Ai－Gang Pan，Jun－Biao Wang，Xian－Jie Zhang，Xiao－Bao Cao

(Shaanxi Engineering Research Center of Digital Manufacturing Technology，Northwestern Polytechnical University，Xi’an 710072，China)

1 Introduction

Advanced portable electric devices such as cell phone，personal digital assistants(PDAs)，pad，etc.are widely used in everyday life.With the advancement in technology，the functionalities of these electronic devices have significantly increased while their sizes and weights have decreased.With becoming smaller and more powerful，electronic devices will create a high heat flux situation in which the limiting design factor is thermal management［1］.An electronic device fails to fulfill its intended function when its application or environmental condition exceeds its application limit.A US Air Force survey indicates that the percentage of temperature related failuresin electronicsexceeds 55%［2］.Traditional air cooling is not sufficient for high heat fluxes，which means new thermal management approaches must be considered.

It iswellknown thattemperature maintains constant or varies in the narrow temperature range when materials absorb or release thermal energy during phase of materials changing process.Based on this physical phenomenon，passive thermal management by using PCMs which can be called phase change thermal control is suitable for applications in which the heat dissipation isintermittentortransient.Substantial research evident that phase change thermal control is a reliable，light weight and low power technology for thermal control of aerospace and portable electric devices［3－6］.

Phase-change thermal control unit(PTCU)filled with PCMs is a thermal control component of electronic devices.The thermal control principle of PTCU is that PCMs in the PTCU absorb energy as latent heat to avoid electronic device exceeding a required operating temperature limit during the electronic device working，and then PCMs release energy to prepare the next working cycle during the electronic device unworking.

In this study，PTCU filled with PCM of Bismuth alloy(49%Bi－21%In－18%Pb－12%Sn)was made for an electric devices thermalprotection.Resistance heating components(RCHs)produced 1，3，5，7，and 10 W for simulating heat generation of electronic devices.The temperatures of RCH with PTCU and without PTCU were measured during heating period and one duty cycle period.The feasibility of these thermal control apparatus was verified according to the time which was the electronic device temperature exceeded the maximum operating temperature.

2 Experimental Setup and Procedures

For analyzing the thermal performance of the PTCU，the values of some important variables must be clarified before the detailed discussion.During experimental testing，the temperature of the heater should be kept under 70℃ since it is typical the operating temperature for most electronic devices to ensure reliability.

2.1 Preparation of Metallic PCM

PCMs use latent heat for energy storage.Many substances have been studied as potential PCMs，but only a few of them are commercialized as so.In Ref.［7］，Abhatconcluded form the information compiled that the main characteristics require of PCMs.The main criteria that govern the selection of phase change heat storage materials are as follows［7］:

1)Possessa meltingpointin the desired operating temperature range(temperature range of application);

2)Possess high latent heat of fusion per unit mass or per unit volume(to achieve high storage density compared to sensible heat storage);

3)High specific heatproviding additional significant sensible heat storage effects;

4)High conductivity，so that the temperature gradients for charging and discharging the storage material are small;

5)Low volume change due to the phase change， which means a simple container and heat exchanger geometry can be employed;

6)Exhibit little or no undercooling during melting;

7)Possesschemicalstability，no chemical decomposition and corrosion resistance to construction materials;

8)Contain non-poisonous，non-flammable and non-explosive elements/compounds;

9)Available in large quantities at low cost.

Abhat classified PCMs as inorganic materials and organic materials［7］.Belén et al.［8］summarized a complete review of the types of material which have been used，advantages and disadvantages，as shown in Table 1.Among inorganic materials are hydrated salts and solution salts，and among organic materials are alkanes，waxes or paraffins.Comparing with these PCMs，metallic PCMs have the advantages of higher density， higher conductivity and little or no undercooling during melting in the aspect of energy storage characteristics require.

Table 1 Advantages and disadvantages of organic and inorganic materials［8］

In this paper，a required operating temperature limit is 70 ℃，which means the temperature of electronic devices should be kept below 70℃ avoiding failure.So the melting point of PCMs filled in PTCU must be lower than 70℃ and higher than the ambient temperature.Bismuth alloy(49%Bi－21%In－18%Pb－12%Sn，mass fraction)，a low melting point alloy which melts at 57℃［9］has fine physical character on density and conductivity，was employed as PCM in this study.

The Bismuth alloy used in this study was prepared with vacuum pot in laboratory.The thermal properties of consisted components of alloy are indicated in Table 2.For determining the behavior of the PCM during the melting process the measurements have been performed with TA instruments differential scanning calorimeter(DSC).The DSC curve for the melting of the Bismuth alloy is shown in Fig.1.Table 3 shows the thermophysical properties and the phase change ranges which were obtained in this study for the Bismuth alloy used for the present investigation.

Table 2 Thermal properties of consisted components

Fig.1 DSC signal for the Bismuth-alloy(49%Bi-21%In-18%Pb-12%Sn)

Table 3 Thermophysical properties of the Bismuth alloy(49%Bi-21%In-18%Pb-12%Sn)

2.2 Preparation of PTCU

A PTCU mainlyconsistsofthePCMsand encapsulating material.The structure of PTCU is shown in Fig.2.Because of fine conductivity，aluminum was used as encapsulating material.A segment of aluminum square tube was filled with metal PCMs and sealed with sealant in both ends.Sealant can prevent the metallic PCMs from leaking during PCMs in liquid phase.Meanwhile，the sealant with good flexibility can absorb thermal expansion of aluminum square tube during PCMs phase or temperature changing.As shown in Fig.2，the dimension of the aluminum container is 10 mm×20 mm×60 mm and 0.5 mm thickness.

The PTCU manufacture process is divided into four steps as shown in Fig.3.Step 1:Select an aluminum square tube with suitable size;Step 2: Inject sealant into one end of aluminum container made in Step 1;the thickness of the sealant in aluminum square tube is about 2 mm;Step 3:After the sealant solidifying，inject liquid metallic PCMs in aluminum container;Step 4:After the liquid metallic PCMs solidifying，close another end of the aluminum square tube with sealant.A PTCU(Fig.4)manufactured through aforementioned step is 92.2 g weight，in which the weight of PCM is 82.5 g.

Fig.2 Structure and dimensions of PTCU(unit:mm)

Fig.3 Manufacture process of PTCU

Fig.4 PTCU

2.3 Experimental Setup

The structure and principle of the experimental platform are shown in Fig.5.The experimental platform consists of three subsystems which are power supply subsystem，electric heating subsystem，temperature measurement and recording subsystem.Power supply subsystem includes programmable logic controller (PLC)，power driver and 24 V AC－DC power supply. Output voltage of the power drive is adjusted between 0 V and 24 V by writing routine into PLC，and is loaded to the resistance heating components(RHCs).The electric heating subsystem consists of two switches and two RHCs used as electronic devices heater.Temperature measurement and recording subsystem includes two thermocouples(Probe A and Probe B are located as shown in Fig.6)，two signal converters，one PC with temperature recoding software.Thermocouples Probe A and Probe B for the temperature measurement are on the top of RHCs respectively.Signal converters are used forconverting the currentsignalfrom thermocouples into the digital signal which can be recognized by computer.Temperature recoding software programed by C++shown in Fig.7 is used for recording time-dependent temperature.As shown in Fig.7，Channel1 and Channel2 were recording the temperature of temperature transmitter Probe A and Probe B respectively.

Fig.5 Principle of the experimental platform

Fig.6 Probe A and Probe B positions configurations

Fig.7 Temperature recording software

Heating power of the RHCs depend on its voltage which is controlled by the PLC.The data acquired from Channel 1 and Channel 2 of the recording subsystem reflected the time-dependenttemperature ofRHC without PTCU and RHC with PTCU，respectively.The operation steps are as follows:

1)Put the thermocouples Probe A and Probe B on the top of RHCs without PTCU and with PTCU respectively;

2)Paint thin layer of Thermalbond 4952 epoxy［10］(conductivity of epoxy is about 1.34 W/(m·K))on all interfaces between the RHCs，PTCU as well as thermocouples for decreasing the effects of thermal contact resistance;

3)Adjust the voltage loaded on the RHCs by PLC;

4)Open the temperature software on PC，and set the recording period;

5)Turn on the switches，the RHCs start working and record the temperature from the two thermocouples by PC;

6)Process data and plot curves of temperature vs.time.

3 Experimental Results and Discussion

In this experimental study，heating power levels were set at 1，3，5，7，and 10 W giving power density values of at 16.13，48.39，80.65，112.90，and 161.29 kW/m2，respectively.The temperatures of the RHC with PTCU and without PTCU were measured in the situation ofnature convection during heating period.Also， the temperatures of the heating resistances with PTCU were tested in the situation of forced convection(the angular velocity of fans is 2400 rad/min for CUP cooling)at a constant heating power during one duty cycle period.It is noticed that because of the high conductivity of PCM and container，the temperature of RHC reflects the status of the PCM in the PTCU.

3.1 Results of Heating Period

The curves of temperature vs.time for the RHC are shown in Fig.8 at various heating power levels.The results indicate that the temperature of the RHC without PTCU reaches the highest operating temperature limit(70℃)within 210，57，33，23，and 15 s respectively.It is undesirable to have conditions like this to occur in an electronic device as the service temperature of the electronic components would be quickly exceeded.

At the same time，another experiment without PTCU was tested for direct comparison of thermal performance without PTCU. Fig.9 shows the temperature vs.time curves for the RHC with PTCU at 1，3，5，7，and 10 W respectively.The results indicate that the temperature of the RHC with PTCU raises more slowly than the case without PTCU at the beginning stage of experiment，and the temperature maintains within a“temperature plateau”at about 57℃.At the beginning，the PTCU absorbs a part of thermal energy from RHC as sensible heat of PCM which makes the temperature increase slowly.Until the temperature of the RHC with PTCU up to about 57℃，solid PCM in the PTCU starts melting，and plenty of heatis absorbed as latentheat.Therefore the temperature increase of the RHC with PTCU is hold back.It is observed that a lower heating power results in longer time for reaching the highest operating temperature limit.However，power input is decided by electronic devices itself not the users.Hence，various heating powerhave to be considered in design.Obviously，the PTCU postpone appearance of RHC reaching the required operating temperature limit，as shown in Fig.10.This is desirable as the PTCU extends the working time of the electronic devices and it is effective for electronic devices thermal control.

Fig.8 Temperature-Time curves of RHC without PTCU

Fig.9 Temperature-Time curves of RHC with PTCU during heating period

Fig.10 Time of heating resistances reached 70℃ under various heating power

3.2 Results of One Duty Cycle

An experimental setup with the forced convection was built and tested.At the beginning of experiment，the RCH was heated up under a heating power;when the temperature reached about 70℃，the switch was turned off and it waited for the RCH cooling down to the ambient temperature.Fig.11 shows the temperature vs.time curves of the RHCs with PTCU under different heating power during one duty cycle(including heating period and cooling period).It is found that the maximum temperature is maintained at 43℃ and hardly reaches 70℃ at 1 W，which means the equilibrium temperature of RCH is about 43℃.In this status，the heat dissipation is equal to heat absorption， therefore the temperature of RCH maintains at the equilibrium temperature.It seems that the RCH can be maintained below the 70℃ with PTCU and with a forced convection condition at 1 W.Also，as can be seen from Fig.11，with the increase of heating power from 3 to 5 to 7 to 10 W，the time for the RCH reached 70℃ decreased from 3000 to 1230 to 720 to 420 s during heating period.And the time for RCH cooling down to the ambient temperature were about 3600 s at 3，5，7，and 10 W.As the mentioned above，the time of RHC working during one duty cycle were about 6600，4800，4300，and 4000 s respected the heating power of 3，5，7，and 10 W.

Fig.11 Temperature-Time curves of RHC with PTCU during one duty cycle period

Series similar experiments but without PTCU were tested at 1 W and 3 W.Fig.12 shows the comparison of the RCH with PCTU and without PCTU temperature vs.time curves.It is observed that，for keeping the temperature of RCH below 70℃，at 1 W，the RCH with PTCU can be maintained switching-on，while the RCH without PTCU should be switching-on within about 540 s，and switching-off within about 700 s for preparing the next duty cycle.And at 3 W，the RCH with PTCU can be switching-on for about 3000 s，and switching-off for about 3600 s preparing the next duty cycle，while the RCH without PTCU should be switching-on within about90 s，and switching-off within about 700 s for preparing the next duty cycle.During 6600 s，RCH with PTCU works for about 3000 s in one duty cycle and RCH without PTCU works for 720 s in 8 duty cycles.This is desirable as the PTCU extends the period of duty cycle and the time of operation for the electronic devices.

Fig.12 Temperature vs.time curves of RCH with PCTU and without PCTU at 1 W and 3 W

3.3 Influences of Heating Power Level

The comparison of temperature vs.time curves of the RHC with PTCU at various power levels are shown in Fig.9.The temperature vs.time curves can be classified into three stages.In stage 1，the temperature of the solid PCM will increase rapidly to its melting temperature(at about 57℃).In stage 2，phase change occurs as the solid PCM melts at constant temperature，and the“temperature plateau”appeared in the each curve.In stage 3，the PCM has melted completely and thermal energy is not absorbed as latent heat any more，and the temperature of the RHC starts increasing rapidly again.

A desired operation situation of electronic devices is assumed that the working period is 20 min and the required operating temperature limit is 70℃.As shown in Fig.9，the PTCU is not fit for the heating power of 7 W and 10 W，because the temperature of RHC exceeds 70℃ at 12 min and 7 min respectively.The PTCU needs more PCM to absorb thermal energy as latent heat.At 1 W and 3 W，the PTCU keeps the temperature of RHC below 70℃ exceeding 30 min and isnotsatisfy operation situation proposed above.Because the PCM in the PTCU do not melt completely，a part of PCM which does not participate in the thermal control process is invalid and can be wiped off to reduce the weight of PTCU.At 5 W，the PTCU keeps the temperature of RHC below 70 ℃ just about 20 min，and the weight of PCM in the PTCU exactly satisfies the assumed operation situation.It is concluded that the weight of molten PCMs in the PTCU is determined by the period and heating power levels of the thermal control object.Fewer PCMs will not meet the thermal control requirements and more PCMs will increase the weight of electronic devices.

4 Numerical Model

A numerical model was developed to design and optimization of the PTCU for arbitrary set of usage conditions and heating power.A 2D model in Fig.13 was developed to reduce the computational domain.

Fig.13 2D physical model

4.1 Modeling of Phase Change

In this study，enthalpy method［11］is employed to solve phase change problems where the phase change in a range of temperature.In the enthalpy formulation，the enthalpy function H(T)，enters the problem as a dependent variable along with the temperature.The following assumptions are employed:

1)The PCM in each phase is homogeneous and isotropic;

2)The thermophysical properties of the PCM are independent of temperature，and they are the average value of properties of solid and liquid phase;

3)The effect of natural convection in the liquid PCM is not considered because the Bismuth-alloy has a good performance in conductivity;

4)In the case of PCM which changes phase over a range of temperatures，the H－T relationship during phase changes are assumed to be linear;

5)The convective heat transfer coefficient，h，is 10 W/m2K in situation of nature convection and 20 W/m2K in situation of forced convection.

The governing equation can be written as

The associated boundary conditions are

Initially at t=0，the PCM，the container and the RHC are all at the same temperature which is lower than the melting temperature of PCM:

The H－T relations for the three regions，i.e.solid phase，solid—liquid phase change temperature range and the liquid phase are written as:

where L is the latent heat of the PCM;C is specific heat of PCM;H is enthalpy of PCM;the subscript‘s’and‘l’stand for the solid and liquid states of PCM，and CP=(Cl+Cs)/2;ε reflects the half range of the phase change temperature. The thermophysical properties of the PCM used in the simulations are shown in Table 4.

4.2 Comparison Between Numerical and Experimental Results

Comparison of predicted and measured temperature vs.time curves at power levels of 1 W and 3 W are shown in Fig.14.Generally，it is observed that numerical results have good agreement with experimental results.However，the temperature of predicted result is higher comparing to measured results，and lower heating power，the decreasing agreement between predicted resultand measured results is observed in Fig.14.Several possible factors may account for these differences.

Firstly，it is observed that thin layer of epoxy are painted on all interfaces between the RHCs，PTCU as well as thermocouples for decreasing the effects of thermal contact resistance.However， less thermal contact resistance exists.Considering that no thermal contact resistance in the numerical model，the temperature of predicted result is higher than measured results.

Table 4 Thermophysical properties of the PCM for the simulation

Fig.14 Comparison of results of simulations and experiments

Secondly，the study is tested under an assumption that the specific heat of PCM is the average value of solid and liquid phase in the numericalmodel.Actually，the average value of specific heat is higher than the solid PCM and lower than the liquid PCM.Therefore，the predicted temperature rises faster than the measured temperature，and they are observed at low power levels.

Thirdly，the interfaces of heat convections are on the top，the left，and the right of PTCU in the numerical model.The front and the back of PTCU are the interfaces ofheatconvections in the actual experiments.It is probably a reason for the predicted temperature rising faster than the measured temperature.

Finally，the actual heat transfer coefficient values may be larger than in the numerical model，resulting in faster computed temperature rises than the experiments.Larger heat transfer coefficients are tried to fit the predicted temperature profile on the experimental data before phase change.However，this results in longer melting periods and lower steady state temperature.

5 Conclusions

The studies were performed to investigate the effectivenessand performance ofa phase-change thermal control unit(PTCU)as a thermal management device for electronic components.RCH at various heating power levels were utilized to simulate the thermalloads ofelectronic devices.The results indicated that:

1)PTCU can provide energy savings during electronic devices working and extend the usage time of the electronic device.During heating period，PTCU delays the time of RCH reaching maximum operating temperature for 4890，2223，1167，587，and 405 s at heating power of 1，3，5，7，and 10 W，respectively.

2)During duty cycles，PTCU provides a long period of duty cycle and little temperature varying of electronic devices.Stable temperature avoids the failure casing by thermal stresses.

3)The actual energy savings depend on the mass of PCM and the geometry of the PTCU which depend on the electronic devices operating situation(including heating power， operating time， and energy consumption).

Briefly，PTCU has an optimum performance for the electronic devices thermal control and the PTCU can be a suitable method for the electronic devices thermal control.It is noted that the discharging of energy of the PTCU need to be considered along with any additional energy consumption associated with heat sinks.

［1］Fok S C，Shen W，Tan F L.Cooling of portable hand-held electronic devices using phase change materials in finned heat sinks.International Journal of Thermal Sciences，2010，49(1):109－117.

［2］Yeh L.Review of heat transfer technologies in electronic equipment.Journal of Electronic Packaging，1995，117 (4):333－339.

［3］Gurrum S P，Joshi Y K，Kim J H.Thermal management of high temperature pulsed electronics using metallic phase change materials.Numerical Heat Transfer，Part A，2002，42(8):777－790.

［4］Kandasamy R，Wang X Q，Mujumdar A S.Application of phase change materials in thermal management of electronics.Applied Thermal Engineering，2007，27(17/ 18):2822－2832.

［5］Saha S K，Dutta P.Cooling of electronics with phase change materials.International Conference on Modeling，Optimization，and Computing.AIP Publishing，2010，1298 (1):31－36.

［6］Tan F L，Tso C P.Cooling of mobile electronic devices using phase change materials. Applied Thermal Engineering，2004，24(2):159－169.

［7］ Abhat A.Low temperature latent heat thermal energy storage:heat storage materials.Solar Energy，1983，30 (4):313－332.

［8］Belén Z，Marín J M，Cabeza L F，et al.Review on thermal energy storage with phase change:materials heat transfer analysis and applications，Applied Thermal Engineering，2003，23(3):251－283.

［9］Pal D，Joshi Y K.Application of phase change materials to thermal control of electronic modules:a computational study.Journal of Electronic Packaging，1997，119(1): 40－50.

［10］Aavid Thermalloy.Thermal adhesives.http://www.aavid.com/product-group/interface/adhesive.

［11］Sharnsundar N，Sparrow E M.Analysis of multidimensional conduction phase change via the enthalpy model.Journal of Heat Transfer，1975，(3):333－340.