems involving sharp gradients and discontinuities. Comparisons with analytical results for the linear case, and other numerical results using classical explicit and compact finite difference schemes and weighted essentially non‐oscillatory schemes are included. Copyright © 2010 John Wiley & Sons, Ltd.     

Effective macroscopic adhesive contact behavior induced by small surface roughness

In this paper we study a model contact problem involving adhesive elastic frictionless contact between rough surfaces. The problem's most notable feature is that it captures the phenomenon of depth-dependent-hysteresis (DDH) (e.g., see Kesari et al., 2010), which refers to the observation of different contact forces during the loading and unloading stages of a contact experiment. We specifically study contact between a rigid axi-symmetric punch and an elastic half-space. The roughness is represented as arbitrary periodic undulations in the punch's radial profile. These undulations induce multiple equilibrium contact regions between the bodies at each indentation-depth. Assuming that the system evolves so as to minimize its potential energy, we show that different equilibrium contact regions are selected during the loading and unloading stages at each indentation-depth, giving rise to DDH. When the period and amplitude of our model's roughness is reduced, we show that the evolution of the contact force and radius with the indentation-depth can be approximated with simpler curves, the effective macroscopic behavior, which we compute. Remarkably, the effective behavior depends solely on the amplitude and period of the model's roughness. The effective behavior is useful for estimating material properties from contact experiments displaying DDH. We show one such example here. Using the effective behavior for a particular roughness model (sinusoidal) we analyze the energy loss during a loading/unloading cycle, finding that roughness can toughen the interface. We also estimate the energy barriers between the different equilibrium contact regions at a fixed indentation-depth, which can be used to assess the importance of ambient energy fluctuations on DDH.     

Vortex formation process in gravity-driven starting jets

Experiments were conducted to investigate the vortex formation process in gravity-driven starting jets at three Reynolds numbers ( 3,528 and 4,716, where D is the nozzle diameter, the average discharging velocity and ν the kinematic viscosity). Planar laser induced fluorescence (PLIF) was used to visualize the flow while particle image velocimetry (PIV) was used to quantify the 2D velocity and vorticity fields. Vortex leapfrogging was found to appear at Re = 2,358, while pinch-off for the leading vortex was observed in the other two cases. Conditions for the vortex pinch-off appeared to be different from those found in the piston-driven starting jets. Although the leading vortex attained the maximum circulation and energy level very quickly after the jet was initiated, its detachment from the main stem may not be achieved necessarily. Attempt had been made to elaborate the different flow characteristics for respective cases from the related distribution of circulation and energy level via the measured velocity field by PIV.