Thermal Prediction of Data Centers during Power Failures

(23 votes, average 4.61 out of 5)
Category : Industrial Projects
Project Code : EDU-PRJ-PG-CFD-005
Project Subscribers : 12



Industrial Projects 7.0/10 Data Centers

With the growing adoption to modern technology by means of smartphones, personal computers & tablets there has been increased demand for the data center processing worldwide. For every handheld computing device attached to the cloud, a data center is processing the required information resulting in the growth of data centers and is expected to continue at the rate of 14% per year. To meet to these challenges the data center designers and operators are struggling to have optimum designs that can conserve energy.

Over the past several years Computational Fluid Dynamics (CFD) has proven to be a reliable tool to model data centers and is gaining popularity to study the performance of cooling of the aisles and racks in a data center. Through thermal mapping with the help of CFD one can expose problem areas (hotspots) that otherwise don’t show up in simple heat and flow balance calculations.

A data center is a dedicated space where companies can store and operate most of the information and communications technology (ICT) infrastructure that supports their business. These can be the servers and storage equipment that run application software and process and store data and content. For some companies this might be a simple cage or rack of equipment while for others can be a room housing a few or many cabinets, depending on the scale of their operations. The space can typically have a raised floor with cabling ducts running underneath to feed power to the cabinets and carry the cables that connect the cabinets together.

Different components of a data center are :

  • CRAC: Computer Room Air Conditioning unit is a device that monitors and maintains the temperature, air distribution and humidity in a network room or data center.
  • RACK: A rack is a standardized frame or enclosure for mounting multiple equipment modules.
  • PDUs and UPS: A PDU is a power distribution unit designed to fit into a server rack in either a vertical or horizontal position.
  • Perforated Tiles and Ceiling Grills: Perforated tiles are placed beneath computer systems to direct conditioned air directly to them.
  • Duct and Diffusers: Overhead ducting system to channelize air flow as desired 
  • Fans: To recirculate air flow and allow mixing of high and low temperature air

ASHRAE (American Society of Heating, Refrigerating, and Air-Conditioning Engineers) is an organization devoted to the advancement of indoor-environment-control technology in the heating, ventilation, and air conditioning (HVAC) industry. One of the most important functions of the organization is to promote research and development in efficient, environmentally friendly technologies. The temperature and humidity in a data center is controlled through air conditioning. ASHRAE's "Thermal Guidelines for Data Processing Environments" recommends a temperature range of 18 – 27 °C (64.4 – 80.6 °F), a dew point range of 5–15 °C (41–59 °F), and a maximum relative humidity of 60% for data center environments.

Continuous supply of electrical power to mission critical facilities is essential not only for uninterrupted operation of servers but also for providing continued cooling to maintain supply air temperatures within the acceptable range of 18 - 27 C (64.4 – 80.6 F) (as recommended by ASHRAE thermal guidelines. However, during power outage situations servers continue to operate by the power provided by uninterrupted power supply (UPS) units while the supply of cooling air is completely halted until alternate means of powering the cooling system are activated. During this time servers continue to generate heat and the server fans continue to circulate room air several times though the servers. This can result in sharp increase in the room air temperature to undesirable levels, which in turn can lead to automatic shutdown of servers and in some cases can even cause thermal damage to servers.

Often alternate means such as standby generators are provided to continue cooling operation during the power outage period. However it takes some time to restart these systems and resume normal cooling operation. It is crucial to understand the rate of temperature rise of the room air during this off cooling period and how long servers can sustain such a situation without automatic thermal shutdown.

Learning and Skill Sets Required

The execution of this project demands the following theoretical knowledge: 

  • Fluid Dynamics & Heat Transfer: CFD is based on Fluid Dynamics equations. Air flow movements are based on momentum and continuity equations. Heat transfer in data center is governed majorly by conduction and convection physics. It is very important for student to be comfortable with governing equations of fluid dynamics & heat transfer to do CFD simulation and interpretation of CFD results.
  • CFD Fundamentals: CFD Fundamentals is studying about the governing physical equations, and how the fluid flow & heat transfer problems are solved on computers using numerical methods, the backbone of any CFD code. Students who use commercial CFD software to complete their project works, often refer to user’s manual or tutorial guide, to make a choice of numerical technique, or turbulence model, or the type of boundary condition to apply. But most of the tutorial guides let them down, by not providing sufficient explanation of the theoretical background and justification for using a particular numerical scheme for the given problem. So, knowing the fundamentals of CFD becomes very important in the process of using CFD as a tool for design analysis.
  • Basics of Data center & its thermal management systemStudent needs to be familiar to the basic concepts of data center components and its thermal management system. Student needs to know & understand the meaning of various global terms used in data center cooling, viz. CRACs, RACKs, Cold Aisle, Hot Aisle, etc. Student should have knowledge of various data center floor layouts and air circulation patterns.

The execution of this project demands the following CFD software skills:

  • ANSYS ICEM CFD: This is pre-processing software that can be used for Mesh generation, which is nothing but a discrete representation of the geometry. Also, ANSYS ICEM CFD has advanced CAD/geometry readers and repair tools to allow the user to do the CAD cleanup work. In this project, ANSYS ICEM CFD will be used to create the 3D data center model and generate structured hexahedral mesh through Multi-block approach. So, the knowledge of the software GUI, CAD tools, and blocking tools in ANSYS ICEM CFD is important to take up this project.
  • ANSYS FLUENT: ANSYS FLUENT is a simulation tool that contains the broad physical modeling capabilities needed to model flow, turbulence, etc. This simulation software allows one to predict, the impact of fluid flows on the design or vice versa. Also, it has post-processing tools to extract simulation results and understand them. In this project, a steady state simulation replicating the normal working of a data center is to be carried out whose result is going to be used as the input to the an unsteady simulation replicating the power failure in data center scenario using ANSYS FLUENT. Later, the student would come up with a script to automate the whole FLUENT work. So, a knowledge on the software GUI, solver set-up, visualization techniques, and FLUENT customization are compulsory to work on this project.

The execution of this project requires knowledge of following programming languages:

  • ANSYS FLUENT TUI commands: ANSYS FLUENT software has a strong text user interface, which is alone capable enough to perform almost every possible operation one can think of in the software.
  • UDF (C Program): ANSYS FLUENT software uses C program to expand its own capabilities by providing users to create their own functions within the software. FLUENT provides standard Macros to the user to create custom functions as well as access FLUENT’s variables to perform various operations which are written in C language.
  • Scheme Programming within FLUENT: This is a programming language required for automating ANSYS Fluent. Scheme language can be used to do solver setup, running the simulation and for saving files with any kind of quantitative and qualitative results.

Necessary LearnCAx Courses

Students opting for this project will have to go through online courses suggested by the LearnCAx mentor. This is to learn the required skillsets before starting the project work. Following is the LearnCAx course required to execute this project.

The access to this course will be provided to the student as a part of the mentoring program and the validity of access exists till the project completion. 

Project and domain specific knowledge is not included in these course(s). The courses are designed to teach CFD methods in-general. The application of knowledge acquired through these course(s) to this specific project has to be done by student. During project execution stage, mentor will guide student to apply the course(s) knowledge for executing the project. Some of the project or domain specific training might not be directly covered in above course. Mentor will provide necessary guidance to students about from where they can acquire the project specific knowledge.

Who can take the project ?

  • Complexity Level -  8.0/10 (0-Low; 10-High)ANSYS ICEM CFD would be used to create the 3D datacenter model and to generate a hexahedral mesh of the same. ANSYS FLUENT would be used to run a steady-state simulation, involving basic fluid flow, turbulence and heat transfer, representing the scenario prior to power failure. The result data from this simulation would be used to carry out an unsteady-state simulation representing the power failure scenario. The next phase of the project is to automate both the simulations using ANSYS FLUENT UDFs & scheme programming languages. Considering physics involved, unsteady nature and level of automation required in the problem, the project is inclined towards higher complexity.
  • Project Level - M.E./M.Tech./M.S.Generally, any Flow problem that involves additional physical models like Heat transfer models, Multiphase models, Rotating machinery related models, Dynamic mesh models, or includes scripting and automation, etc…, we consider it as M.E. level project. In this project we are automating a 3D unsteady heat transfer involved flow inside a data center model. So, the level of the project is best suited for the post graduate students.
  • Duration - 6 MonthsAssuming the student can spend 2-3 hours of time per day and considering the amount of work involved in both learning (2 months) and working (4 months) on the project, we feel this project can be completed in 6 months. This duration might vary based on the amount of dedicated time; the student spends on the project work.

Benefits for students

Your academic project is one of the most important aspects of your degree. It is so important that it always decides what’s going to be next for you. Let it be higher studies or industrial job, the whole career path is based on the project work. With fierce competition powered by a rapid change in the world economy, every graduate/post-graduate is fighting a tough career battle today in the job market. All students look for an initial breakthrough in their careers and each one of them requires a good educational qualification complemented with a good project work.

Knowing CFD software is one important aspect for being CFD engineer, but using the CFD software for solving complex industrial problem is must when it comes to paving your path for career as CFD engineer. 

Working on this project will not only give you knowledge of solving a specific problem of HVAC, but it will also give you enough confidence to solve general HVAC problems using CFD. This project would also open your doors to the ever increasing domain of Data center and its thermal management systems. This project experience will boost your academic profile and make it suitable for companies looking for CFD engineer with HVAC expertise. Work experience on this project will open an opportunity for students in industries like Infrastructure design HVAC, Data centers, and Vehicle HVAC industry.

Following are few in-built benefits you will get when working on this project:

  • Opportunity to work on challenging projects of Data center thermal management
  • Opportunity to present project work and get reviews from industry experts
  • Project certification done by industry
  • Opportunity to understand and open doors to the huge world of data centers
  • Opportunity to learn non-technical aspects of project execution followed in industry
  • Opportunity to sharpen the domain expertise and shape future career path

This project is to be executed using ANSYS ICEM CFD and ANSYS FLUENT software. When it comes to customization and automation of CFD, ANSYS FLUENT is one of the best choices for industries. Working on this project will give you exposure to automation of CFD using ANSYS FLUENT and will open large number of opportunities in the latest and ever growing demand of automation in CFD industry.

Photo: Tooltip cfd-bg-pr-004-cfd-modeling-using-star-ccm4 Add a photo

Project Details

Mentor Project Details

A company has bought a complete room from a data center provider to setup its servers which would serve the company’s online business requirements. The company has provided all the server requirements to the data center. Now it’s time for the data center to assign this work to one of its finest employees. Imagine yourself to be in that position and try to visualize what all types of work you are going to undertake to make it a success.

The responsibilities lying on you would be :

  • Generate a layout of the data center room considering the requirements of the client and constraints of the data center building structure & cooling system
  • Perform CFD study of that layout to confirm that ASHRAE thermal standards are satisfied
  • Perform a power failure analysis considering residual cooling of the cooling units to find out the time window available to power up the cooling units by alternate sources

The input data available with you would be :

  • Drawings of the data center building cross sections
  • Location of Cooling units which is based on the piping network done inside the data center building
  • Server heat load data shared by the company
  • Number of Servers, Racks, etc. requested by the company

You, as that data center employee would design the layout by consulting the thermal experts in your team. You would then create a 3D data center model in ANSYS ICEM CFD, generate a mesh, setup boundary conditions and launch a steady-state run using ANSYS FLUENT. Based on the results, you would modify the design if needed and finally freeze the layout. The next phase is to find out the amount of time; servers would sustain without thermal breakdown in case of power failures for the finalized layout. You would then perform an unsteady simulation in ANSYS FLUENT replicating the power failure scenario. 

The work in this project would also be in line with the above discussion. The student has to create a 3D data center model, generate a hexahedral mesh and perform a steady-state simulation as well as an unsteady-state power failure simulation using ANSYS FLUENT software manually and finally automate the process and submit the automation scripts. 


Objective of the Study

The objective of this CFD study is to generate a parameterized automation scripts for data center power failure analysis. This script would be able to accept any 3D data center model’s mesh in .msh file format, perform the complete set of simulations, and generate post processing/result data. In order to achieve the above objective, the student needs to develop a CFD approach for solving the real world data center’s power failure scenario and then would find a scripting logic suitable to support the above objective.

Project inputs to be shared with students

LearnCAx discussed with CCTech to extract the inputs required for the execution of this project. These inputs will be shared with the students before they start with the project work. Following is brief information on the project inputs and how to use them. 

Data Center Model: Student will receive data center’s layout and drawings. The student will also receive list of data center components to be placed inside that room. He will be given guidelines to design the data center layout. He will be provided with the thermal specifications of each of the components present in that room. 

Steady-state analysis prior to Power failure scenario: The CFD objective of this simulation with required help to fetch boundary condition data from the data center component details provided earlier will be provided to the student once the student verifies the hexahedral mesh of the data center from the mentor.

Power failure scenario’s analysis: The plot showing residual cooling capacity of the cooling units versus time and the plot showing reduction in air flow rate through the cooling units versus time will be provided to the student after completion of the first simulation. These plots will help the user to provide input into ANSYS FLUENT for carrying out the second simulation.

Expected deliverable from student

After completion of the project, set of deliverables has to be submitted to LearnCAx and CCtech. These deliverables include detailed project report and necessary software files. Following are the details of these deliverables. These deliverables will be used to review the quality of work done based on which grading and certification will be done.

Project Report

The report should be in MS Word format and expect to provide the project details starting from the problem description, 3D data center model, Meshing details, case setup & results of the steady-state simulation, unsteady-state power failure simulation strategy, solver setup & results, automation strategy, process & files and finally conclusion at the end. An overview of the expected content is provided below.

  • Introduction: A brief introduction to the project domain along with the need for the study is expected in this section.
  • Project Overview: In this section, explain in detail about the project or the problem and also specify the objective of the problem.
  • Steady-State Simulation Details: The objective of this phase of the project is to get the converged flow and thermal data of the pre power failure steady-state situation. During this period, the student has been put on a learning curve on the data center industry, world standards in its thermal analysis, optimized CFD solver setup used for data center HVAC applications, etc. So the details about the 3D data center model, problem definition, setup, results, ANSYS FLUENT case setup, UDFs used and the learnings from this study are to be provided in this section.
  • Power failure Simulation Problem Understanding and CFD methodology: Student is supposed to submit a detailed understanding of the problem definition and create a strategy to carry out the CFD simulation.
  • Unsteady-State Simulation: Student would provide the ANSYS FLUENT case setup details, as well as the results of the run.
  • Automation Process: The student is expected to understand the need of automation of this unsteady simulation as per CCTech’s requirement, and design a plan to automate it. The student is also expected to submit the scripts which would accept the mesh file of the data center model, setup the steady-state simulation and run it, use it as input for the unsteady-state simulation, do necessary case setup, run the simulation and save the results in ANSYS FLUENT software.  
  • Conclusion: Finally end the report with a summary or conclusion of the complete project work. 

Files to be submitted

All the necessary files related to this project are to be submitted to the company for their future reference. This includes both CoolSim 4.3 and ANSYS FLUENT associated files, as listed below.

  • 3D data center model (final setup) : .tin file
  • ANSYS ICEM CFD Block file : .blk file
  • ANSYS ICEM CFD Mesh file : .uns file
  • ANSYS FLUENT mesh file : .msh file
  • Simulation 1 case file : .cas file
  • Simulation 1 data file : .dat file
  • Simulation 2 case file (final time step) : .cas file
  • Simulation 2 data file (final time step) : .dat file
  • UDF codes : .c file
  • Scheme scripts : .scm file
  • Journal scripts : .jou file (if any)


Mentor Project Mentor

LearnCAx mentor program connects Students, Mentors, and Industrial/University Projects together. Through this unique program, we give an opportunity for students to work on challenging projects offered by industry or assigned by your university. Main aim of LearnCAx mentor program is to give all necessary knowledge and guidance to students, so that they can work on challenging projects. For success of this program, it is very critical for students to understand how this program works, what is role of LearnCAx mentor, and what is role of student.

To get an overall idea about LearnCAx mentor program, visit Overview and How it Works? articles.

Every project has different challenges and requires specific domain expertise. LearnCAx has team of mentors. Every mentor has expertise in CFD and large work experience in executing industrial projects. They have developed domain experts in specific domain by executing industrial project in the domain for more than 5 years. When you enroll for LearnCAx mentor program, you get a dedicated mentor. Mentor is decided based on the project definition and required expertise. 

The complete LearnCAx mentor program is based on the theme of “Learn – Try – Execute”. This is student centric approach, where it is expected that student would learn and acquire all required knowledge, try the knowledge on simple problems and then execute the project. LearnCAx mentor is a guide/mentor who will be with student during every phase, let it be learning or executing the project. Mentor will provide all required guidance to student enrolled for this program. 

Following are the few responsibilities of LearnCAx mentor:

  • To check if project is feasible using CFD or not
  • To design the learning path for students which will include required courses and domain knowledge
  • Guide student to break the project into intermediate stages
  • Guide student to make required assumptions and simplify the problem
  • Guide student during their learning phase
  • Guide student during the project execution phase
  • Review the project work at regular intervals
  • Review project work and provide feedback

Our main focus is to give student a working experience on challenging project. Student will execute all the stages of project by acquiring required skill sets. Mentor will provide necessary guidance. Following are few things LearnCAx mentor will not do:

  • Provide customized training specific for the assigned project
  • Work on any of the project execution stage including meshing and simulation
  • Prepare the project report/presentation


Mentor Project Certification

About Company Offering this Project

This is an industrial project offered by Centre for Computational Technologies Pvt. Ltd. (CCTech). CCTech is a venture started by a group of IITians and industry professionals with extensive experience in CAD/CFD application, development, and testing. The average experience of a CAD/CFD professional at CCTech is more than 6 years. Members of the advisory board and principal consultants are specialists in various applications of CAD/CFD, empowering CCTech to handle complex CAD/CFD problems.

CCTech has always taken new challenges in terms of problem complexity and project time lines. It has successfully carried out various projects in high speed aerodynamics, HVAC of automobile, data center cooling, analysis of automobile defrost and ventilation ducts, volute design for pump, fluidized bed simulation, soot formation in IC engines etc.

This project is offered by CCTech’s CFD consultancy division. CFD consultancy division offers design, analysis and optimization services for various industries and successfully completed more than 100 projects. With its quality of work and capability of handling challenging project, CFD consultancy division is one of the preferred choices for many industries including automobile, heat exchanger, and control valve manufacturers, oil & gas design and consultancy firms. Working on this project will give you an opportunity to work with expert engineers in the CFD consultancy division and it would be unique learning experience. To know more about the company, visit

Assessment Process

This project will go through two levels of assessment. The first level of assessment will be done by project mentor. Second level of assessment will be done by review panel from Centre for Computational Technologies Pvt. Ltd. (Company offering this project). The assessment process is designed to make sure that a student has gone through all the necessary learning and project execution stages. The assessment process is also designed to grade the project work for quality of work done by student.

Project mentor’s assessment is a continuous monitoring process. The assessment process is designed to make sure that student executes each and every stage of project successfully with desired output and learning. Project mentor will do assessment at following stages:

  • Learning done by student to make sure that student has acquired skills to execute the project
  • Literature survey and problem understanding by student to make sure that student has understood the complexity of project and knows the execution path
  • Geometry and meshing techniques used to make sure its quality
  • Simulation and methods used to make sure that it will satisfy the objective of simulation
  • Project report review and presentation to make sure that the project objectives are satisfied

After completion of project, the final assessment and review will be done by team from Centre for Computational Technologies Pvt. Ltd. The review will be done based on the project report submitted by student. The project work will be graded based on following criteria

  • Aim and objective of the simulation work done by student
  • Geometry and physics simplifications done by students and its validity
  • Meshing method used, cell count and its quality
  • CFD models, boundary conditions used and its validity
  • Agreement of CFD results with data available with CCTech
  • Simulation results and student’s interpretation about the results


After successful completion of the project, student will get a certificate issued by Centre for Computational Technologies Pvt. Ltd. This industrial project certificate will add a great value in student’s profile and will lay a foundation for their career in CFD domain.

Student’s project work will go through a rigorous review process set by CCTech. A review team will grade students work and assign grading out of 10. CCTech will give a project completion certificate with acquired grade to the student.


Mentor Project FAQ

1. Will I get project completion certificate from LearnCAx or the company?

You will get the certificate from the company. You can appear for a test and get course completion certificate separately from us. This is independent of the project completion.

2. Why there is cost associated with this project?

Although this project is offered by industry, there is need of courses and dedicated mentor to guide the student throughout project journey. The total cost has two components, one the cost associated with required LearnCAx courses and cost associated with guidance provided by mentor. Some part of the total cost is also for continuous assessment and project certification.

3. You have mentioned that the project is for ME or MTech or MS students. I am a BE or undergraduate student, can I do this project?

Ideally this project is for ME/MTech level students. If you are a BE student you will need more time than the mentioned project duration. If you are interested in the project and have 8 months to an year to work on the project then you can consider this project.  

4. We are a group of 3 to 4 students. Can we as a group work on this project?

No. This project is for a single individual only. A group of student cannot take this project.

5. As a undergraduate or BE student can I do this project individually?

Yes. If you are an undergraduate or BE student having interest and appropriate time and background required for this project you can take this project. But you will need to do this project individually and not in group. Also as this is a ME level project, you should expect that it will be challenging at BE level. A BE level student must have at-least an year to work on the project.


Last Reviews

There are yet no reviews for this product.
Now please write a (short) review....(min. 100, max. 2000 characters)
First: Rate the product. Please select a rating between 0 (poorest) and 5 stars (best). Rating:
Characters written:

Browse Knowledge Base

Recommended By

Get Instant Updates

Subscribe to get instant updates about CFD courses, projects, blogs, webinars, software tutorials & CFD jobs