The rapid global increase in energy demand for data centres requires the continuous improvement of cooling solutions and techniques implemented. In a standard data centre approximately a third of the total power consumption can be attributed to the cooling infrastructure, resulting in raised power usage effectiveness (PUE) values. The main culprits of raised PUE are outdated legacy (air cooled) data centres, where only low grade waste heat is available and its capture and re-use remains a challenge. This study investigates numerically the potential for energy recuperation by a server-level internal layout optimization for a hybrid air/liquid cooled server. The study uses multi-objective genetic algorithms (MOGA) and entropy generation minimization (EGM) techniques to incorporate the multiple objectives involved in solving this problem, and examines the cooling performance and waste heat recovery potential. In order to evaluate the potential of the waste heat recovery the term ̇, is introduced in this study. Effect of modified server component layout on pressure drop and maximum outlet temperature were of main interest, due to their role in fan power usage and energy recuperation potential. The base unmodified configuration CFD models were validated through experimental pressure measurements conducted on a real server blade module. The research concluded that a basic server layout optimization such as changing the
Numerical models of manufacturing processes are useful and provide insight for the practitioner, however, model verification and validation are a prerequisite for expedient application. This paper details the code-to-code verification of a thermal numerical model for the Bridgman solidification process of alloys in a 2-dimensional axisymmetric domain, against an established commercial code (ANSYS Fluent); the work is considered a confidence building step in model development. A grid sensitivity analysis is carried out to establish grid independence, this is followed by simulations of two transient solidification scenarios: pulling rate step change and ramp input; the results of which are compared and discussed. Good conformity of results is achieved, hence the non-commercial model is codeto-code verified; in addition, the ability of the non-commercial model to deal with radial heat flow is demonstrated. The introduction of front tracking to model the macroscopic growth of dendritic mush and the region of undercooled liquid is identified as the next step in model development. http://mc.manuscriptcentral.com/hff
In a recent study, the present authors reported an analysis of a transient process of directional solidification of TiAl alloys in the absence of convection [Battaglioli et al., 2017]. The adopted front tracking model, also coupled with indirect methods for predicting columnar to equiaxed transition (CET), showed how the development of different grain regions, namely axial columnar, radial columnar and equiaxed, depends on process parameters such as temperature distribution and applied cooling rates, as well as on properties such as the degree of inoculation of the melt and nucleation undercooling required for equiaxed growth. In this paper, the previous front tracking model is significantly developed, by including the solution of Navier-Stokes equations in order to predict thermal convection in the liquid region as well as in the columnar mush (treated as an isotropic porous medium). This improvement is introduced with the aim to investigate TiAl alloys solidification under different gravity conditions. Accordingly, the simulation setup employed in the study reproduces the one used in experimental campaigns carried out on the MAXUS 9 sounding rocket (microgravity) and on ESA's Large Diameter Centrifuge (hypergravity), within the framework of ESA's GRADECET (Gravity Dependence of CET in TiAl alloys) project. The ability of the model in predicting thermal convection is demonstrated by considering several case studies. Results show that the evolution of fluid flow patterns in the samples depends on the external forces considered, combined to the transient change of axial and radial temperature gradients that occur during the solidification process. A parametric study of the directional solidification process on a centrifuge is performed by changing the value of the centrifuge arm and rotation rate and results are compared to predictions derived from nondimensional considerations.
A transient 2D axisymmetric numerical model for the Bridgman solidification process for a stationary furnace and moving sample is presented. The model is able to predict the evolution of temperature and solid fraction of binary alloys in cases where buoyancy induced convection is negligible, such as in microgravity conditions. A dimensionless form of the governing equations was derived in order to identify the dimensionless parameters that characterize the process, those being the Stefan, Péclet, and Biot numbers. The problem was solved using a finite volume method and an explicit time stepping scheme. To test the efficacy of the model, simulated results were compared with experimental data from the literature and acceptable agreement was obtained. Finally, a parametric analysis was performed for understanding the influence of the process parameters on solidification. One key feature of this study was the inclusion of a term describing the advection of latent heat due to the translation of the mushy zone with varying solid fraction. This thermal transport mechanism was shown to be significant, since its magnitude was comparable to the advection of sensible heat. It was also found that when small Biot numbers were due to low values of the heat transfer coefficients at the surface of the sample, rather than to small sample radii, advective mechanisms were enhanced resulting in more convex shapes of the liquidus isotherm. This highlighted the importance of considering both axial and radial heat fluxes when describing the process.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.