Optimization for cooling

Patent metadata

Field	Value
Publication number	US-9740801-B2
Application number	US-201214342472-A
Country	US
Kind code	B2
Filing date	Aug 29, 2012
Priority date	Sep 3, 2011
Publication date	Aug 22, 2017
Grant date	Aug 22, 2017

Abstract

A design optimization system ( 100 ) and a method for achieving design optimization for cooling are described herein. According to an implementation, the method includes obtaining an inlet value of at least one flow parameter at a small geometric length scale and determining an outlet value of the at least one flow parameter at the small geometric length scale based on the inlet value. Further, a flow behavior is modeled based on the inlet and outlet values of the at least one flow parameter, and based on the modeled flow behavior an optimized design for cooling is ascertained.

First claim

We claim:

1. A computer implemented method for designing a data center optimized for cooling of components in the data center, the method comprising:
generating, by a processor, a small geometric length scale model of the data center, wherein the small geometric length scale model includes geometrical specifications of a small component of the data center, the small component comprising at least one of a processor chip and vent-tile;
ascertaining, by the processor, for the small geometric length scale model, outlet values of at least one flow parameter associated with a cooling medium, based on input inlet values of the at least one flow parameter and a fluid dynamics simulation of circulation of the cooling medium in the small geometric length scale model;
determining, by the processor, an empirical model indicative of fluid behaviour of the cooling medium in the small geometric length scale model using a data-based modeling framework, based on the input inlet values and the ascertained outlet values of the at least one flow parameter;
generating, by the processor, a full-scale physics-based model based on a full-scale model of the data center, the full-scale physics based model being indicative of geometry of a full-scale of the components of the data center without the geometrical specifications of the small component;
integrating, by the processor, the empirical model with the full-scale physics-based model to obtain an integrated physics-based model for simulating fluid flow behaviour of the cooling medium in the data center;
determining, by the processor, full-scale outlet values of the at least one flow parameter of the cooling medium for the data center, based on inlet values of the at least one flow parameter in the data center and the integrated physics-based model, for optimizing the data center for cooling of the components; and
wherein integrating the empirical model further comprises validating the empirical model comprising:
selecting the at least one correlation;
providing test data as input to the at least one correlation, wherein the test data comprises inlet values of the at least one flow parameter;
determining test outlet values of the at least one flow parameter, based on the at least one correlation and the test data; and
comparing the test outlet values with ascertained outlet values of the at least one flow parameter to validate the empirical model.

2. The method as claimed in claim 1 , wherein the ascertaining the outlet values comprises simulating fluid flow behaviour in the small geometric length scale model, using a physics-based model.

3. The method as claimed in claim 2 , wherein the physics-based model is a computational fluid dynamics (CFD) based model.

4. The method as claimed in claim 1 , wherein the determining the empirical model comprises ascertaining at least one correlation and one or more coefficients of the at least one correlation, based on the inlet values and the outlet values of the at least one flow parameter.

5. The method as claimed in claim 1 , wherein the data-based modeling framework is an artificial neural network (ANN)-based model.

6. A design optimization system for designing a data center, the design optimization system comprising:
a processor; and
a memory coupled to the processor, the memory comprising,
a determination module configured to,
generate a small geometric length scale model of the data center, wherein the small geometric length scale model includes geometrical specifications of a small component of the data center, the small component comprising at least one of a processor chip and vent-tile;
ascertain, for the small geometric length scale model, outlet values of at least one flow parameter associated with a cooling medium, based on input inlet values of the at least one flow parameter and a fluid dynamics simulation of circulation of the cooling medium in the small geometric length scale model; a modeling module configured to:
determine an empirical model indicative of fluid behaviour of the cooling medium in the small geometric length scale model using a data-based modeling framework, based on the input inlet values and the ascertained outlet values of the at least one flow parameter;
generate a full-scale physics-based model based on a full-scale model of the data center, the full-scale physics based model being indicative of geometry of a full-scale of the components of the data center without the geometrical specifications of the small component;
integrate the empirical model with the full-scale physics-based model to obtain an integrated physics-based model for simulating fluid flow behaviour of the cooling medium in the data center;
determine full-scale outlet values of the at least one flow parameter of the cooling medium for the data center, based on inlet values of the at least one flow parameter in the data center and the integrated physics-based model, for optimizing the data center for cooling of the component;
select a primary flow parameter from the at least one flow parameter; and
determine the empirical model based on the selected primary flow parameter using the data-based modeling framework.

7. The design optimization system as claimed in claim 6 , wherein the determination module is configured to:
obtain the inlet values of the at least one flow parameter in the small geometric length scale model of a data center; and
compute the outlet values of the at least one flow parameter in the small geometric length scale model of a data center based on a computational fluid dynamics (CFD) based model.

8. The design optimization system as claimed in claim 6 , further comprising a validation module configured to validate an accuracy of the empirical model determined by the modeling module, wherein the validation module is configured to:
determine test outlet values of the at least one flow parameter, based on test data and the empirical model;
compare the test outlet values with available outlet values of the at least one flow parameter computed by the modeling module; and
validate the accuracy of the empirical model based on the comparison.

9. A non-transitory computer-readable medium having a set of computer readable instructions that, when executed, perform acts comprising:
generating a small geometric length scale model of a data center, wherein the small geometric length scale model includes geometrical specifications of a small component of the data center, the small component comprising at least one of a processor chip and vent-tile;
ascertaining for the small geometric length scale model, outlet values of at least one flow parameter associated with a cooling medium, based on input inlet values of the at least one flow parameter and a fluid dynamics simulation of circulation of the cooling medium in the small geometric length scale model;
determining an empirical model indicative of fluid behaviour of the cooling medium in the small geometric length scale model using a data-based modeling framework, based on the input inlet values and the ascertained outlet values of the at least one flow parameter;
generating a full-scale physics-based model based on a full-scale model of the data center, the full-scale physics based model being indicative of geometry of a full-scale of the components of the data center without the geometrical specifications of the small component;
integrating the empirical model with the full-scale physics-based model to obtain an integrated physics-based model for simulating fluid flow behaviour of the cooling medium in the data center;
determining full-scale outlet values of the at least one flow parameter of the cooling medium for the data center, based on inlet values of the at least one flow

Assignees

• Singh Umesh
• Singh Amarendra Kumar
• Sivasubramaniam Anand
• Tata Consultancy Services Ltd

Inventors

• Singh Umesh
• Singh Amarendra Kumar
• Sivasubramaniam Anand

Classifications

• G06F30/20

  Design optimisation, verification or simulation (optimisation, verification or simulation of circuit designs G06F30/30) · CPC title

• G06F1/20 Primary

  Cooling means · CPC title

• G06F30/13 Primary

  Architectural design, e.g. computer-aided architectural design [CAAD] related to design of buildings, bridges, landscapes, production plants or roads · CPC title

• G06F2113/02

  Data centres · CPC title

• G06F17/5004 Primary

  Physics · mapped topic