# Computer chassis cooling system design using fea software package

Published:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

### 1. Introduction

As mentioned in the initial report, electronics parts are the core of all aspects of human life since everything now is computerized and artificial intelligence runs everything: schools, universities, governments and even the coffee machine at the office. The microprocessors of these computers get heated due to analyzing the electrical signals coming to it and sending another electrical signal to act upon received inputs. Because of this repetitive operation that can occur for million times within a second, these microprocessors generate, relatively, a huge amount of energy in the form of heat that may lead to malfunctioning the item in the short term or reduce its life cycle in the long term.

From these facts, the need for objects that enhances the disposal of the generated amounts of heats is becoming more vital in order to push the wheel of microprocessor technology forward, and this is the topic of this research.

### 2. Draft Contents

In the third section of this report, the objectives will be clarified in order to give an indication about what the project will accomplish. This section will also show the importance of the project and the area it can be applied in.

The fourth section will describe how these objectives will be accomplished; it will provide the reader with the methodology and the flow of the work that will be done in order to reach the required results of the project.

An approximately 2000 words literature review about the former studies that were made in this field will be provided and books that talks about the theories involved with the topic.

The time plan that this project's progress will work according to it will be given in the sixth section. All efforts will be made to stick with this plan.

Seventh section will provide all the references that were used in this report with proper documentation.

The appendix in the last pages will present the initial report with new changes indicated in bold font.

### 3. Objectives

Generally, when electronic device is in use, they generate a lot of heat due to the resistance of electrons flowing through the material of the CPU or the microprocessor. The more processing operations within these devices the more resistance and thus more heat generation. Usually, a normal desktop computer operating at average usage factor generates between 60 to 2501 watts of heat, this amount of heat could cause serious damage to the computer if not disposed of properly. The most efficient way of getting rid of that heat is the usage of a fan to increase convective heat transfer rates, and sinks to increase the surface area exposed to the ambient air.

From the initial report and the introduction of this progress report, the aim of this project is to try developing a model that relates the heat sink design with the air flow rate to accomplish the best heat rejection object. This model will enhance the heat transfer operation from electronic equipment in order to achieve better results. The main objectives that the project will try to accomplish are:

1- Design a heat sink that is efficient, reasonably sized and rejects the required amounts of heat.

2- Establish a numerical model that relates the heat transfer factor all together (geometrical dimensions and air velocity) and make optimization over wide ranges to reach the best heat transfer sink-fan combination.

3- Simulate the produced model using a simulation program to prove the results.

Michael Bluejay, (2009), ©1998-2009 Michael Bluejay, Inc. All Rights Reserved. Available at: http://michaelbluejay.com/electricity/computers.html, accessed on December, 30th 2009.

### 4. Methodology

First, the amount of heat rejected by common microprocessors is obtained in order to know what is the heat transfer ratio being targeted, and that will be obtained from the literature review and external sources.

The second step will be determining the convection heat transfer coefficient at various air velocities. This may be accomplished by a series of equations and procedures:

### First, obtain the Reynolds number:

Then, obtain Prandtl number from thermodynamics tables. For air, Pr is expected to be between 0.7 and 1.

After the two dimensionless variables have been obtained, the Nusselt number can be computed using the proper empirical relation, and equaling it to its equivalent equation, in other words:

Where ρ is the air's density, v is the air velocity, is the dynamic viscosity, Lc is the characteristic length, h is the heat transfer coefficient and k is the thermal conductivity of the air.

Once h is obtained, the third phase can be started which is determining the design and geometry of the heat sink. The operation will need a lot of work since it involves selecting many variables, such as fin shape, length, width, thickness, diameter, orientation size and number of fins.

When initial designs and selections are completed, the results and mathematical equations will be assigned into a computer program (most likely Microsoft Excel) and examine the results on a range of numbers in order to obtain the best forced air convection cooled sink that gives the required amount of heat rejection. A graphical representation of the data will be provided in order to give full information about the relation between heat rejection ratio and any variable in the designed system.

After the design has been selected and set, a simulation will be performed using a computer program (SolidWorks for the time being) to ensure that the selected model is capable of doing what it is designed for.

### 5. Literature Review

(Cengel, 2006) According to Newton's law of cooling (), there are two way to increase heat transfer rate when temperatures are fixed due to design consideration, increase heat transfer coefficient h or surface area As, and since increasing h is difficult and not accurate, the best way to increase Q is by adding more surface area extended to convection and radiation and that's by attaching extended, high thermal conductivity surfaces called "Fins". In heat transfer analysis for fins, heat transfer coefficient is considered to be the same at any point of the fin, no heat generation occurs and the thermal conductivity is constant. In reality, h value varies along the fin, and that's because of the ability of the fluid to move around the fin, in other words, h is low at fin's base because the solid surfaces impedes the fluid's motion while on the tip of it the h value noticeably increases because of higher fluid motion. Increasing the number of fins may retard the heat transfer operation.

There geometrical parameters that strongly affect fin heat transfer are: width, thickness, length and the space between two adjacent fins. When designing fins, length of the fin must be chosen wisely because increasing the fin length doesn't mean that heat transfer will be enhanced, on the contrary, it may be reduced, and that due to the fact that as the length of the fin increases, the further points from the base are reaching the ambient temperature, and thus, no heat transfer will occur because there is no temperature gradient and all the extra material employed will be useless, adding to that, the extra lengths will reduce fluid ability to move around and thus decreasing h and the overall heat transfer. Fins should coincide with the equation: , where is the characteristic thickness of the fin.

Fin efficiency is defined as the ratio of the actual heat transfer rate from the fin to the ideal heat transfer rate from the fin if the entire fin was at the base temperature. Since fins basic mission is to enhance heat transfer process, then installing them must be justified economically and technically, so the term fin Effectiveness is used to express this aspect, it is defines as the ratio of heat transfer rate from the fin of base area to the heat transfer from the surface of the area, from definition, effectiveness must be greater than 1to make fins installation justified.

Fins that are used to cool electronic equipment are usually called heat sinks, they have very complex geometrical designs and are usually expressed with their thermal resistances by the equation: .

Convection heat transfer is the process by which heat transfer occurs due to fluid motion over hot or cool surfaces, continuously bringing new portions of the fluid in contact with surface and thus transferring heat by conduction between them. It differs from conduction in that it needs the motion of a fluid. At the same conditions, convection is much more efficient than conduction because it keeps bringing new quantities of fresh fluid to contact with the surface, maintaining the same temperature gradient and thus the same transfer ratio. Convection is mainly classified in to two types depending on the type of motion of the fluid, Natural: by which air moves due to density change with temperature variance and Forced when a fan or a blower is used to make motion.

There are quantities that are obtained in order to obtain the heat transfer that occurs by convection, usually they are dimensionless numbers because the give and indication about something and don't actually have a physical meaning, another reason why they are nondimensionalized is to reduce the total number of variables. Nusselt number is quantity that represents improvement of heat transfer operation by fluid convection relative to conduction of the same fluid, that is, if Nu equals 1, the operation is purely conduction. Nu is given by the equation: . Prandtl number represents the ratio of the molecular diffusivity of momentum to the molecular diffusivity of heat, to clarify this, heat diffuses very slow in oils, so Pr>>1, while in metals the operation is fast so Pr<<1. In order to describe the characteristics of the fluid motion, Reynolds number is used (Re). Reynolds number expresses the regime of the flow by surface geometry and roughness, velocity, kind of the fluid and the temperature of the surfaces over which the fluid is passing, i.e., the ratio of the inertia forces to viscous. The importance of the previous quantities is that Re and Pr are used to obtain Nu by empirical relations involved with each case, and through Nu, convection heat transfer coefficient h can be obtained and Newton's law of cooling can be obtained to determine the amount of heat transfer.

Laminar flow is the one which is characterized by smooth sharp and organized motion. Turbulent is a flow that involves fluctuations and disorder in the motion of the particles of the fluid. Heat transfer in Turbulent flows is much higher than laminar ones because its fluctuation and vortices allows for all the parts of the fluid to be in contact with the surface from which heat is transferred from, while lamina flow doesn't allow for such mixing and thus heat transfer is retarded. Type of the fluid is recognized by the Reynolds number. Change of the fluid characteristic doesn't occur suddenly, there is a transformation phase called Transition.

(Arularasan, et. al., 2008) This study investigates the proper dimensions of parallel plate fins in order to achieve the maximum heat rejection from generated in the electronic components working within electrical systems such as computers. The aim of the study is to test these fins at different geometric dimensions. The study has reached to the result that the optimum fin dimensions for its height, thickness, base height and space between two fins are 48, 1.6, 8 and 4 mm, respectively. The Importance of the study arises from the fact that as time moves on, the need for lighter, faster and smaller computers that are capable of processing greater amount of data in shorter times are becoming more needed, and since these required qualities will cause the electronic components to emit more heat in order to keep functioning, and so the need for more efficient heat disposing objects are becoming more crucial. Enhanced heat transfer, smaller pressure losses, simpler and easier structures manufacturing and relative low costs should be considered when designing heat sinks (fins). The simulation and modeling processes of CFD calculations are performed in three steps, the pre-processing in which goals are set and the calculating arrays are prepared, the second stage is the solving execution where the boundary conditions and the computational methods are set to start the solver to start processing data until convergence is obtained and finally the post-processing when obtained results are examined.

(Tiselg, et. al., 2004) This study investigates the numerical results of experiments studying the heat transfer operation between water and silicon triangular micro channels that have a hydraulic diameter of 160 Micro meters at a Reynolds number of 3.2-64. It was noticed that heat transfer is reversed at the end of the channel at higher Reynolds number and that the change of temperature between bulk water and walls is not linear. It was proven concluded that at a mass flow rate of 0.0365g/s and after water has transported 0.12m within the tunnel the temperature of water and micro channels is identical, after that the heat transfer is reversed. The maximum axial heat transfer occurs at the inlet.

(Qu, 2006) The flow development and pressure drop for a rectangular micro channel was examined numerically and experimentally. The experiment conditions were: Re=196 - 2215, width, depth and length dimensions of 222, 694 and 120000 Micro meters, respectively, one dimensional water flow and temperature difference of zero. The study proves that the Navier-Stokes is accurate which makes it a powerful tool in designing micro-heat sinks. When high Reynolds number is present at the entrance, vortices are formed at that point and it affects the pattern of the flow until it leaves the channel.

(Islam, 2009) This study investigates the performance of heat transfer inside a finned tube; the fins are T-section shaped. The experiment was conducted at Reynolds number ranging between to at which, air was fully turbulent developed. Electrical heater was used as a source of heat. It was discovered that in comparison with smooth tubes, finned tubes friction was five times higher and the heat transfer coefficient was doubled. For smooth tubes, the study shows that the temperature of the wall at various locations increases almost linearly with distance X inside the tube, however, when it reaches .55m, it becomes constant until it starts dropping after .85m due to the end effect. In the case of finned tube, the temperature of the wall increases almost linearly with the length of the tube with no drop downs, the temperature of the fins tips were always less than the wall's temperature by 6-10 due to two factors: internal resistance of the fins and because it was always in contact with colder fluid. In both cases, it was proven that with increasing Reynolds numbers; the temperature of the walls decreased. The local heat transfer coefficient was inversely proportional with the linear distance (x/L) until it reached 0.4 after that it became almost constant, and was enhanced with higher Reynolds numbers. The study also reached to other main conclusion: Fully developed thermal region started on x/L=0.3 with an x/D ≈6 for smooth tubes and x/L=0.4 and x/D≈8 for finned tubes at the same Reynolds number and nominal diameter.

(Teerstra, 2004) Analytical models are established to study the effects of size and location of the slotted heat sinks for forced convection heat transfer, this study is conducted on a wide range of Reynolds number. It was proven in a previous study (R-Theta, 1999) that slotted plate fin heat sinks provide higher heat transfer rates because it allows the creation of new thermal boundary layers on each section which will provide the object with higher heat rejection ratios but still may decrease the heat transfer because of the reduction of surface area. An optimization between number of slots, forced convection surface area, thermal conductivity of the material, location and dimensions of the heat sink is made to reach to the best heat rejection ratios. The study has reached for many interesting conclusions, it was indicated that the dimensionless heat transfer coefficient was improved significantly at the upper and lower bounds of the slotted plate heat sinks. The experimental data almost coincided with the numerical testing results, with a slight shifting to the upper limits. Wide ranges were covered in the research: Reynolds number ranging between 40 to 180, S/P ratio of 0.5, and a P/L ratio between 0.059 and 0.4. The New proposed models depended on the arithmetic mean of the bounds, with an increase of RMS by 12% from the old fins that employed normal plate fins. The staggered fins showed better heat transfer characteristics than inline fins, this may be regarded to the fact that staggered fins establishes turbulent flow, which by default enhances the heat transfer as in the case of bundle tubes. Care should be taken when sizing the slotted fin in order to avoid dramatic decrease in the convective heat transfer area and there production should be economically feasible. Enough data were collected to establish a model for relating the slot size with other independent parameters.

(Chapman, et. al., 1994) A comparison made between extruded, cross-cut rectangular and elliptical aluminum fins. This study was carried out to compare the elliptical pins with other fins kinds. Elliptical fins were designed to minimize pressure losses by eliminating the effect of the vortices and increasing the heat transfer surface area, and thus, the heat transfer operation was enhanced. The experimental result where compared with those obtained from CFD analysis carried out by using Sauna Heat sink simulation program and were found to be very close to each other, the tests were performed at low air velocities. It was also noticed that the thermal resistance of straight fins were the lowest compared to the other fins due to enhanced lateral thermal conductivity and lower flow bypass properties. The elliptical fins showed the highest âˆ†T/Q ratio at the same air velocity, cross-cut pins were slightly lower than them.

### REFERENCES:

- Cengel,Yunus, (2006), HEAT AND MASS TRANSFER A Practical Approach, McGraw Hill.

- Arularasan, R., Velraj R., (2008), International Journal of The Computer, the Internet and Management Vol. 16.No.3, Department of Mechanical Engineering, SSN College of Engineering, India.I.

- Tiselj I., Hetsroni G., Mavko B., Mosyak A., Pogrebnyak E., Segal Z., (2004), Effect of axial conduction on the heat transfer in micro-channels, International Journal of Heat and Mass Transfer, Available at sciencedirect.com.

- Qu, Weilin, Mudawar, Issam, Lee, Sang-Youp, Wereley, Steven, (2006) Experimental and Computational Investigation of Flow Development and Pressure Drop in a Rectangular Micro-channel, Journal of Electronic Packaging, ASME.

- Md. Islam, Asharful, Mozumder A., (2009), Journal of Mechanical Engineering, Vol. ME 40, No. 1, Transaction of the Mech. Eng. Div., The Institution of Engineers, Bangladesh.

- Teertstra P., Culham J.R., M.M. Yovanovich, (2004), Analytical Modeling of Forced Convection in Slotted Plate Fin Heat Sinks, Microelectronics Heat Transfer Laboratory, Department of Mechanical Engineering, University of Waterloo, Canada.

- Chapman Christopher, Lee, Seri, Schmidt, Bill, (1994), THERMAL PERFORMANCE OF AN ELLIPTICAL PIN FIN HEAT SINK, Tenth IEEE SEMI-THE, IEEE.

- Michael Bluejay, (2009), ©1998-2009 Michael Bluejay, Inc. All Rights Reserved. Available at: http://michaelbluejay.com/electricity/computers.html

### Appendix: Adjusted initial report

### Chapter one: Introduction

### 1.1 Background

Electronics can be considered the nerve of the modern science, in which it control all major engineering application from computers to electromechanical systems, adding to that a numerous amount off applications that cannot be mentioned in this part of the project. According to this fact it should be aware that appropriate conditions should surround any electronic system to maintain the correct electronic behavior of such electronic components, this may causes lose of its normal characteristics due to the up normal variation in the work atmosphere, which is going to drive any electronic systems into improper work output, or in some cases may drive the system into a catastrophic results in the output or and total malfunction, or in some cases to a hardware breakdowns.

All these facts mentioned above leads to the clear result that to maintain a correct operation of any electronic system, correct atmospheric conditions should surround the electronic board. This job should be done very carefully, and one of the main sections of this subsystem is the air flow system that cools the electronic parts in the electronic board, this system usually uses an electrical fan that push or draws air to or from the electronic system.

So in this project it is attend to do some CFD analysis using SolidWorks Flow Simulation, which provides a good and easy to use CFD analysis tool which is going to carry the calculation of the analyzed system. Also in the project the effect of the fan on the cooling process in the computer chassis will be studied with all the variables connected to this issue (Paul M.Kurowski 2004).

References should play a key role in this project, it is needed to see similar past researches that talked about this topic, and meanwhile it is also necessary to see maybe a brochure or a repair manual that explains the heat management cycle in the computer's chassis. Also a good training manuals or courses in SolidWorks Flow Simulation is needed to fully understand how to navigate correctly in the SolidWorks Simulation environment due to the complexity in the CFD in general, also it is important to fully understand how to collect necessary results from the analyzed case(Paul M.Kurowski 2004).

### 1.1 Project Aims

To closely investigate the thermal environment that surrounds the computer's chassis, in which the heat produced from the main electrical and electronic parts of the computer is generated and released into the surrounding, and of course to change parameters of the thermal control devices such like the electrical fan in such a way that preserves the sensitive component in the chassis from being thermally corrupted due to overheating caused by improper heat dissipation. There will be a number of parameters that need a fair research and those may just be like the number of fans, there location, the amount of air need to be delivered through each fan...etc.

### 1.2 Objectives:

1- Design a heat sink that is efficient, reasonably sized and rejects the required amounts of heat.

2- Establish a numerical model that relates the heat transfer factor all together (geometrical dimensions and air velocity) and make optimization over wide ranges to reach the best heat transfer sink-fan combination.

3- Simulate the produced model using a simulation program to prove the results.

### Chapter 2 . Literature review

Computer chassis cooling system has been a research topic for many years, so there is very good references about this topic. Also the basic engineering of the design of the electronic should be well study. For example the heat transfer process is a very important topic to study. Cengel,Yunus, (2006), HEAT AND MASS TRANSFER A Practical Approach is considered a very good text book that can be referenced for such a project. Also there should be a very good reference for the electrical parts due to the importance of understanding power consumption for the important parts such like the processor and the rams of the computer. Also patents websites can provide a good reference for the patents that discus the same or similar topics, one of these websites is www.patetns.com.

Very important reference to be mentioned is the sources provided by SolidWorks. The SolidWorks Flow Simulation Training Manual which discussed the air flow through electronic chassis in a very excellent way. One of good books that help in good understanding for Finite Element Analysis and its role in building SolidWorks software package is Finite Element Analysis for Design Engineering by Paul M.Kurowski

### Chapter 3: Methodology and Analysis

In this project SolidWorks- as a numerical technique -Flow Simulation is going to be used, the model of the chassis is going to be inserted and also the electronic components should be modeled as thermal loads, then the thermal conditions of the chassis should be inserted into the software and by applying correct variables into the Simulation correct thermal and flow results should be obtained. Of course before starting to apply data into SolidWorks, it should be collected from outside, using proper datasheets.

That aim and these objectives are assumed to be reached using experimental approach using Flow Simulation that gives the user a very good tool to create study, analyze many thermal and flow situation just like our case of study that will include many parameters that can be researched using SolidWorks. Also I think that an advice from a specialist in the field of the design of thermal analysis for electronic parts is needed to reach a certain level of perfection in our design. Finally it should be mentioned that a very good research about electrical power supply and power consumption in the main electronic parts in the chassis compartments, this is so important to this research to have enough background information about the power consumption and the heat dissipation so that it can be available to us to create a suitable air cooling system.

It should be fully known what kind of chassis and what kind of fans that are going to be used, one of the most important information to be collected is how much air can a single fan draw from the atmosphere, this is very important to make the right calculations about how many fans that we are going to need to supply enough cooling to the chassis.

### 3.1 Conclusion:

Of course using SolidWorks provided an excellent advantage for the project. It provided an excellent and handy tool to acquire results for the discussed topic. But as a disadvantage it cannot easily check the results without data from the manufacturer of the given computer or electronic part.

### 6. References

1. SolidWorks Flow Simulation Manual.

2. Cengel,Yunus, (2006), HEAT AND MASS TRANSFER A Practical Approach, McGraw Hill.

3. Arularasan, R., Velraj R., (2008), International Journal of The Computer, the Internet and Management Vol. 16.No.3, Department of Mechanical Engineering, SSN College of Engineering, India.I.

4. Tiselj I., Hetsroni G., Mavko B., Mosyak A., Pogrebnyak E., Segal Z., (2004), Effect of axial conduction on the heat transfer in micro-channels, International Journal of Heat and Mass Transfer, Available at sciencedirect.com.

5. Qu, Weilin, Mudawar, Issam, Lee, Sang-Youp, Wereley, Steven, (2006) Experimental and Computational Investigation of Flow Development and Pressure Drop in a Rectangular Micro-channel, Journal of Electronic Packaging, ASME.

6. Md. Islam, Asharful, Mozumder A., (2009), Journal of Mechanical Engineering, Vol. ME 40, No. 1, Transaction of the Mech. Eng. Div., The Institution of Engineers, Bangladesh.

7. Teertstra P., Culham J.R., M.M. Yovanovich, (2004), Analytical Modeling of Forced Convection in Slotted Plate Fin Heat Sinks, Microelectronics Heat Transfer Laboratory, Department of Mechanical Engineering, University of Waterloo, Canada.

8. Chapman Christopher, Lee, Seri, Schmidt, Bill, (1994), THERMAL PERFORMANCE OF AN ELLIPTICAL PIN FIN HEAT SINK, Tenth IEEE SEMI-THE, IEEE.

9. Michael Bluejay, (2009), ©1998-2009 Michael Bluejay, Inc. All Rights Reserved. Available at: http://michaelbluejay.com/electricity/computers.html

10. Paul M.Kurowski 2004, Finite Element Analysis for Design Engineering, 1st edition , SAE publisher.