ASU Electronic Theses and Dissertations
This collection includes most of the ASU Theses and Dissertations from 2011 to present. ASU Theses and Dissertations are available in downloadable PDF format; however, a small percentage of items are under embargo. Information about the dissertations/theses includes degree information, committee members, an abstract, supporting data or media.
In addition to the electronic theses found in the ASU Digital Repository, ASU Theses and Dissertations can be found in the ASU Library Catalog.
Dissertations and Theses granted by Arizona State University are archived and made available through a joint effort of the ASU Graduate College and the ASU Libraries. For more information or questions about this collection contact or visit the Digital Repository ETD Library Guide or contact the ASU Graduate College at gradformat@asu.edu.
Displaying 1 - 2 of 2
Filtering by
- All Subjects: Computer Science
Description
With the ever-increasing demand for high-end services, technological companies have been forced to operate on high performance servers. In addition to the customer services, the company's internal need to store and manage huge amounts of data has also increased their need to invest in High Density Data Centers. As a result, the performance to size of the data center has increased tremendously. Most of the consumed power by the servers is emitted as heat. In a High Density Data Center, the power per floor space area is higher compared to the regular data center. Hence the thermal management of this type of data center is relatively complicated.
Because of the very high power emission in a smaller containment, improper maintenance can result in failure of the data center operation in a shorter period. Hence the response time of the cooler to the temperature rise of the servers is very critical. Any delay in response will constantly lead to increased temperature and hence the server's failure.
In this paper, the significance of this delay time is understood by performing CFD simulation on different variants of High Density Modules using ANSYS Fluent. It was found out that the delay was becoming longer as the size of the data center increases. But the overload temperature, ie. the temperature rise beyond the set-point became lower with the increase in data center size. The results were common for both the single-row and the double-row model. The causes of the increased delay are accounted and explained in detail manner in this paper.
Because of the very high power emission in a smaller containment, improper maintenance can result in failure of the data center operation in a shorter period. Hence the response time of the cooler to the temperature rise of the servers is very critical. Any delay in response will constantly lead to increased temperature and hence the server's failure.
In this paper, the significance of this delay time is understood by performing CFD simulation on different variants of High Density Modules using ANSYS Fluent. It was found out that the delay was becoming longer as the size of the data center increases. But the overload temperature, ie. the temperature rise beyond the set-point became lower with the increase in data center size. The results were common for both the single-row and the double-row model. The causes of the increased delay are accounted and explained in detail manner in this paper.
ContributorsRamaraj, Dinesh Balaji (Author) / Gupta, Sandeep (Thesis advisor) / Hermann, Marcus (Committee member) / Huang, Huei-Ping (Committee member) / Arizona State University (Publisher)
Created2015
Description
This dissertation introduces FARCOM (Fortran Adaptive Refiner for Cartesian Orthogonal Meshes), a new general library for adaptive mesh refinement (AMR) based on an unstructured hexahedral mesh framework. As a result of the underlying unstructured formulation, the refinement and coarsening operators of the library operate on a single-cell basis and perform in-situ replacement of old mesh elements. This approach allows for h-refinement without the memory and computational expense of calculating masked coarse grid cells, as is done in traditional patch-based AMR approaches, and enables unstructured flow solvers to have access to the automated domain generation capabilities usually only found in tree AMR formulations.
The library is written to let the user determine where to refine and coarsen through custom refinement selector functions for static mesh generation and dynamic mesh refinement, and can handle smooth fields (such as level sets) or localized markers (e.g. density gradients). The library was parallelized with the use of the Zoltan graph-partitioning library, which provides interfaces to both a graph partitioner (PT-Scotch) and a partitioner based on Hilbert space-filling curves. The partitioned adjacency graph, mesh data, and solution variable data is then packed and distributed across all MPI ranks in the simulation, which then regenerate the mesh, generate domain decomposition ghost cells, and create communication caches.
Scalability runs were performed using a Leveque wave propagation scheme for solving the Euler equations. The results of simulations on up to 1536 cores indicate that the parallel performance is highly dependent on the graph partitioner being used, and differences between the partitioners were analyzed. FARCOM is found to have better performance if each MPI rank has more than 60,000 cells.
The library is written to let the user determine where to refine and coarsen through custom refinement selector functions for static mesh generation and dynamic mesh refinement, and can handle smooth fields (such as level sets) or localized markers (e.g. density gradients). The library was parallelized with the use of the Zoltan graph-partitioning library, which provides interfaces to both a graph partitioner (PT-Scotch) and a partitioner based on Hilbert space-filling curves. The partitioned adjacency graph, mesh data, and solution variable data is then packed and distributed across all MPI ranks in the simulation, which then regenerate the mesh, generate domain decomposition ghost cells, and create communication caches.
Scalability runs were performed using a Leveque wave propagation scheme for solving the Euler equations. The results of simulations on up to 1536 cores indicate that the parallel performance is highly dependent on the graph partitioner being used, and differences between the partitioners were analyzed. FARCOM is found to have better performance if each MPI rank has more than 60,000 cells.
ContributorsBallesteros, Carlos Alberto (Author) / Herrmann, Marcus (Thesis advisor) / Adrian, Ronald (Committee member) / Chen, Kangping (Committee member) / Huang, Huei-Ping (Committee member) / Lopez, Juan (Committee member) / Arizona State University (Publisher)
Created2019