Patent Application
20110174390
1. Field of the Invention
The present invention generally relates to thin film channels, microfluidic devices, biosensors, electronic cooling, control of fuel flow prior combustion and insulating assemblies.
2. Description of the Related Art
Thin films are used in a variety of devices, including electrical, electronic, chemical, and biological devices, for modulating or controlling flow and heat characteristics in the devices. See e.g. Vafai & Wang (1992) Int. J. Heat Mass Transfer 35:2087-2099, Vafai et al. (1995) ASME J Heat Transfer 117:209-218, Zhu & Vafai (1997) Int. J. Heat Mass Transfer 40:2887-2900, and Moon et al. (2000) Int. J. Microcircuits and Electronic Packaging 23:488-493 for flat heat pipes; Fedorov & Viskanta (2000) Int. J. Heat Mass Transfer 43:399-415, Lee and Vafai (1999) Int. J. Heat Mass Transfer 42:1555-1568, and Vafai & Zhu (1999) Int. J. Heat Mass Transfer 42:2287-2297 for microchannel heat sinks; Lavrik et al. (2001) Biomedical Microdevices 3(1):35-44, and Xuan & Roetzel (2000) Int. J. Heat Mass Transfer 43:3701-3707 for biosensors and nanodevices.
For many of these applications, modulation and control of the flow and heat characteristics in the devices is desired. Unfortunately, the prior art methods for modulating and controlling the flow and heat are difficult or problematic. For example, a two phase flow in a microchannel is capable of removing maximum heat fluxes generated by electronic packages, but instability occurs near certain operating conditions. See Bowers & Mudwar (1994) ASME J. Electronic Packaging 116:290-305. Further, the use of porous medium for cooling electronic devices enhances heat transfer via the increase in the effective surface area, but the porous medium results in a substantial increase in the pressure drop inside the thin film. See Huang & Vafai (1993) Int. J. Heat Mass Transfer 36:4019-4032, Huang & Vafai (1994) AIAA J. Thermophysics and Heat Transfer 8:563-573, Huang & Vafai (1994) Int. J. Heat and Fluid Flow 15:48-61, and Hadim (1994) ASME J. Heat Transfer 116:465-472.
Therefore, a need still exists for methods of modulating or controlling heat and flow characteristics in thin films.
The present invention generally relates to thin film channels, microfluidic devices, biosensors, electronic cooling, control of fuel flow prior to combustion, and insulating assemblies.
The present invention provides methods to modulate flow and heat in a variety of thermal systems including thin film channels, microfluidics, insulating assemblies, and the like with no need for external cooling or flow controlling devices.
The present invention provides several devices for modulating flow and heat. Several devices provided herein reduce the temperature as the thermal load increases as related to electronic cooling and cooling of engine applications. Several devices provided herein reduce the flow rate as the thermal load increases which are important to internal combustion applications where fuel rate needs to be reduced as the engine gets overheated. Several devices provided herein conserve thermal energy as the temperature increases and to reduce leakage from microfluidics. These devices have applications related to thermal insulations and biosensor devices among others.
An apparatus can be configured which includes a first gas compartment comprising a first substrate and a second substrate. Such an apparatus can also include a second gas compartment comprising the second substrate and a third substrate. The first substrate can include a face in contact with at least one hot medium and having another face in contact with a main gas and having fixed supports. Additionally, the second substrate can include a face in contact with the secondary gas and another face in contact with a main gas. Additionally, the third substrate can include a face in contact with the secondary gas and having another face in contact with at least one cold medium.
Such an apparatus can also include a first at least one flexible seal attached to one end of the second substrate and to the opposing end of the first substrate to form at least one closed enclosure comprising the main gas so that the second substrate moves when the main gas undergoes volumetric thermal expansion. Such an apparatus can further include a second at least one flexible seal attached to other end of the second substrate and to the opposing end of the third substrate to form at least one closed enclosure comprising the secondary gas so that this enclosure contracts when the main gas undergoes volumetric thermal expansion. Additionally, the second gas compartment can be vented such that the secondary gas pressure remains constant and allows a maximum volumetric thermal expansion of the main gas.
In another embodiment of such an apparatus, the secondary gas compartment can further include a plurality of elastic balloons filled with the main gas. In yet another embodiment of such an apparatus, the third substrate is not fixed and the second at least one flexible seal can be attached to one end of the first substrate and to the opposing end of the third substrate. In still another embodiment, the main gas can possess a thermal conductivity that is smaller than that of the secondary gas so that effective thermal insulation properties of said apparatus are improved when the main gas undergoes a volumetric thermal expansion. In other embodiments, the flexible seal can comprise a closed-cell form seal that can sustain high temperatures. An additional enhancement in insulating property can be obtained utilizing a series of the first and second gas compartments.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are intended to provide further explanation of the invention as claimed. The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute part of this specification, illustrate several embodiments of the invention, and together with the description serve to explain the principles of the invention.
This invention is further understood by reference to the drawings wherein:
on the dimensionless exit mean bulk temperature for a single layer flexible microchannel heat sink.
on the dimensionless average lower plate temperature for a single layer flexible microchannel heat sink.
on the dimensionless average convective heat transfer coefficient for a single layer flexible microchannel heat sink.
on UReo and UF for a single layer flexible microchannel heat sink.
on the pressure drop ratio and the friction force ratio between single and double layered flexible microchannel heat sinks.
The present invention provides methods for modulating or controlling heat and flow characteristics in a variety of devices. In particular, the present invention provides flexible seals for modulating or controlling heat and flow characteristics in devices comprising thin films, such as thin film channels, microchannels, microfluidics and the like. The present invention also provides a method to control heat and flow inside other thermal systems, such as insulating assemblies and fuel flow passages. As used herein, a “flexible seal” refers to a material that can be deformed significantly according to the load acting upon it. Examples of these materials include elastmors, polymers, natural rubber, closed rubber cell foams, and the like. In some embodiments, the present invention provides flexible complex seals for modulating or controlling heat and flow characteristics in devices comprising thin films, such as microchannels and microfluidics. As used herein, a “flexible complex seal” refers to a flexible seal comprising at least one closed cavity of stagnant fluid. In preferred embodiments, the stagnant fluid has at least one point of contact with the heated surface of the device. In preferred embodiments, the stagnant fluid has a large value of the volumetric thermal coefficient. As used herein, a “fluid” refers to a continuous amorphous substance that tends to flow and to conform to the outline of a container, such as a liquid or a gas, and may be used in accordance with the present invention. As used herein, the term “stagnant fluid” refers generally to a fluid that is not circulating or flowing and in preferred embodiments, the stagnant fluid can be surrounded by a flexible seal and/or the surfaces of a device such that the average translational velocity of the fluid is zero.
As used herein, “primary fluid” refers to the fluid that the devices of the present invention control or modulate its flow rate or its temperature. As used herein, “secondary fluid” refers to an auxiliary fluid utilized in the present invention to achieve additional control and modulation features for the primary fluid flow rate and temperature. As provided herein, the stagnant fluid in the complex flexible seals can have characteristics that are the same as or different from the characteristics of the primary fluid, the secondary fluid, or both. As used herein, “biofluid” refers to the fluid that contains at least one species of a biological substance that needs to be measured. As provided herein, the primary fluid can be a biofluid.
The flexible seals and flexible complex seals of the present invention are typically found between a first substrate and a second substrate of a thin film or other thermal systems such as the insulating assemblies. As used herein, “substrate” includes plates which may be inflexible or flexible according to part 6 herein below. In some preferred embodiments, the elastic modulus for the seals of the present invention, the ratio of the applied stress on the seal to the induced strain, range from about 103N/m2 to about 107N/m2. The seals of the present invention may comprise at least one closed cavity of a fluid such as air or the like in order to minimize their effective elastic modulus. The deformation of the flexible seals of the present invention can be guided by special guiders to attain maximum or desired deformations. In preferred embodiments, the flexible seals comprise different cross-sectional geometries, such as circular cross-section, rectangular cross-section and the like. As used herein, “thin films” include fluidic devices that have the thickness of their fluidic layers of an order of about a millimeter or less such as, microchannels and microfluidic devices. Thin films comprise at least two substrates, lower and upper substrates, and at least one fluidic layer. As used herein, an “insulating assembly” means an assembly of at least two insulating substrates and at least one fluid layer placed consecutively in series.
The flexible seals and flexible complex seals of the present invention are typically found between a first substrate and a second substrate of a thin film or other thermal systems such as the insulating assemblies. As used herein, “substrate” includes plates which may be inflexible or flexible according to part 6 herein below. In some preferred embodiments, the elastic modulus for the seals of the present invention, the ratio of the applied stress on the seal to the induced strain, range from about 103N/m2 to about 107N/m2. The seals of the present invention may comprise at least one closed cavity of a fluid such as air or the like in order to minimize their effective elastic modulus. The deformation of the flexible seals of the present invention can be guided by special guiders to attain maximum or desired deformations. In preferred embodiments, the flexible seals comprise different cross-sectional geometries, such as circular cross-section, rectangular cross-section and the like. As used herein, “thin films” include fluidic devices that have the thickness of their fluidic layers of an order of about a millimeter or less such as, microchannels and microfluidic devices. Thin films comprise at least two substrates, lower and upper substrates, and at least one fluidic layer. As used herein, an “insulating assembly” means an assembly of at least two insulating substrates and at least one fluid layer placed consecutively in series.
As disclosed herein, modulating the thermal characteristics of a device may be conducted by modifying the thin film thickness, the thermal load, the flow rate, or a combination thereof. For example, additional cooling can be achieved if the thin film thickness is allowed to increase by an increase in the thermal load, pressure gradient or both which will cause the coolant flow rate to increase. As provided herein, the enhancement in the cooling due to the flexible complex seals used is substantial at larger thermal loads for stagnant liquids while this enhancement is much larger at lower temperatures for stagnant fluids, especially ideal gases. This is because the volumetric thermal expansion coefficient increases for liquids and decreases for gases as the temperature increases. Moreover, the enhancement in the cooling due to flexible seals is substantial at larger pressure gradients for single layered thin films while it is significant for double layered thin films at lower pressure gradients.
Khaled and Vafai analyzed the enhancement in the heat transfer inside thin films supported by flexible complex seals. See Khaled & Vafai (2003) ASME J. of Heat Transfer 125:916-925, which is herein incorporated by reference. Specifically, the applied thermal load was considered to vary periodically with time in order to investigate the behavior of expandable thin film systems in the presence of a noise in the applied thermal load. As provided herein, a noticeable enhancement in the cooling capacity can be achieved for large thermal loads especially in cooling of high flux electronic components (q≈700 kW/m2) since they produce elevated working temperatures. Also, the generated squeezing effects at the mobile and inflexible substrate can be minimized when nanofluids are employed in the coolant flow. As used herein, “nanofluids” are mixtures of a working fluid, such as water, and suspended ultrafine particles in the fluid such as copper, aluminum, or the like with diameters of an order of about the nanometer range. See Eastman et al. (2001) Applied Physics Letters 78: 718-720, which is herein incorporated by reference.
The flexible seals, flexible complex seals, or both of the present invention may be used in two-layered thin films in order to regulate the flow rate of the primary fluid layer such that excessive heating in the secondary fluid layer results in a reduction in the primary fluid flow rate. For example, the flexible seals, flexible complex seals, or both of the present invention may be applied in the internal combustion industry where the fuel flow rate should be reduced as the engine gets overheated. In this example, the primary fluid flow is the fuel flow while the secondary fluid flow can be either flow of combustion products, flow of engine coolant or flow of any other auxiliary fluid. The flexible seals, flexible complex seals, or both of the present invention may be used to modulate or control exit thermal conditions in devices comprising two-layered thin films. For example, the flexible seals, flexible complex seals, or both of the present invention may be used to minimize bimaterial effects of various biosensors, including microcantilever based biosensors, which are sensitive to flow temperatures. See Fritz et al. (2000) Science 288:316-318, which is herein incorporated by reference. In this example, the primary fluid flow is flow of a biofluid while the secondary fluid flow can be either flow of the external surrounding fluid or flow of any auxiliary fluid.
As provided herein, thin films comprising flexible seals, flexible complex seals, or both are modeled and designed in order to alleviate the thermal load or modulate the flow. These systems according to the present invention provide noticeable control of the flow rate, reduce thermal gradients within the primary fluid layer at relatively large external thermal loads, and minimize fluctuation at the mobile and inflexible substrate in the presence of nanofluids.
As disclosed herein, the present invention provides a method for modulating or controlling the insulating properties of a device, an insulating assembly having insulating substrates separated by fluid layers and flexible seals. The fluid layers were supported by flexible seals in order to allow for volumetric thermal expansion of the primary fluid layers while the secondary fluid layers are vented to the atmosphere such that the secondary fluid pressure remains constant. The volumetric thermal expansion of the primary fluid layers within the insulating assembly were determined taking into consideration the variation in the fluid pressure due to the elastic behavior of the supporting flexible seals. The volumetric thermal expansion of the primary fluid layers was correlated to the increase in the equivalent thermal resistance of the fluid layers. The volumetric thermal expansion of the primary fluid was found to approach its isobaric condition value as the primary fluid layer thickness decreases. Also, the insulating properties were found to be enhanced when the primary fluid had a minumum thermal conductivity and when relatively high temperatures were experienced. The insulating properties deteriorate at large temperatures when the primary fluid has a relatively large thermal conductivity.
The following Table 1 provides the various symbols and meanings used in this section:
Generally, thermal losses increase at large working temperatures. The present invention provides a device that has desirable insulative attributes even at high working temperatures. That is, the present invention better conserves thermal energy especially at high temperatures as compared to similar devices that do not comprise flexible seals. An example of a device of the present invention is shown in
When the operating temperature (high temperature source) increases, the average fluid temperature of the primary fluid layer increases. Accordingly, the volume of the primary fluid layer expands accompanied by a shrinkage in the secondary fluid layer. As such, an increase in the equivalent thermal resistance of the insulating assembly can be attained as long as the thermal conductivity of the primary fluid layer is smaller than that for the secondary fluid layer. Preferably, the heated substrate has a relatively small thickness and a relatively large thermal conductivity so that the thermal expansion of the primary fluid layer is maximized.
Forces on elastic materials, such as seals, are usually proportional to the elongation of this material. See R. L. Norton (1998) M
wherein
T is the average temperature of the primary fluid layer;
K* is the stiffness of the supporting seals;
AS is the surface area of the intermediate insulating substrate;
ho is the reference thickness of the primary fluid layer;
Δh1 is the corresponding expansion in the primary fluid layer thickness;
m1 is the mass of the primary fluid; and
R1 is the primary fluid constant.
The first term on the left hand side of Equation 1 represents the pressure inside the primary fluid layer. The reference thickness ho corresponds to the thickness of the primary fluid layer when the primary fluid pressure is equal to the atmospheric pressure. Equation 1 can be solved for Δh1 and the expansion is found to be:
wherein
and
C
2
=
n
1
R
1
T
*h
o
2
−
atmAS
*ho
Eq.4
In order to maximize the expansion in the primary fluid layer which in turn results in better insulating properties, i.e. increased effective thermal resistance of the insulating assembly, the parameter C2 needs to be maximized. This can be accomplished by considering minimum values of K*ho while the following relationship provided in Equation 5 is preferred to be satisfied:
The following parameters were considered for studying the flexible seals of the present invention: K*=48000 N/m, AS=0.0036 m2 and patm=0.1 Mpa. The parameter m1R1 was evaluated at the reference condition when the primary fluid pressure was equal to the atmospheric pressure. This condition which causes the expansion to be zero in Equation 1 was assumed to be at T=To=283K and ho=0.004 m. This leads to m1R1=5.088×10−3J/K. Accordingly, the relation between the volumetric thermal expansion of the primary fluid layer and its average temperature is illustrated in
Equation 2 reduces to the following linearized model for relatively low volumetric thermal expansion levels
where To is the average temperature of the primary fluid layer at the reference condition. The reference condition corresponds to the condition that produces a zero net force on the seals. That is, thermal expansion is zero when the primary fluid layer is kept at To. At this condition, the primary fluid layer thickness is ho. The relative volumetric thermal expansion, Δh1/ho, approximated by Equation 6 is similar to that for isobaric expansion with the average primary fluid temperature being increased by the parameter
This parameter is denoted as ΔTo.
The error associated with Equation 6 is further reduced if
The latter inequality means that the insulating system exhibits relatively large volumetric thermal expansion by having a small increase in the primary fluid pressure due to the elastic behavior of the flexible seal.
The efficiency of the volumetric thermal expansion CF of the primary fluid layer is defined as the ratio of the expansion in the primary fluid layer when the flexible seal is present to the expansion when under constant pressure as expressed in the following Equation 7:
wherein (Δh1)Isobaric/ho=(T−To)/To. For the linearized model shown in Equation 6, the efficiency CF will be:
According to Equation 8, the values of CF which approaches unity as ΔTo decreases are provided for various ΔTo in Table 2 as follows:
The equivalent thermal resistance of the fluid layers during volumetric thermal expansion is given by the following Equation 9:
wherein
k1 is the thermal conductivity of the primary fluid; and
k2 is the thermal conductivity of the secondary fluid.
Both fluid layers are assumed to have a similar thickness prior to thermal expansion equal to ho. Based on Equation 1 and Equation 3, the increase in the equivalent thermal resistance ΔRth, the third part on the right of Equation 9, was correlated to the relative expansion in the primary fluid layer according to the following Equation 10:
wherein Rtho is the equivalent thermal resistance of both layers prior to thermal expansion.
The parameter Rtho is the sum of the first two terms on the right of Equation 9. When the parameter ηR is positive, the thermal resistance of the insulating assembly increases while it decreases as it becomes negative. Therefore, Rtho represents the dimensionless increase in the thermal resistance. Various properties of different gases are provided in the following Table 3:
According to Table 3, xenon can be used to enhance the insulating properties while helium is preferable to deteriorate the insulating properties especially at large operating temperatures as can be noticed from the last column in Table 3.
In the following analysis, the temperature at the lower side of the primary fluid layer was assumed to be kept under T1. See
wherein
hc is the convective heat transfer coefficient at the lower temperature side; and
T∞ is the temperature of environment facing the lower temperature side.
The surface area of the insulating assembly that faces the seal is relatively small. Therefore, the heat transfer through the seal portion is neglected in Equation 11 and Equation 12. For the previous example along with hc=5 W/m2K, T∞=275 K and Kins=0.04 W/mK, the temperature Te as a function of T1 is illustrated in
For the insulating assembly shown in
wherein Δh1/ho1 can be shown to be equal to the following Equation 13b:
where in
ho1 is the reference primary fluid layer thickness;
ho2 is the reference secondary fluid layer thickness;
(hins)i is the thickness of the ith insulating substrate;
(kins)i the thermal conductivity of the ith insulating substrate;
To is the primary fluid layer temperature that causes the primary fluid pressure to be equal to the atmospheric pressure; and
T*1 represents the average primary fluid layer temperature.
The parameter T*1 can be measured experimentally or determined theoretically using an iterative scheme. Equation 13a is based on the assumption that the heat transfer through the flexible seals is negligible when compared to the total heat transferred through the insulating assembly.
The solution of Equation 13a and Equation 13b can be used to produce pertinent engineering correlations. For example, percentage difference between the heat flux including thermal expansion effects and the heat flux at reference condition, qref, where thermal expansion is ignored, and correlated to T1, Te, To, k1 and ΔTo. The obtained family of correlations has the following functional form:
wherein a, b, c, d, e, m, n and the correlation coefficient R2 for different values of ho1 are listed in Table 4 as follows:
This correlation was obtained over the following range of parameter variations: 310<T1<400 K, 270<To<290 K, 50<ΔTo<150 K, 270<Te<300 K, 0.001<k1<0.017 W/m K, ho2=ho1,
and k2=0.028 W/m K.
1F. Examples of Insulating Assemblies with Maximum Enhanced Insulating Properties
As provided herein, the effects of both external squeezing and internal pressure pulsations were studied on flow and heat transfer inside non-isothermal and incompressible thin films supported by flexible seals. The laminar governing equations were non-dimensionalized and reduced to simpler forms. The upper substrate (mobile and inflexible substrate) displacement was related to the internal pressure through the elastic behavior of the supporting seals. The following parameters: squeezing number, squeezing frequency, frequency of pulsations, fixation number (for the seal) and the thermal squeezing parameter are the main controlling parameters. Accordingly, their influences on flow and heat transfer inside disturbed thin films were determined and analyzed. As provided herein, an increase in the fixation number results in more cooling and a decrease in the average temperature values of the primary fluid layer. Also, an increase in the squeezing number decreases the turbulence level at the upper substrate. Furthermore, fluctuations in the heat transfer and the fluid temperatures may be maximized at relatively lower frequency of internal pressure pulsations.
The following Table 5 provides the various symbols and meanings used in this section:
In certain thin film applications, external disturbances, such as unbalances in rotating machines or pulsations in external ambient pressures due to many disturbances, can result in an oscillatory motion at the upper substrate boundary. In addition to external disturbances, internal pressure pulsations such as irregularities in the pumping process, can produce similar oscillatory motion. Even small disturbances on the substrates of the thin film can have a substantial impact on the cooling process as the thickness of thin films is very small. These disturbances are even more pronounced if the thin film is supported by flexible seals. Accordingly, the dynamics and thermal characterization of thin films will be altered.
The chambers for chemical and biological detection systems such as fluidic cells for chemical or biological microcantilever probes are examples of thin films. See Lavrik et al. (2001) Biomedical Devices 3(12):35-44, which is herein incorporated by reference. Small turbulence levels that can be introduced into these cells by either flow pulsating at the inlet or external noise that may be present at the boundaries which result in a vibrating boundary can produce flow instabilities inside the fluidic cells. These disturbances substantially effect the measurements of biological probes, such as microcantilevers which are very sensitive to flow conditions.
The flow inside squeezed thin films, such as the flow inside isothermal oscillatory squeezed films with fluid density varying according to the pressure, has been studied. See Langlois (1962) Quarterly of Applied Math. XX:131-150, which is herein incorporated by reference. The heat transfer inside squeezed thin films (not oscillatory type) has been analyzed. See Hamza (1992) J. Phys. D: Appl. Phys. 25:1425-1431, Bhattacharyya et al. (1996) Numerical Heat Transfer, Part A 30:519-532, and Debbaut (2001) J. Non-Newtonian Fluid Mech. 98:15-31, which are herein incorporated by reference. The flow and heat transfer inside incompressible oscillatory squeezed thin films has been analyzed. See Khaled & Vafai (2002) Numerical Heat Transfer Part A 41:451-467, which is herein incorporated by reference. The effects of internal pressure pulsations have been studied on flow and heat transfer inside channels. See Hemida et al. (2002) Int. J. Heat Mass Transfer 45:1767-1780, and Joshi et al. (1985) J. Fluid Mech. 156:291-300, which are herein incorporated by reference.
Unfortunately, the prior art fails to account for the effects of both internal and external pressure pulsations on flow and heat transfer inside thin films, wherein the gap thickness will be a function of both pulsations.
Therefore, as provided herein, the upper substrate of a thin film was considered to be subjected to both external squeezing effects and the internal pressure pulsations. The influence of internal pressure pulsations on the displacement of the upper substrate was determined by the theory of linear elasticity applied to the seal supporting the substrates of an incompressible non-isothermal thin film. The laminar governing equations for flow and heat transfer were properly non-dimensionalized and reduced into simpler equations. The resulting equations were then solved numerically to determine the effects of external squeezing, internal pressure pulsations and the strength of the seal on the turbulence inside the disturbed thin films as well as on thermal characteristics of these thin films.
A two dimensional thin film that has a small thickness, h, compared to its length, B, was considered. The x-axis was taken in the direction of the length of the thin film while y-axis was taken along the thickness as shown in
h=ho(1−βcos(γωt)) Eq. 15
wherein
γ is the dimensionless frequency;
β is the dimensionless upper substrate motion amplitude; and
ω is a reference frequency.
The fluid was assumed to be Newtonian with constant properties.
The general two-dimensional continuity, momentum, and energy equations for the laminar thin film are given as follows:
wherein
T is the fluid temperature;
ρ is the density;
p is the pressure;
μ is the dynamic viscosity;
cp is the specific heat; and
k is the thermal conductivity of the fluid.
Equations 16-19 are non-dimensionalized using the following dimensionless variables:
wherein
T1 is the inlet temperature of the fluid; and
Vo is a constant representing a reference dimensional velocity.
As provided in the above equations, ΔT is equal to T2−T1 for constant wall temperature conditions (CWT), T2 will be the temperature of both lower and upper substrates, and is equal to
for uniform wall heat flux conditions (UHF). The variables X, Y, τ, U, V, Π and θ are the dimensionless forms of x, y, t, u, v, p and T variables, respectively. The above transformations except for dimensionless temperature have been used in the art along with the perturbation parameter ε
See Langlois (1962) Quarterly of Applied Math. XX:131-150, which is herein incorporated by reference.
Most flows inside thin films are laminar and could be creep flows especially in lubrications and biological applications. Therefore, the low Reynolds numbers flow model was adopted here. The application of this model to Equations 16-19 results in the following reduced non-dimensionalized equations:
wherein
σ is the squeezing number; and
PS is the thermal squeezing parameter.
The squeezing number and the thermal squeezing parameter are defined as:
The inlet dimensionless pulsating pressure is considered to have the following relation:
Πi=Πo+βp sin(γpωt+φp) Eq. 26
wherein
βp is the dimensionless amplitude in the pressure;
πi is the inlet dimensionless pressure;
πo is the mean dimensionless pressure;
γp is the dimensionless frequency of the pressure pulsations parameter; and
φp is a phase shift angle parameter.
Due to both pulsations in internal pressure and external disturbances, the dimensionless film thickness H, (H=h/ho), can be represented by Equation 27 by noting the principle of superposition:
H=1−βcos(γωt)+Hp Eq. 27
wherein Hp is the dimensionless deformation of the seals resulting from pulsations in the internal pressure.
The lower substrate was assumed to be fixed (immobile and inflexible substrate) and that the upper substrate (mobile and inflexible substrate) of the thin film is rigid such that the magnitude of the deformation in the seals is similar to displacement of the upper substrate (mobile and inflexible substrate). The dimensionless deformation in the seals due to variations in the external pressure is the second term of Equation 27 on the right. The dimensionless frequency γ is allowed to be different than γp.
The dimensionless pressure gradient inside the thin film as a result of the solution to the Reynolds Equation 23 is:
The reference velocity Vo that was used to define the dimensionless pressure, axial dimensionless velocity and the squeezing number was taken to be related to the average velocity, um, inside the thin film at zero β and βp and the dimensionless thickness of the thin film that results from the application of the corresponding inlet mean pressure, Hm, through the following relation:
The previous scaled reference velocity is only a function of the mean pressure, viscosity and the reference dimensions of the thin film and results in the following relation between the inlet mean dimensionless pressure to the squeezing number:
Πo=12−σ Eq. 30
Accordingly, the dimensionless pressure gradient, the dimensionless pressure and the average dimensionless pressure ΠAVG inside the thin film were related to the squeezing number through the following equations:
The displacement of the upper substrate due internal pressure pulsations was related to the ΠAVG through the theory of linear elasticity by the following relation:
H=FnΠAVG Eq. 34
wherein Fn is equal to:
The parameters E and ds in the previous equation are the modulus of elasticity of the flexible seals of the present invention and a characteristic dimension for the seal, respectively. The quantity ds is equal to the effective diameter of the seal's cross section times the ratio of the length of the seals divided by the thin film width. The effective diameter for seals having square cross section is equal to ho. The term Fn will be called the fixation number of the thin film.
The fixation parameter Fn represents a ratio between viscous shear force inside thin films to the elastic forces of the flexible seals. Moreover, Equation 34 is based on the assumption that transient behavior of the seal's deformation is negligible. The values of Fn are about 0.001 to about 0.1 for long thin films supported by flexible seals.
The first set of dimensionless boundary conditions used were for constant wall temperatures (CWT) at both the lower and the upper substrates while the second set used assumed that the lower substrate was at uniform wall heat flux conditions (UHF) and the upper substrate is insulated. As such the dimensionless boundary conditions can be written as:
The last condition of Equation 36 is based on the assumption that the flow at the exit of the thin film is thermally fully developed. Moreover, the last thermal condition of Equation 37 was derived based on an integral energy balance at the exit of the thin film realizing that the axial conduction is negligible at the exit. The calculated thermal parameters considered were the Nusselt numbers at the lower and upper substrates, and the dimensionless heat transfer from the upper and lower substrates, Θ, for CWT conditions, which are defined according to the following equations:
wherein θm and Um are the dimensionless mean bulk temperature and the dimensionless average velocity at a given section and are defined as follows:
Due to symmetric flow and thermal conditions for CWT, Nusselt numbers at lower and upper substrates were expected to be equal.
The dimensionless thickness of the thin film was determined by solving Equations 27, 33 and 34 simultaneously. Accordingly, the velocity field, U and V, was determined from Equations 21 and 22. The reduced energy equation, Equation 24, was then solved using the Alternative Direction Implicit techniques (ADI) known in the art by transferring the problem to one with constant boundaries using the following transformations: τ*=τ, ξ=X and
Iterative solution was employed for the ξ-sweep of the energy equation for CWT conditions so that both the energy equation and the exit thermal condition, last condition of Equation 36, are satisfied. The values of 0.008, 0.03, 0.002 were chosen for Δξ, Δη and Δτ*.
The effects of pressure pulsations on H are clearly seen for large values of Fn as shown in
For σ=12 where the time average of the average gage pressure inside the thin film is zero, the variation in H decreases as Fn increases. This effect can be seen from Equation 33 and Equation 34 and will cause reductions in the flow and in the cooling process. However, the mean value of ΠAVG is always greater than zero for other values of σ which causes an increase in the mean value of H as Fn increases resulting in an increase in the mean value of the flow rate inside the thin film.
Flow and heat transfer inside externally oscillatory squeezed thin films supported by flexible seals in the presence of inlet internal pressure pulsations were analyzed. The governing laminar continuity, momentum and energy equations were properly non-dimensionalized and reduced to simpler forms for small Reynolds numbers. The reduced equations were solved by the alternative direction implicit (ADI) method. The turbulence level at the upper substrate increases by increases in both the fixation number and the frequency of the internal pressure pulsations. However, an increase in the squeezing number decreases the turbulence level at the upper substrate. The fluid temperatures and the corresponding fluctuations were found to decrease when the fixation number and the thermal squeezing parameter were increased for both CWT and UHF conditions. Finally, fluctuations in the heat transfer and the fluid temperatures were more pronounced at lower frequency of internal pressure pulsations.
Although thin films are characterized by having laminar flows with relatively low Reynolds numbers leading to stable hydrodynamic performance, the thickness of the thin films is small enough such that small disturbances at one of the boundaries may cause a significant squeezing effect at the boundary. See e.g. Langlois (1962) Quarterly of Applied Math. XX:131-150 (flow inside isothermal oscillatory squeezed films with fluid density varying with the pressure), Khaled & Vafai (2002) Numerical Heat Transfer, Part A 41:451-467 and Khaled & Vafai (2003) Int. J. Heat and Mass Transfer 46:631-641 (flow and heat transfer inside incompressible thin films having a prescribed oscillatory squeezing at one of their boundaries), and Khaled & Vafai (2002) Int. J. Heat and Mass Transfer 45:5107-5115 (internal pressure through the elastic behavior of the supporting seal), which are herein incorporated by reference.
Recently, the situation where the squeezing effect at the free substrate is initiated by thermal effects was studied. See Khaled & Vafai (2003) ASME J. Heat Transfer 125:916-925, which is herein incorporated by reference. As provided herein, flexible seals with closed cavities of stagnant fluids having a relatively large volumetric thermal expansion coefficient, flexible complex seal, were studied. Flexible complex seals in a single layer thin film can cause flooding of the coolant when the thermal load of the thin film is increased over its projected capacity. As a result, an enhancement in the cooling process is attained especially if ultrafine suspensions are present in the coolant, a fluid that exhibits high heat transfer performance. Ultrafine suspensions in the fluid such as copper or aluminum particles with diameters of order nanometer are found to enhance the effective thermal conductivity of the fluid. See Eastman et al. (2001) Applied Physics Letters 78: 718-720, which is herein incorporated by reference.
As provided herein, the flow and heat transfer inside an oscillatory disturbed two-layered thin film channel supported by flexible complex seals in the presence of suspended ultrafine particles was studied. Oscillatory generic disturbances were imposed on the two-layered thin film channels supported by flexible complex seals in the presence of suspended ultrafine particles, which correspond to disturbances in the upper substrate temperature and in the inlet pressure of the secondary fluid layer. The governing continuity, momentum and energy equations for both layers were non-dimensionalized and categorized for small Reynolds numbers and negligible axial conduction. The deformation of the supporting seals was linearly related to both the pressure difference across the two layers and the upper substrate's temperature based on the theory of the linear elasticity and the principle of the volumetric thermal expansion of the stagnant fluid filling the closed cavities of the flexible complex seals.
As provided herein, the flow rate and heat transfer in the main thin film channel can be increased by an increase in the softness of the seals, the thermal squeezing parameter, the thermal dispersion effect and the total thickness of two-layered thin film. However, the flow rate and heat transfer in the main thin film channel decrease as the dimensionless thermal expansion coefficient of the seals and the squeezing number of the primary fluid layer increase. Both the increase in thermal dispersion and the thermal squeezing parameter for the secondary fluid layer were found to increase the stability of the intermediate or the mobile and inflexible substrate. Furthermore, the two-layered thin film channel was found to be more stable when the secondary fluid flow was free of pulsations or it had relatively a large pulsating frequency. Finally, the proposed two-layered thin film supported by flexible complex seals, unlike other controlling systems, does not require additional mechanical control or external cooling devices, i.e. is self-regulating for the flow rate and temperature of a primary fluid layer.
The following Table 6 provides the various symbols and meanings used in this section:
The present invention provides flexible complex seals. The flexible complex seals may be used in two-layered thin films are utilized in order to regulate the flow rate of the primary fluid layer such that excessive heating in the second layer results in a reduction in the primary fluid flow rate. The flexible complex seals of the present invention may be used in internal combustion applications where the fuel flow rate should be reduced as an engine gets overheated. The flexible complex seals of the present invention may be used to minimize bimaterial effects of many biosensors that are sensitive to heat and flow conditions. See Fritz et al. (2000) Science 288:316-318, which is herein incorporated by reference.
The heat flux of the upper substrate can be independent of the primary fluid flow or can be the result of external processes utilizing the primary fluid flow as in combustion processes. The latter can be used for controlling the primary fluid flow conditions while the former may model the increase in the ambient temperature in a fluidic cell application, thereby preventing an increase in the average fluid temperature in an ordinary fluidic cell avoiding a malfunctioning of a device such as a biosensor.
The sealing assembly of the upper layer contains flexible complex seals, closed cavities filled with a stagnant fluid having a relatively large volumetric thermal expansion coefficient. The upper layer also contains flexible seals in order to allow the intermediate substrate to move in the normal direction. Any excessive heating at the upper substrate results in an increase in the upper substrate's temperature such that the stagnant fluid becomes warmer and expands. This expansion along with the increase in inlet pressure in the upper layer, if present, causes the intermediate substrate to move downward. Thus, a compression in the film thickness of the lower layer is attained resulting in reduction in mass flow rate within the primary fluid flow compartment. This insulating assembly may be used to control combustion rates since part of the excessive heating and increased pressure due to deteriorated combustion conditions can be utilized to prescribe the heat flux at the upper substrate. Thus, the flow rate of the fuel in the primary fluid layer can be reduced and combustion is controlled.
In fluidic cells, excessive heating at the upper substrate causes compression to the primary fluid layer's thickness. Thus, average velocity in the primary fluid layer increases, when operated at constant flow rates, enhancing the convective heat transfer coefficient. This causes the average fluid temperature to approach the lower substrate temperature, thereby reducing the bimaterial effects. When it is operated at a constant pressure or at a constant velocity, the compression of the primary fluid layer due to excessive heating at the upper substrate reduces the flow rate. Thus, the fluid temperatures approach the lower substrate temperature at a shorter distance. As such, bimaterial effects are also reduced. The flexible seals can be placed between guiders as shown in
As provided herein, upper and lower thin films that have small thicknesses h1 and h2, respectively, compared to their length B and their width D1 and D2, respectively, were analyzed. The x-axis for each layer is taken along the axial direction of the thin film while y-axis for each layer is taken along its thickness as shown in
Both lower and upper substrates were assumed to be fixed (immobile and inflexible substrates) while the intermediate substrate was free to move only in the normal direction due to the use of flexible complex seals (mobile and inflexible substrate). The generic motion of the intermediate substrate due to both variations of the stagnant fluid temperature in the secondary fluid flow passage and the induced internal pressure pulsations within both primary fluid and secondary fluid flow passages is expressed according to the following Equation 41:
wherein
ho is a reference thickness for the primary fluid passage;
H1 is the dimensionless motion of the intermediate substrate;
HT is the dimensionless motion of the intermediate substrate due to the volumetric thermal expansion of the stagnant fluid; and
Hp is the dimensionless motion of the intermediate substrate due to the deformation in seals as a result of the internal pressure.
The fluid was assumed to be Newtonian having constant average properties except for the thermal conductivity. The general two-dimensional continuity, momentum and energy equations for a laminar thin film are given as follows:
wherein
T is the fluid temperature;
u is the dimensional axial velocity;
v is the dimensional normal velocity;
ρ is the average fluid density;
p is pressure;
μ is the average fluid dynamic viscosity;
cp is the average specific heat of the fluid; and
k is the thermal conductivity of the fluid.
When the fluid contains suspended ultrafine particles, these properties will be for the resulting dilute mixture so long as the diameter of the particles is very small compared to ho. The index “i” is “1” when analyzing the primary fluid layer while it is “2” when analyzing the secondary fluid layer. Equations 42-45 are non-dimensionalized using the following dimensionless variables:
wherein
ω is the reference frequency of the disturbance;
T1o is the inlet temperature for the primary fluid flow;
T2o is the inlet temperature for the secondary fluid flow;
Tw is the lower substrate temperature;
pe is the reference pressure which represents the exit pressure for both layers;
qo is the reference heat flux at the upper substrate;
k2o is the stagnant thermal conductivity of the secondary fluid;
Voi is the reference dimensional velocity for the lower layer;
Vo2 is the reference dimensional velocity for the upper layer; and
ε is the perturbation parameter,
The prescribed heat at the upper substrate, qu, as well as the dimensionless inlet pressure, π2n, for the secondary fluid flow vary according to the following generic relationships:
q
u
=q
o
+βq sin
ωt
Eq. 47
Π2n=Π2o+βp sin
pωt+φp
Eq. 48
wherein
βq is the dimensionless amplitude of upper substrate's heat flux;
βp is the dimensionless amplitude for the inlet pressure for the secondary fluid flow;
γ is the dimensionless frequency for the upper substrate heat flux; and
γp is the dimensionless frequency for the inlet pressure for the secondary fluid layer.
The variables Xi, Yi, τ, Ui, Vi, Πi and θi are the dimensionless forms of xi, yi, t, ui, vi, pi and Ti variables, respectively.
For the two-layered thin film shown in
wherein (πAVG)1 and (πAVG)2 are the average dimensionless pressure in the primary fluid and the secondary fluid layers, respectively. The parameter E*i will be referred to as the softness index of the supporting seal in layers “1” or “2” and will be denoted as E* when E*1=E*2. It has the following functional form:
wherein K* is the effective stiffness of the seals that support the intermediate substrate. The dimensionless parameter L* is introduced to account for the elastic contribution of the intermediate substrate in the calculation of the displacement.
As provided herein, the analysis was performed for relatively small thermal load frequencies in order to ascertain that squeezing generated flows are in the laminar regime. For these frequencies, Equation 49 was applicable and the inertia effect of the intermediate substrate was negligible. Moreover, the increase in the thickness due to a pressure increase in the thin film causes a reduction in the stagnant fluid pressure. This action stiffens the insulating assembly. Therefore, the stiffness K* was considered to be the effective stiffness for the insulating assembly and not for the seal itself. From the practical point of view, the closed cavity width G was taken to be large enough such that a small increase in the stagnant fluid pressure due to the thermal expansion can support the associated increase in the elastic force on the seal.
The dimensionless displacement of the intermediate substrate due to the thermal expansion was related to the dimensionless average temperature of the upper substrate, u
AVG, by the following linearized model:
H
T
=−F
T(θu)AVG Eq. 51
wherein FT is named the dimensionless thermal expansion parameter and is equal to:
The coefficient A* depends on the closed cavities dimensions and their geometry. The parameter βT is the volumetric thermal expansion coefficient of the stagnant fluid in its approximate form:
evaluated at the pressure ps1 corresponding to the stagnant fluid pressure in the closed cavities when the secondary fluid flow temperature was kept at inlet temperature of the secondary fluid layer T2o. The closed cavity volumes Vso, Vs1 and Vs represent the closed cavity volume at the reference condition (h2=ho), the closed cavity volume when the pressure in the closed cavities is ps1 and the closed cavity volume at normal operating conditions where the average stagnant fluid temperature is Ts, respectively. The factor CF represents the volumetric thermal expansion correction factor. This factor was introduced in order to account for the increase in the stagnant pressure due to the increase in the elastic force in the seal during the expansion which tends to decrease the effective volumetric thermal expansion coefficient. It approaches one as the closed cavity width G increases and it can be determined theoretically using methods known in the art.
The parameter FT is enhanced at elevated temperatures for liquids and at lower temperatures for gases because βT increases for liquids and decreases for gases as the temperature increases. Dimensionless thermal expansion parameter is further enhanced by a decrease in ko, an increase in qo, an increase in E*i or an increase in ho. Equation 51 is based on the assumption that the stagnant fluid temperature is similar to the average upper substrate temperature since closed cavity surfaces were considered insulated except for the region facing the upper substrate in order to provide a maximum volumetric thermal expansion to the closed cavities. Moreover, the heat flux on the upper substrate was assumed to be applied to the portion that faces the secondary fluid flow.
The thermal conductivity of the fluid was considered to vary with the flow speed in order to account for thermal dispersion effects when suspended ultrafine particles were present in the secondary fluid flow. Induced squeezing effects at the intermediate substrate due to time variations in the thermal load or inlet pulsative pressures were expected to enhance the heat transfer inside fluid layers due to thermal dispersion effects. To account for this increase, a linear model between the effective thermal conductivity and the fluid speed was utilized as provided by Equation 53. See Xuan & Roetzel (2000) Int. J. Heat and Mass Transfer 43:3701-3707, which is herein incorporated by reference.
k
i
i
,Y
i,τ=
o
+λi√{square root over (U2
i,Yi,τ
+Λi2V2
i,Yi,τ
=)}
o
φi
i,Yi,τ
Eq.53
where λi and Λi are the dimensionless thermal dispersion coefficient and reference squeezing to lateral velocity ratio which are:
The coefficient C* depends on the diameter of the ultrafine particle, its volume fraction and both fluid and the particle properties. The parameter (ρcp)fi is the density times the specific heat of the fluid resulting from the mixture of the pure fluid and the ultrafine particles suspensions within the ith layer while (ko)i is the stagnant thermal conductivity of the working fluid in the ith layer that contains ultrafine particles. This stagnant thermal conductivity is usually greater than the thermal conductivity of the pure fluid. See Eastman et al. (2001) Applied Physics Letters 78:718-720, which is herein incorporated by reference. All the fluid properties that appear in Equations 42-45 should be replaced by the effective mixture properties which are functions of the pure fluid and the particles and that the diameter of the ultrafine particles are so small that the resulting mixture behaves as a continuum fluid. See Xuan & Roetzel (2000) Int. J. Heat and Mass Transfer 43:3701-3707, which is herein incorporated by reference.
Flows inside thin films are in laminar regime and could be considered creep flows in certain applications as in lubrication and biological applications. Therefore, the low Reynolds numbers flow model was adopted and applied to Equations 42-44 and the results of dimensionalizing the energy equation result in the following reduced non-dimensionalized equations:
The axial diffusion term in the dimensionalized energy equation, Equation 59, is eliminated because it is of order ε2. The parameters σi and (PS)i are called the squeezing number and the thermal squeezing parameter, respectively, and are defined as:
The dimensionless thickness of the lower layer and the upper layer are defined as:
The reference thickness ho can be determined using the force balance across the intermediate substrate due to the flow exit pressures of both layers at static conditions using methods known in the art. The reference thickness ho can be controlled by either varying flow exit pressures for each layer prior injecting of both flows, by a proper selection to the undistorted thickness of the supporting seals in each layer or by using both, according to methods known in the art. Therefore, the dimensionless thicknesses H1 and H2 are related to each other through the following relation as both lower and upper substrates are fixed (immobile and inflexible substrates):
H
1
+H
2
=H
t Eq. 62
wherein Ht is a constant representing the dimensionless total thickness of the two-layered thin film.
Two conditions will be imposed for the inlet flow rate of the primary fluid layer. In applications that require minimizations of thermal effects due to an increase in heat transfer from the environment such as for fluidic cells of biological and chemical sensing devices, the inlet flow rate for the lower layer is assumed to be constant and referred to as the CIF condition. However, constant inlet pressure was assumed to model flow of fluids in combustion applications such as flow of fuel prior to the mixing section and is referred as the CIP condition. The previously defined reference velocities Vo1 and Vo2 represent the velocity in the flow passages at zero values of the parameters E*1, E*2 and FT. Accordingly, the inlet dimensionless pressures vary with the squeezing numbers according to following relations for the CIP condition:
Π1n=12−σ1 Eq. 63
Π2n=−σ2
+βp sin
pτ+φp
Eq. 64
Therefore, the solution of the Reynolds equations for the CIP condition will reveal the following relationships for the dimensionless pressure gradient, the dimensionless pressure and the average dimensionless pressure ΠAVG inside lower and upper layers:
For the CIF condition, the dimensionless pressure gradient, the dimensionless pressure and the average dimensionless pressure ΠAVG inside lower layer were changed to the following:
The dimensionless initial and thermal boundary conditions for the previously defined problem were taken as follows:
Based on physical conditions, the intermediate substrate was taken to be insulated and the Nusselt number at the lower and the upper substrates are defined as:
wherein hcl and hcu, are the convective heat transfer coefficients for the lower and upper substrates, respectively.
The quantities θim and Uim are the sectional dimensionless mean bulk temperature and the dimensionless average velocity for the ith layer and are given as:
wherein Uim, is the dimensionless average velocity at a given section for the ith layer. For the primary fluid passage, the dimensionless heat flux at a given section is defined as follows:
The obtained dimensionless film thickness for the primary fluid layer H1 can be used to determine the dimensionless flow rate of the fluid in the primary fluid passage at the mid section for the CIP condition. The latter is an important parameter should be controlled and is referred to as ΨX=0.5 where X=0.5 denotes the location at X1=0.5. This parameter can be calculated from the following relation:
wherein QX=0.5 is the dimensional flow rate at X=0.5 in the main thin film.
The procedure for the numerical solution is summarized as follows:
1. Initially, a value for HT is assumed.
2. The dimensionless thicknesses for the lower and upper layers H1 and H2 are determined by solving Equations 41, 49, 62, 69, and 70 simultaneously, using an explicit formulation. The velocity field, Ui and Vi is then determined from Equations 55, 56, 65, and 66.
3. Reduced energy equations, Equation 59, are solved by first transferring them to a constant boundary domain using the following transformations: τ*=τ, ξl=Xi and
Tri-diagonal algorithm was implemented along with a marching scheme. See Blottner (1970) AIAA J. 8:193-205, which is herein incorporated by reference. Backward differencing was chosen for the axial convective and transient terms and central differencing was selected for the derivatives with respect to ηi. The values of 0.008, 0.03, 0.001 were chosen for Δξi, Δηi and Δτ*, respectively.
4. HT is updated from Equation 51 and steps (2) to (4) is repeated until:
5. The solution for the flow and heat transfer inside the two layers is determined.
6. Time is advanced by Δτ* and steps (1) to (5) are repeated.
Numerical investigations were performed using different mesh sizes and time steps to assess and ascertain grid and time step independent results. Any reduction in the values of Δξ, Δη and Δτ* below Δξ=0.008, Δη=0.03 and Δτ*=0.001 cause less than about 0.2 percent error in the results.
The maximum value of the parameters PS is chosen to be 1.0. Beyond this value, the error associated with the low Reynolds number model will increase for moderate values of the dimensionless thermal expansion parameter, softness index of the seals, and the Prandtl number. As an example, the order of transient and convective terms in the momentum equations is expected to be less than 5.0 percent that of the diffusive terms for PS=1.0, Pr=6.7, E*1=E*2=0.3, FT=0.15, βq=0.2 and σ1=3.0, σ2=6.0. The parameters correspond, for example, to a main thin film filled with water and having B=D=60 mm, ho=0.3 mm, ω=1.7s−7, Vo=0.1 m/s and K*=33000 N/m.
Ideal gases produce about a 15 percent increase in the closed cavity volume under typical room conditions for a 45° C. temperature difference. Further, about a 60 percent increase in the convective heat transfer coefficient for about a 2 percent volume fraction of copper ultrafine particles has been reported. See Li & Xuan (2002) Science in China (Series E) 45:408-416, which is herein incorporated by reference. Accordingly, the parameters FT and λ2 were varied until comparable changes have been attained in the dimensionless thin film thickness and the Nusselt number.
For the CIP condition, the increase in the dimensionless thermal expansion parameter FT of the upper flexible complex seals causes a reduction in ΨX=0.5 values and an increase in the disturbance at intermediate substrate. Consequently, the parameters ΘAVG and (θu)AVG decrease as FT increases. These observations are shown in
As the squeezing number for the primary fluid flow passage increases, the net pressure force on the intermediate substrate decreases as dictated by Equation 49. Therefore, the primary fluid layer film thickness decreases causing a reduction in the values of ΨX=0.5, ΘAVG and (θu)AVG as shown in
The increase in the value of the thermal squeezing parameter PS2 of the upper layer causes an enhancement in the upper substrate cooling as shown by reductions in (θu)AVG in
Due to their random motions, ultrafine particles tend to increase the heat exchange within the fluid causing the thermal dispersion effect. Therefore, as the dimensionless thermal dispersion parameter λ increases, the thermal conductivity increases causing the upper substrate temperature (θu)AVG to decrease. Thus, in turn, the values of ΨX=0.5 and ΘAVG are increased while variations in dH1/dτ are decreased as λ increases. See
wherein the maximum and minimum values corresponds to the steady periodic values.
It should be noted that ΔΨX=0.5 and ΔΘAVG are unpredictable at relatively lower frequencies of pulsations and the primary fluid layer becomes more stable for large values of γp. See
As provided herein, flow and heat transfer inside thin films supported by flexible complex seals, flexible seals having closed cavities of a stagnant fluid possessing a large coefficient of volumetric thermal expansion βT, were studied in the presence of suspended ultrafine particles and under periodically varying thermal load conditions. The governing continuity, momentum and energy equations are non-dimensionalized and reduced to simpler forms. The deformation of the seal is related to the internal pressure and lower substrate's temperature based on the theory of linear elasticity and a linearized model for thermal expansion. As provided herein, enhancements in the cooling may be achieved by an increase in the volumetric thermal expansion coefficient, thermal load, thermal dispersion effects, softness of the supporting seals and the thermal capacitance of the coolant fluid. Further, thermal dispersion effects were found to increase the stability of the thin film. The noise in the thermal load was found to affect the amplitude of the thin film thickness, Nusselt number and the lower substrate temperature; however, it had a negligible effect on the mean values.
Thin films are widely used in cooling of many heating sources such as electronic components. These elements are used in thin films in cooling systems such as in flat heat pipes or microchannel heat sinks. See Moon et al. (2000) Int. J. Microcircuits and Electronic Packaging 23:488-493, Fedorov & Viskanta (2000) Int. J. Heat and Mass Transfer 43:399-415, and Zhu & Vafai (1999) Int. J. Heat and Mass Transfer 42:2287-2297, which are herein incorporated by reference. A two phase flow in microchannel is capable of removing maximum heat fluxes generated by electronic packages yet the system may become unstable near certain operating conditions. See Bowers & Mudawar (1994) ASME J. Electronic Packaging 116:290-305, which is herein incorporated by reference. Further, the use of porous medium in cooling of electronic devices was found to enhance heat transfer due to increases in the effective surface area. See Hadim (1994) ASME J. Heat Transfer 116:465-472, which is herein incorporated by reference. However, the porous medium creates a substantial increase in the pressure drop inside the thin film.
As provided herein, additional cooling can be achieved if the thin film thickness is allowed to increase by an increase in the thermal load which will cause the coolant flow rate to increase using flexible complex seals of the present invention, i.e. flexible seals having closed cavities of a stagnant fluid having a large value of the volumetric thermal expansion coefficient βT.
In the presence of periodic external thermal loads, the thickness of a thin film supported by a flexible complex seal is expected to be periodic. This is because the stagnant fluid expands during maximum thermal load intervals allowing for a relaxation in the thin film thickness which causes a flooding of the coolant. On the other hand, the thin film is squeezed during minimum thermal loads intervals due to the contraction in the stagnant fluid in the closed cavities of the flexible complex seals.
One of the advantages of using flexible complex seals is that the increase in the coolant flow rate because of thermal expansion effects produces an additional cooling in the presence of suspended ultrafine particles. See Li & Xuan (2002) Science in China (Series E) 45:408-416, which is herein incorporated by reference. This is because the chaotic movement of the ultrafine particles, the thermal dispersion, increases with the flow where it is modeled in the energy equation by introducing an effective thermal conductivity of the coolant. See Xuan & Roetzel (2000) Int. J. Heat and Mass Transfer 43:3701-3707, which is herein incorporated by reference. Further, large fluctuation rates that can be generated in the flow during severe squeezing conditions tend to increase the chaotic motions of the particles in the fluid which increases the energy transport in the coolant.
As provided herein, the enhancement in the cooling process inside thin films supported by flexible complex seals in the presence of suspended ultrafine particles was analyzed. The lower substrate of the examined thin film was considered to be under a periodically varying heat flux. The thin film thickness was related to the thermal load and the internal pressure through the volumetric thermal expansion coefficient of the stagnant fluid and the theory of linear elasticity applied to the supporting seals. The governing equations for flow and heat transfer were properly non-dimensionalized and reduced into simpler equations for low Reynolds numbers. The resulting equations were then solved numerically to determine the effects of the thermal load, volumetric thermal expansion coefficient of the stagnant fluid, the softness of the seal, thermal capacitance of the working fluid and the squeezing number on the dynamics and thermal characteristic of the thin films supported by flexible complex thin films. As provided herein, the flexible complex seals of the present invention are useful in enhancing the cooling and can be used for additional purposes such as for diagnosing functions for heating sources so long as they possess large thermal expansion coefficient.
The following Table 7 provides the various symbols and meanings used in this section:
The analysis is concerned with a thin film that has a small thickness h compared to its length B and its width D. Therefore, a two-dimensional flow is assumed. The x-axis was taken along the axial direction of the thin film while y-axis was taken along its thickness as shown in
The lower substrate of the thin film was assumed to be fixed (immobile and inflexible substrate) and in contact with or adjacent to a heating source while the upper substrate was attached to the lower substrate by flexible complex seals allowing it to expand (mobile and inflexible substrate). The motion of the upper substrate due to both internal variations in the stagnant fluid temperature and the induced internal pressure pulsations as a result of oscillating thermal loads is expressed according to the following relation:
wherein
h is the thin film thickness;
ho is a reference film thickness;
H is the dimensionless thin film thickness;
HT is the dimensionless motion of the upper substrate due to the thermal expansion of the stagnant fluid; and
Hp is the dimensionless motion of the upper substrate as a result of the deformation of seals due to the average internal pressure of the working fluid.
The fluid is assumed to be Newtonian having constant average properties except for the thermal conductivity. The general two-dimensional continuity, momentum and energy equations for a laminar flow of the working fluid inside the thin film are given as:
wherein
T is temperature;
u is the dimensional axial velocity;
v is the dimensional normal velocity;
ρ is the average density;
p is pressure;
μ is the average dynamic viscosity;
cp is the average specific heat; and
k is the thermal conductivity.
The previous fluid properties are for the pure working fluid in the case where the fluid is free from any suspensions. In the presence of suspended ultrafine particles, the previous properties will be for an approximated new continuum fluid composed from the mixture of the pure fluid and the suspensions. See Xuan & Roetzel (2000) Int. J. Heat and Mass Transfer 43:3701-3707, which is herein incorporated by reference. The new properties of the mixture are related to the fluid and the particle properties through the volume fraction of the suspended particles inside the thin film and the thermal dispersion parameter.
The following dimensionless variables were used to non-dimensionalized Equations 81-84:
wherein ω, T1, pe, qo and Vo are the reference frequency of thermal load, inlet temperature of the fluid, a constant representing the exit pressure, reference heat flux and a constant representing a reference dimensional velocity, respectively. The term ko corresponds to the working fluid thermal conductivity in the absence of any suspensions while it is the stagnant thermal conductivity, free from the dispersion term, for the dilute mixture between the fluid and the ultrafine suspensions. The stagnant thermal conductivity has usually an enhanced value when compared to that of the pure fluid for metallic particles. See Eastman et al. (2001) Applied Physics Letters 78:718-720, which is herein incorporated by reference.
The upper substrate is assumed to be insulated to simplify the analysis and that the lower substrate was subjected to a periodically varying wall heat flux qL condition according to the following relation:
q
L
=q
o(1=βq sin(βωt)) Eq. 86
wherein βq and β are the dimensionless amplitude of the lower substrate's heat flux and a dimensionless frequency, respectively. The variables X, Y, τ, U, V, Π and θ are the dimensionless forms of x, y, t, u, v, p and T variables, respectively. The parameter ε appearing in Equation 85f is the perturbation parameter,
For the thin film shown in
Hp=FnπAVG Eq. 87
This is based on the fact that the upper substrate is assumed to be rigid and that the applied force on an elastic material, the flexible seal, is assumed to behave as an elastic material, is proportional to the elongation of this material. See Norton (1998) Machine Design; An Integrated Approach Prentice-Hall, New Jersey, which is herein incorporated by reference. The parameter Fn is referred to as the fixation parameter and is a measure of the softness of the seal, flexible seals have large Fn values, and is equal to:
wherein E and ds are the effective modulus of elasticity for the complex seal and a characteristic parameter which depends on the seal's dimensions and the thin film width D, respectively. The quantity ds is equal to the effective dimension of the seal's cross section times the ratio of the total length of the seal divided by the thin film width D. The seal is considered to have isotropic properties. Further, the effective dimension of the seals times their total length represents the contact area between the seals and the upper or lower substrates when the seals have a rectangular cross section as shown in
As provided herein, the analysis was performed for relatively small thermal load frequencies in order to ascertain that squeezing generated flows have relatively small Reynolds numbers. For these frequencies, Equation 87 is applicable and the inertia effect of the upper substrate is negligible. Moreover, the increase in the thickness due to a pressure increase in the thin film causes a reduction in the stagnant fluid pressure. This action stiffens the insulating assembly. Therefore, the parameter E is considered to be the effective modulus of elasticity for the insulating assembly not for the seal itself. Practically, the closed cavity width G is assumed to be large enough such that a small increase in the stagnant fluid pressure due to the expansion can support the associated increase in the elastic force on the seal.
The dimensionless displacement of the upper substrate due to thermal expansion is related to the dimensionless average temperature of the lower substrate, (θW)AVG, by the following linearized model:
H
T
=F
T(θW)AVG Eq. 89
wherein FT is named the dimensionless thermal expansion parameter and is equal to:
wherein A* is a constant depending on the closed cavities dimensions and geometry. The parameter βT is the volumetric thermal expansion coefficient of the stagnant fluid in its approximate form: βT (1/VSo)[(VS−VS1)/(Ts-T1)]|ps1 evaluated at the pressure ps1 corresponding to the stagnant fluid pressure at the inlet temperature T1. The quantities VS1 and VS represent the closed cavity volumes at normal operating conditions when the stagnant fluid is at T1 and at the present stagnant fluid temperature Ts, respectively. The parameter VSo represents the closed cavity volume at the reference condition. The factor CF represents the volumetric thermal expansion correction factor. This factor was introduced in order to account for the increase in the stagnant pressure due to the increase in the elastic force in the seal during the expansion which tends to decrease the effective volumetric thermal expansion coefficient. It approaches one as the closed cavity width G increases and may be determined theoretically using methods known in the art.
The parameter FT is enhanced at elevated temperatures for liquids and at lower temperature for gases because βT increases for liquids and decreases for gases as the stagnant temperature increases. Dimensionless thermal expansion parameter is also enhanced by a decrease in ko, an increase in qo, an increase in Fn, or by increases in ho. Equation 89 is based on the assumption that the stagnant fluid temperature is similar to the lower substrate temperature since entire closed cavity surfaces were considered insulated except that facing the lower substrate. Furthermore, the heat flux of the heating source is applied on the portion of the lower substrate that is facing the working fluid. The other portion which faces the seals is taken to be isolated from the heating source and the environment to minimize the variation in the lower substrate temperature along the width direction.
In the presence of suspended ultrafine particles in the working fluid, the thermal conductivity of the working fluid composed from the pure fluid and suspensions is expected to vary due to the thermal dispersion. To account for these variations, the following model which is similar to the Xuan & Roetzel ((2000) Int. J. Heat and Mass Transfer 43:3701-3707) model that linearly relates the effective thermal conductivity of the working fluid to the fluid speed is utilized:
k(X,Y,τ)=ko=λ√{square root over (U2
,Y,τ
+Λ2V2
,Y,τ
)}=koφ(X,Y,τ) Eq. 91
wherein λ and Λ are the dimensionless thermal dispersion coefficient and the reference squeezing to lateral velocity ratio which are:
λ=C*cp
ho
o+ωB
Eq. 92a
wherein C* is the coefficient of the thermal dispersion which depends on the diameter of the ultrafine particles, its volume fraction (ratio of the particles volume to the total thin film volume), and both fluid and ultrafine particles properties. Ultrafine particles include particles that are extremely small compared with the thickness of the thin film.
The coefficient C* is expected to increase by an increase in the diameter of the particles, their volume fraction, their surface roughness and the working fluid Prandtl number, Pr=(ρcpυ)/ko. On the other hand, the stagnant thermal conductivity ko increases with an increase in both the volume fraction and the surface area of the particles. A dilute mixture of ultrafine suspensions and water produce no significant change in the pressure drop compared to pure water which reveals that the viscosity is a weak function of the fluid dispersion for a dilute mixture.
Generally, flows inside thin films are in laminar regime and could be creep flows as in lubrication. Therefore, the low Reynolds numbers (the modified lateral Reynolds number ReL=(Voho)ε/υ and the squeezing Reynolds number Res=o2ω
υ) flow model was used herein. These insulating assemblies neglect the transient and convective terms in momentum equations, Equations 82 and 83. These terms become incomparable to the pressure gradient and diffusive terms for small squeezing frequencies and reference velocities. Application of these insulating assemblies to Equations 82-84 and the outcome of dimensionalizing the energy equation, Equation 85, result in the following reduced non-dimensionalized equations:
Note that Equation 96 is based on the assumption that the axial conduction is negligible when compared to the transverse conduction. The parameters σ and PS are referred to as the squeezing number and the thermal squeezing parameter, respectively, and are defined as:
Both inlet and exit dimensionless pressures were assumed constant and the following relationship was obtained between the inlet dimensionless pressure and the squeezing number based on the assumption that the reference velocity Vo represents the average velocity in the thin film at zero values of FT and Fn:
Πi=12−σ Eq. 98
Accordingly, the dimensionless pressure gradient, the dimensionless pressure and the average dimensionless pressure ΠAVG inside the thin film are related to the squeezing number through the following equations:
The dimensionless thermal boundary conditions for the previously defined problem are taken as follows:
Based on the physical conditions, the Nusselt number is defined as:
The parameter θm is the dimensionless mean bulk temperature and is given as:
wherein Um is the dimensionless average velocity at a given section.
The procedure for the numerical solution is summarized as follows:
1. Initially, a value for HT is assumed.
2. At the present time, the dimensionless thickness of the thin film H is determined by solving Equations 80, 87, 88, and 101 simultaneously, using an explicit formulation. The velocity field, U and V, is then determined from Equations 93, 94, and 99.
3. At the present time, the reduced energy equation, Equation 96, is transferred into one with constant boundaries using the following transformations: τ*=τ, ξ=X and
A tri-diagonal solution was implemented along with a marching scheme. See Blottner (1970) AIAA J. 8:193-205, which is herein incorporated by reference. Backward differencing was chosen for the axial convective and transient terms and central differencing was selected for the derivatives with respect to η. The values of 0.008, 0.03, 0.001 were chosen for Δξ, Δη and Δτ*, respectively.
4. HT is updated from Equation 89 and steps (2) to (4) is repeated until:
5. The converged solution for the flow and heat transfer inside the thin film is determined at the present time.
6. Time is advanced by Δτ* and steps (1) to (5) are repeated.
Numerical investigations were performed using different mesh sizes and time steps to assess and ascertain grid and time step independent results. Any reduction in the values of Δξ, Δη and Δτ* below Δξ=0.008, Δη=0.03 and Δτ*=0.001 results in less than about a 0.2 percent error in the results.
In the results, the maximum value of the parameters PS is chosen to be 1.0. Beyond this value, the error associated with the low Reynolds number model will increase for moderate values of the dimensionless thermal expansion parameter, fixation parameter, and the Prandtl number. As an example, the order of transient and convective terms in the momentum equations were found to be less 1.0 percent that of the diffusive terms for PS=1.0, Pr=6.0, Fn=0.05, FT=0.25, βq=0.1 and σ=6.0. The parameters correspond, for example, to a thin film filled with water and having B=D=60 mm, ho=0.3 mm, ds=0.5 mm, ω=2.0s−1, Vo=0.12 m/s and E=2(105)pa.
Ideal gases produce about a 15 percent increase in the closed cavity volume at room conditions for a 45° C. maximum temperature difference. Further, about a 60 percent increase in the convective heat transfer coefficient for a volume fraction of copper ultrafine particles of about a 2.0 percent has been reported. See Li & Xuan (2002) Science in China (Series E) 45:408-416, which is herein incorporated by reference. Accordingly, the parameters FT and λ were varied until comparable changes have been attained in the dimensionless thin film thickness and the Nusselt number.
Due to the reduction in the lower substrates temperatures as λ increases, the dimensionless thin film thickness decreases as λ increases as depicted in
The reduction in thermal resistance across the transverse direction when λ increases causes the temperature profiles to be more flattened as λ increases as seen in
As provided herein, the effects of both fluid leakage and wall slip conditions were studied analytically and numerically on the fluctuation rate in the flow inside non-isothermal disturbed thin films supported by flexible seals within a fluidic cell. Flow disturbances due to internal pressure pulsations and external squeezing are considered in this work. The main controlling parameters are the dimensionless leakage parameter, softness of the seal, squeezing number, dimensionless slip parameter, the thermal squeezing parameter and the power law index. Accordingly, their influences on the fluctuation rate and heat transfer characteristics inside disturbed thin films were determined. As provided herein, an increase in the dimensionless leakage parameter, softness of the seal-upper substrate assembly and the wall slip parameter result in more cooling and an increase in the fluctuation level in the flow. However, an increase in the squeezing number and the fluid power index decreases decrease flow fluctuations.
Thin films are utilized in various chemical and biological systems such as in biosensing devices. See Lavrik et al. (2001) Biomedical Microdevices 3(1):35-44, which is herein incorporated by reference. These biosensing devices have the advantage to accurately, quickly, and economically screen patients for the presence of various diseases or can be used to detect many bio-warfare agents. Many biosensors in the art comprise at least one microcantilever, wherein detection of a desired agent is based on the deflection of the free end of the microcantilever that is caused by the imposed stresses at least one of its surfaces. See U.S. Pat. No. 7,288,404, issued on Oct. 30, 2007, which is herein incorporated by reference. This surface stress is due to the reaction, interaction, or binding between a given agent in a fluid sample inside the thin film and a second agent, such as a receptor, that reacts, interacts, or binds with the given agent, that is immobilized on the surface of the microcantilever.
Examples of reactions in biomolecular (receptor/analyte) applications which occur within a fluidic cell include: antibody-antigen (receptor/analyte) bindings or DNA hybridization of a pair of DNA strands (receptor/analyte) having complementary sequences, and the like. An example of antibody-antigen bindings includes the binding of polyclonal anti-PSA (prostate-specific antigen) antibody and free PSA (fPSA). See Wu et al. (2001) Nature Biotechnology 19:856-860, which is herein incorporated by reference. In many cases, disturbances exist in the flow which can disturb the deflection of the microcantilever and produce a noise in the measurement. See Fritz et al. (2000) Science 288:316-318, which is herein incorporated by reference.
Part of the noise in the measurement is due to the fact that oscillations in the flow may produce an oscillatory drag force on the microcantilever surface causing it to vibrate. For example, a 100 nm deflection in the microcantilever due to initial flow disturbances in the fluidic cell while the microcantilever deflection due to receptor/analyte binding was of the order of 10 nm has been reported. See Fritz et al. (2000) Science 288:316-318, which is herein incorporated by reference.
Meanwhile, flow oscillations may change the microcantilever temperature causing it to produce an additional noise where the microcantilever is composed of two layers (bimaterial) having different coefficients of thermal expansion. For example, microcantilevers having a 50 nm deflection due to bimetallic effects, which was five times the microcantilever deflection due receptor/analyte binding, has been reported. See Fritz et al. (2000) Science 288:316-318, which is herein incorporated by reference. The rate of receptor/analyte binding changes with the flow velocity has been demonstrated. See Prichard et al. (1995) J. Biomechanics 28:1459-1469, which is herein incorporated by reference. As flow oscillations add extra noise due to surface stresses, minimizations of flow oscillations in fluidic cells are desired and may be achieved according to the present invention.
Flow disturbances can be due to external disturbances or due to internal pressure pulsations when the pumping process is irregular. Even a small change in the internal pressure of the fluidic cell can have a substantial impact since the thickness of the thin film is very small. The impact is more pronounced if the thin film is supported by flexible seals. Accordingly, the dynamics and thermal characterization of thin films will be altered producing a noise in the biosensor measurement which is proportional to the fluctuation rate in the flow. Another source for flow disturbances is the flow leakage which can seriously affect the operation of the microcantilever. See Raiteri et al. (2000) Electrochimica Acta 46:157-163, which is herein incorporated by reference.
Flow inside oscillatory disturbed thin films has been studied. See Langlois (1962) Quarterly of Applied Math. XX:131-150, which is herein incorporated by reference. Laminar pulsating flows has been studied as well as effects of internal pressure pulsations on oscillatory squeezed laminar flow and heat transfer inside thin films supported by flexible seals. See Hemida et al. (2002) Int. J. Heat and Mass Transfer 45:1767-1780 and Khaled & Vafai (2002) Int. J. Heat and Mass transfer 45:5107-5115, which are herein incorporated by reference. Unfortunately, the prior art fails to analyze the effects of fluid leakage on pulsating flow and heat transfer inside thin films in the presence of flexible seals. Such an analysis and understanding is important as the effects of fluid leakage contribute to flow disturbances.
Thus, as provided herein the effects of fluid leakage on pulsating flow and heat transfer inside thin films in the presence of flexible seals were analyzed. Further, as provided herein, flow inside disturbed fluidic cells under wall slip conditions with different fluid types is analyzed in order to determine their best operating conditions that cause minimum flow fluctuations. Wall slip conditions can be achieved either when the fluid contains suspensions or when the substrates are coated with water repellent resigns. See Watanabe & Udagawa (2001) AIChE J. 47:256-262, which is herein incorporated by reference. Also, wall slip occurs when the size of the thin film is so small that the Kundsen number, a ratio of the molecular mean free path to the characteristic length of the fluidic cell, is between about 10−3 to about 10−1 as for flow of gases in microchannels. See Shiping & Ameel (2001) Int. J. Heat and Mass Transfer 44:4225-4234, which is herein incorporated by reference. Thus, as provided herein flow and heat transfer associated with side leakage, wall slip condition and non-Newtonian effects inside disturbed thin films supported by flexible seals are analytically and numerically examined in order to provide improved fluidic cell designs.
A two-dimensional thin film fluidic cell that has a small thickness h compared to its length 2B and its width D was considered. The inlet of this fluidic cell is taken to be at its center forming a symmetrical fluidic cell, as shown in
The lower substrate of the thin film is assumed to be fixed or immobilized (immobile and inflexible substrate) while the upper substrate is attached to the lower substrate by flexible seals and therefore capable of movement (mobile and inflexible substrate). The average dimensionless motion of the upper substrate H is expressed according to the following relation:
wherein h, ho and Hp are the dimensional average thin film thickness, a reference thin film thickness, and the average dimensionless change in the film thickness due to internal pressure forces, respectively.
The following dimensionless variables will be utilized in the analysis herein:
wherein ω, T1, pe, Vo, μ, k, and ε are the reference frequency of internal pulsations, inlet temperature of the fluid, a constant representing the exit pressure, a constant representing a reference dimensional velocity, dynamic viscosity of the fluid, thermal conductivity of the fluid and the perturbation parameter (ε=ho/B), respectively. The pressure at the exit and the outside pressure were assumed to be at the exit pressure. The lower substrate is maintained at a uniform wall heat flux condition qo. The variables t, u, v, w, p and T are the time, axial velocity component, normal velocity component, lateral velocity component, pressure and the temperature, respectively. The dimensionless variables X, Y, Z, τ, U, V, W, Π and θ are the dimensionless forms of x, y, z, t, u, v, w, p and T variables, respectively.
The average dimensionless change in the film thickness was related to the average dimensionless pressure inside the thin film fluidic cell ΠAVG through the theory of linear elasticity and assumes that the change in the pressure force on the upper substrate is linearly proportional to the average change in the thin film thickness by the following relation:
HP=FnπAVG Eq. 108
wherein Fn is named, the fixation parameter. A larger Fn value indicates softer seal-upper substrate assembly. See Boresi et al. (1978) Advanced Mechanics of Materials Wiley, NY, which is herein incorporated by reference. The inertia of the upper substrate is negligible because the frequency of pulsations is usually small. The fixation parameter Fn that appears in Equation 108 is equal to:
wherein E and hs are the effective modulus of elasticity and the effective dimension of the seal (hs=ho for a square seal cross section), respectively. The factor K* reflects the contribution of the elastic behavior of the upper substrate. The parameter Fn becomes apparent when the thin film thickness is very small (ho less than about 0.15 mm).
Most flows inside thin films possess relatively small Reynolds numbers and could be creep flows as in biological applications, i.e. the modified Reynolds numbers, ReL=Vohoε/υ and ReS=ho2ω/υ, are less than one. Therefore, the low Reynolds number flow model was adopted. Accordingly, the continuity and momentum equations for the flow inside the fluidic cell filled with a Newtonian fluid were reduced to the following non-dimensionalized equations along with the non-dimensionalized energy equation:
No slip conditions were assumed at the lower and the upper substrates of the fluidic cell as shown in Equation 110. The parameters σ and PS are called the squeezing number and thermal squeezing parameter, respectively, and are defined as:
According to Equation 112, the leakage inside the thin film is distributed equally on both sides of the thin film and it is relatively small. Thus, linearization of the lateral pressure gradient was used. As seen in Equation 112, side leakage is proportional to the pressure difference between internal and external (at Pe) pressures of the thin film. Equation 113 is the corresponding modified Reynolds equation of the problem. Equation 114 is applicable at the plane of symmetry at Z=0. The parameter ML in Equation 112 will be named the dimensionless leakage parameter and is related to the total leaked mass mL through the following relation:
The inlet pressure due to flow disturbances in the pumping process is considered to have the following relation:
Πi=Πo+βv sin(γωt)
Eq. 116
wherein βv, γ, πi and πo are the dimensionless amplitude in the pressure, dimensionless frequency of the pressure pulsations, inlet dimensionless pressure and the mean dimensionless inlet pressure, respectively. The solution to Equation 113 is obtained as:
The reference velocity Vo is taken to be the velocity inside the thin film in absence of any disturbance and it is related to Πo according to following relation:
Πo=12−σ Eq. 118
The effects of fluid slip at the boundaries and non-Newtonian effects in the presence of an external disturbance were analyzed. The dimensionless oscillation of upper substrate was based on the following generic relationship:
H=1−β cos(γτ) Eq. 119
wherein β and γ are the amplitude of the motion and a selected dimensionless frequency, respectively. The apparent viscosity μ of a non-Newtonian fluid such as a biofluid at low flow rates can be expressed according to the following power-law formula:
where n is a constant representing the power law index. As a result, axial momentum equation for creep flow reduces to the following, μo, replaces μ in Equation 107h:
According to the linear relationship between the wall slip velocity and the shear rate at a solid boundary, the dimensionless boundary conditions for the axial velocity at the substrates are:
wherein βp is the dimensional slip parameter. See Navier (1823) Mem. Acad. Sci. Inst. France 1:414-416, which is herein incorporated by reference. By solving Equation 120 and the continuity equation, the modified Reynolds equation is:
For a constant average inlet velocity condition Vo during the oscillations, Equation 120 can be used for determining the velocity field, U and V, for the lower half of the thin film (Y/H<0.5), which are found to be:
Accordingly, the fluid slip velocity at the wall is obtained as:
The upper substrate was assumed to be insulated while the lower substrate was maintained at a constant heat flux. Accordingly, the dimensionless thermal boundary and initial conditions are:
The dimensionless average thickness of the thin film for the leakage problem was determined by solving Equations 106 and 108 and the average of Equation 117, simultaneously, using the explicit formulation with respect to time. Accordingly, the velocity field U, V and W was determined. The energy equation, Equation 114, was then solved using the Alternative Direction Implicit (ADI) method known in the art by transferring the problem to one with constant boundaries using the following transformation: τ*=τ, ξ=X and
The used dimensionless parameters in the leakage problem were selected according to the following data from the literature: the estimated volume of the fluidic cell,
According to
When the upper substrate assembly employs a flexible seal as for large Fn values, the film thickness will be more sensitive to internal pressure pulsations. As a result, an increase in the fixation parameter Fn, causes an increase in the fluctuation rate at the upper substrate and consequently an increase in flow fluctuations is associated. See Equations 110-114 and 117 and
The resistance against the flow decreases as the dimensionless wall slip parameter βP/ho increases. Thus, the wall slip velocity increases as βP/ho increases. See
Due to the expected increase in wall shear stresses for pseudoplastic (n<1) fluids as the power law index n decreases, the wall slip velocity increases as n decreases. See
The reduction in internal pressures associated with an increase in the leakage rate results in an increase in the inlet flow rate which reduces the average dimensionless lower substrate temperature as seen in
In order to minimize axial, normal and lateral flow disturbances inside thin films, the parameters Fn, mL and β are minimized. Designs of these films as provided herein can satisfy these constraints. For example, the multi-compartment fluidic cell with multiple inlets shown in
Flow fluctuations within a fluidic cell and consequently the noise in the measurement due to flow disturbances, may be minimized by considering the following effects:
minimizing the working velocities;
maximizing the thickness of the upper substrate;
maximizing the thin film width if large leakage rate is involved;
minimizing the thin film width in the absence of leakage;
maximizing the perturbation parameter;
utilizing dilute working fluid; and
maximizing the thin film thickness.
The last three effects may increase the microcantilever deflection due to thermal effects. Thus, in preferred embodiments of the present invention, flow oscillations are reduced by employing fluidic cell designs provided herein.
6A. Systems with Increased Capacity as Thermal Load Increases
Thus, the present invention provides an upper substrate (flexible) assembly that can be bent or flexed in certain direction when exposed to heat. Such an upper substrate assembly is shown in
6B. Systems with Decreased Capacity as Thermal Load Increases
However, under very high operating thermal conditions, as in deteriorated combustions, the supporting seals may not work properly. Thus, the present invention provides an upper substrate (flexible) assembly that may be operated under high thermal conditions. Such a design is shown in
Generally, thermal losses increase at large working temperatures. The present invention provides an insulating assembly having desirable insulative attributes at high working temperatures. That is, its effective thermal resistance increases with an increase in the working temperatures. An example of an insulating assembly of the present invention is shown in
The mathematical modeling for the insulating assembly shown in
wherein To and T* are the original primary fluid layer temperature and the average primary fluid temperature, respectively. The quantity ΔTo is equal to Kho2/1R1
where m1, R1 and K are the mass of the primary fluid, the primary fluid constant and the stiffness of the supporting seals.
In addition, different plausible insulating assemblies which are compact and can provide additional enhancement to the insulating properties utilizing expandable fluid layers. An example of these are shown in
q={dot over (m)}
air
p
ir
2
−T
1
Eq. 128
wherein {dot over (m)}air, (cp)air, T1 and T2 are the mass flow rate of the heated air, specific heat of the air, the inlet mean bulk temperature of the air and the exit mean bulk temperature of the air, respectively. This value represents twice the heat transferred through each insulating assembly in
Experiments may be performed under various inlet air temperatures to investigate the enhancement in the insulating properties. A sample result expected from the experiment is shown in
The motion of the upper substrate shown in
wherein h, ho and H are the thin film thickness, a reference film thickness and the dimensionless thin film thickness, respectively. The variables HT and Hp are the dimensionless motion of the upper substrate due to the thermal expansion of the stagnant fluid and the dimensionless motion of the upper substrate as a result of the deformation of seals due to the average internal pressure of the working fluid, respectively.
The presence of a noise in the thermal load can result in a noise in dimensionless film thickness H which produces fluctuations in the flow rate due to squeezing effects. See Khaled & Vafai (2003) ASME J. Heat Transfer 125:916-925, which is herein incorporated by reference. The flow and heat transfer inside the expandable thin film can be simulated using an iterative procedure that results in solving the momentum and energy equations, Equation 130 and Equation 131, while satisfying Equation 129.
wherein V, T and p are the velocity field vector, temperature, and the fluid pressure, respectively, and ρ, μ, cp, and k are the primary fluid's density, primary fluid's absolute viscosity, primary fluid's specific heat, and the thermal conductivity of the primary fluid, respectively.
For the thin film shown in
wherein D, B, K and pe are the thin film width, thin film length, the effective stiffness of the sealing assembly, and the external pressure, respectively. This is based on the fact that the upper substrate is assumed to be rigid and that the applied force on an elastic material (the flexible seal) is proportional to the elongation of this material. See Norton (1998) Machine Design; An Integrated Approach Prentice-Hall, New Jersey, which is herein incorporated by reference.
The increase in the thickness due to a pressure increase in the thin film causes a reduction in the stagnant fluid pressure. This action stiffens the sealing assembly. Therefore, the parameter K is considered to be the effective stiffness for the sealing assembly and not for the seal itself. When the closed cavities are filled with an ideal gas, the effective K can be shown to be approximately equal to the following when the mass of the stagnant fluid is kept constant for the configuration shown in
wherein m, R and Ksm are the mass of the ideal fluid in the closed cavities, fluid constant and the stiffness for the pure seal material, respectively.
When check valves are used to ensure that the pressure does not fall below the initial stagnant pressure, K is expected to approach Ksm. Practically, the closed cavity width G is assumed to be large enough such that a small increase in the stagnant fluid pressure due to the expansion can support the associated increase in the elastic force of the seal. Moreover, the fixation parameter can be enhanced by replacing segments of the seals at different locations by elastic membranes especially the outermost ones thereby reducing the effective length of the seal.
The dimensionless displacement of the upper substrate due to thermal expansion is related to the difference between the average temperature of the heated substrate, (TW)AVG, and the initial stagnant fluid temperature T1 by the following linearized model:
H
T
=A*β
T
C
F[(TW)AVG−T1] Eq. 134
wherein A* is a constant depending on the closed cavities dimensions and geometry.
The parameter βT is the volumetric thermal expansion coefficient of the stagnant fluid in its approximate form: βT≈(1/Vs1)(Vs−Vs1)/(Ts−T1)|ps1 evaluated at the pressure ps1 corresponding to the stagnant fluid pressure at the inlet temperature T1. The quantities Vs1 and Vs represent the closed cavity volumes at T1 and at the present stagnant fluid temperature TS, respectively. The factor CF represents the volumetric thermal expansion correction factor. This factor is introduced to account for the increase in the stagnant pressure due to an increase in the elastic force in the seal during the expansion which tends to decrease the effective volumetric thermal expansion coefficient. It approaches one as the closed cavity width G increases and it needs to be determined theoretically. For ideal gases and assembly shown in
wherein hpm is the mean value for the dimensional film thickness prior thermal effects.
A model for evaluating different thermal loads is shown in
Two-layered thin films possess enhanced cooling capacity. See Vafai & Zhu (1999) Int. J. Heat and Mass Transfer 42:2287-2297, which is herein incorporated by reference. These two-layered systems also provide a passive control of flow and exit thermal conditions for the main thin film when flexible complex seals are separating the substrates of the two-layered thin film as shown in
The upper layer of the two-layered thin film shown in
The flexible complex seal of the upper layer contains closed cavities filled with a stagnant fluid having a relatively large volumetric thermal expansion coefficient. This sealing assembly also comprises flexible seals in order to allow the intermediate substrate to move in the normal direction. See
In fluidic cells, excessive heating at the upper substrate causes compression of the primary fluid layer's thickness. Thus, average velocity within the primary fluid layer increases, when operated at constant flow rates, enhancing the convective heat transfer coefficient. This causes the average fluid temperature to approach the lower substrate temperature, thereby reducing the bimaterial effects. When this cooling assembly is operated at a constant pressure or at a constant velocity, the compression of the primary fluid layer due to excessive heating at the upper substrate reduces the flow rate. Thus, the bulk fluid temperature approaches the lower substrate temperature within a shorter distance. As such, bimaterial effects are also reduced. Flexible seals can be placed between special guiders as shown in
Both lower and upper substrates were assumed to be fixed (immobile and inflexible substrates) while the intermediate substrate which was separated from the lower and upper substrates was free to move only in the normal direction due to the presence of flexible complex seals (mobile and inflexible substrate). The generic motion of the intermediate substrate due to both variations of the stagnant fluid temperature in the secondary fluid flow passage and the induced internal pressure pulsations within both the primary fluid and secondary fluid flow passages is expressed according to the following relationship:
wherein H1 (H1=h1/ho), (Tu)AVG and T1o are the dimensionless displacement of the intermediate substrate, average temperature at the upper substrate and the initial stagnant fluid temperature, respectively. The subscript “1” indicates the primary fluid layer while “2” indicates the secondary fluid layer. The flow and heat transfer inside the expandable two-layered thin film can be solved using an iterative procedure that results in solving the momentum and energy equations and satisfying Equation 134.
Heat transfer and flow induced by either natural or mixed convection inside vertical channels and open ended cells in the presence of flexible complex seals may be analyzed as provided herein. See
In this section, single layered (SL) and double layered (DL) flexible microchannel heat sinks are analyzed. The deformation of the supporting seals is related to the average internal pressure by theory of elasticity. It is found that sufficient cooling can be achieved using SL flexible microchannel heat sinks at lower pressure drop values for softer seals. Double layered flexible microchannel heat sinks provide higher rate of cooling over SL flexible microchannel heat sinks at the lower range of pressure drops. Single layered flexible microchannel heat sinks are preferred for large pressure drop applications while DL flexible microchannel heat sinks are preferred for applications involving low pressure drops.
The rapid development of microelectronics has created a need for large integration density of chips in digital devices such as VLSI components. These devices require increased current-voltage handling capabilities leading to large amount of dissipated heat within a small space. Microchannel heat sinks are one of the proposed methods that can be used to remove this excessive heating.
Microchannels have a very high heat transfer coefficient. Early works on microchannel heat sinks had shown that parallel micro passages with 50 μm wide and 302 μm deep had thermal resistances as low as 9×10−6K/(W/m2). See Tuckerman & Pease (1981) IEEE Electron Device Lett EDL-2:126-129. This value is substantially lower than the conventional channel sized heat sinks. See Missaggia, L. J., et al. (1989) IEEE J. Quantum Electronics 25:1988-1992; Kleiner, M. B., et al. (1995) IEEE Trans on Components, Packaging and Manufacturing Technology Part A 18:795-804; and Samalam, V. K. (1989) J. Electronics Materials 18:611-617. Microchannel heat sink devices can be used as single layered (SL) micro passage such as those illustrated in the works of Lee and Vafai and Fedorov and Viskanta. See Lee & Vafai (1999) Int. J. Heat and Mass Transfer 42:1555-1568 and Fedorov & Viskanta (2000) Int. J. Heat and Mass Transfer 43:399-415. Double layered (DL) microchannel heat sinks were introduced for the first time in the work of Vafai and Zhu to provide additional cooling capacity for the microchannel and to decrease the axial temperature gradients along the microchannel. See Vafai & Zhu (1999) Int. J. Heat and Mass Transfer 42:2287-2297. Single layered microchannel heat sinks can be either single channel system such as those analyzed in the work of Harms et. al. or multiple channel system. See Harms, T. M., et al. (1999) Int. J. Heat and Fluid Flow 20:149-157 and Lee & Vafai (1999) Int. J. Heat and Mass Transfer 42:1555-1568.
One of the drawbacks of microchannel heat sinks is the increased temperature of the coolant as large amount of heat is carried out by a relatively small amount of coolant. As such, new technologies developed in the works of Vafai and Zhu and Khaled and Vafai provides new solutions for cooling of electronic components utilizing microchannel heat sinks. See Vafai & Zhu (1999) Int. J. Heat and Mass Transfer 42:2287-2297; Khaled & Vafai (2002) Int. J. Heat and Mass Transfer 45:5107-5115; Khaled & Vafai (2003) ASME J. Heat Transfer 125:916-925; and Khaled & Vafai (2004) Int. J. Heat and Mass Transfer 47:1599-1611. The work of Khaled and Vafai is based on utilizing flexible soft seals. The resulting microchannel heat sink system is referred to as “flexible microchannel heat sink”. See Khaled & Vafai (2002) Int. J. Heat and Mass Transfer 45:5107-5115; Khaled & Vafai (2003) ASME J. Heat Transfer 125:916-925; and Khaled & Vafai (2004) Int. J. Heat and Mass Transfer 47:1599-1611. Khaled and Vafai demonstrated that additional cooling can be achieved if flexible thin films including flexible microchannel heat sinks are utilized. See Khaled & Vafai (2002) Int. J. Heat and Mass Transfer 45:5107-5115. In this work, the expansion of the flexible thin film including flexible microchannel heat sink is directly related to the internal pressure. Khaled and Vafai have demonstrated that significant cooling inside flexible flexible thin films can be achieved if the supporting seals contain closed cavities which are in contact with the heated surface. See Khaled & Vafai (2003) ASME J. Heat Transfer 125:916-925. They referred to this kind of sealing assembly as “flexible complex seals”. Moreover, Khaled and Vafai demonstrated that flexible complex seals along with thin films have important applications in design and control of the flow and thermal characteristics of these types of systems. See Khaled & Vafai (2004) Int. J. Heat and Mass Transfer 47:1599-1611.
In this work, the enhancement in the cooling process inside SL and DL flexible microchannel heat sinks is investigated. The theory of linear elasticity applied to the supporting seals is utilized to relate the average internal pressure to the thickness of the flexible microchannel heat sinks. The resulting equations are then solved numerically and analytically to determine the effects of the pressure drop, softness of the supporting seals, the Prandtl number and the coolant mass flow rate on the thermal characteristics of both SL and DL flexible microchannel heat sinks.
The following Table 8 provides the various symbols and meanings used in this section:
Consider flow inside a two dimensional microchannel heat sink with a height H and axial length B. The x-axis is aligned along the channel length while the y-axis is in the traverse direction as shown in
leads to the following dimensionless energy equation:
wherein q, T1 and Re are the heat flux at the heated plate, the inlet temperature and the Reynolds number (Re−(ρumH)/μ), respectively. Pr and ε are the Prandtl number (Pr=υ/α) and the perturbation parameter (ε=H/B). The mean velocity is related to the pressure drop across the channel, Δp, through the following relation:
wherein μ is the dynamic viscosity of the coolant.
For microchannel heat sinks supported by flexible soft seals, the separation between the microchannel's plates can be expressed according the following assuming that the seals are linear elastic materials:
wherein Ho, W, and K are a reference thickness of the microchannel heat sink, the width of the microchannel heat sink and the stiffness of the supporting seal, respectively. As such, the Reynolds number and the perturbation parameter can be expressed according to the following relations:
Re=Re
o
+Re
o
F
3 Eq. 141
ε=εo+ReoF
Eq. 142
wherein Reo and εo are the Reynolds number and the perturbation parameter evaluated at the reference microchannel thickness and the parameter F is the fixation parameter. These parameters are defined as:
The parameter Reo can be interpreted as the dimensionless pressure drop parameter. The temperature normalized with respect to the reference parameters, θ* is defined as follows:
The normalized mean bulk temperature, obtained from the solution of integral form of Eq. 138 is:
The uncertainty in (θ*)m, Δ(θ*)m, is:
U
=Δ
*
=U
Re o
ΔRe
o
+U
H
ΔF Eq. 148
wherein UReo, and UF are defined as:
The lower plate is assumed to have a uniform wall heat flux and the upper plate is considered to be insulated. As such the dimensionless boundary conditions can be written as:
The Nusselt number is defined as:
wherein (θ*)W is the heated plate temperature normalized with respect to the reference parameters. Under fully developed thermal conditions, Nusselt number approaches the following value:
wherein (θ*)W is the dimensionless lower plate temperature under fully developed thermal conditions. Thus, it can be expressed according to the following:
Minimizing this temperature at the exit results in the following value of the fixation parameter:
As such, the corresponding Reynolds number and the perturbation parameters are:
wherein the subscripts 1 and 2 are for the lower and the upper layers, respectively. The corresponding boundary conditions are:
The intermediate plate is taken to be made from a highly conductive material like copper such that temperature variation across this plate is negligible. The following parameters are introduced in order to compare the performance of the DL flexible microchannel heat sink compared to SL flexible microchannel heat sink:
Lower values of the cooling factors κm and κW indicate that DL flexible microchannel heat sinks are preferable over SL flexible microchannel heat sinks.
Another factor that will be considered is the ratio of the total friction force in DL flexible microchannel heat sinks to that for SL flexible microchannel heat sinks delivering the same flow rate of coolant. It can be shown that this factor is equal to:
wherein (Reo)DL and (Reo)SL are related through the following:
e
o
L
+
e
o
L
F
=2eo
L
+
eo
LF
Eq.164
As such, the delivered dimensionless mass flow rate by both SL and DL flexible microchannel heat sinks is:
wherein m is the dimensional mass delivered by both flexible microchannel heat sinks.
Equations 137, 158 and 159 were descritized using three points central differencing in the transverse direction while backward differencing was utilized for the temperature gradient in the axial direction. The resulting tri-diagonal system of algebraic equations at X=ΔX was then solved using the well established Thomas algorithm. See Blottner, F. G. (1970) AIAA J. 8:193-205. The same procedure was repeated for the consecutive X-values until X reached the value of unity. For equations 158 and 159, the temperature distribution at the intermediate plate was initially prescribed. Equations 158 and 159 were solved as described before. The thermal boundary condition at the intermediate plate was then used to correct for intermediate plate temperatures. The procedure was repeated until all the thermal boundary conditions were satisfied.
In most of the cases considered here, the minimum value of Re was taken to be 50 while the maximum Re value was allowed to expand to 2100 for Reo=50 and F=0.05. The maximum Re corresponded to a microchannel heat sink that was substantially expanded due to the presence of soft seals. The thickness for the latter limiting case (Re=2100) was found to be 3.5 times the thickness of the former limiting case (Re=50). The maximum fixation parameter was taken to be 0.05. This represented a thin film microchannel heat sink filled with water, having B=60 mm, W=20 mm ho=0.3 mm, and K=1000 N/m.
Convective heat transfer coefficient is reduced as F increases at low dimensionless pressure drops as shown in
Heat transfer inside SL and DL flexible microchannel heat sinks have been analyzed in this work. The deformation of the supporting seals was related to the average internal pressure by theory of linear elasticity. Increases in the fixation parameter and the dimensionless pressure drop were found to cause enhancements in the cooling process. These enhancements are significant at lower pressure drop values. Moreover, DL flexible microchannel heat sinks were found to provide additional cooling which were significant at lower values of pressure drop for stiff seals. It is preferred to utilize SL flexible microchannel sinks over DL microchannel heat sinks for large pressure drop applications. However, at lower flow rates the DL flexible microchannel heat sink is preferred to be used over SL flexible microchannel heat sinks especially when stiff sealing material is utilized.
Heat transfer enhancements are investigated inside channels by controlling thermal dispersion effects inside the fluid. Different distributions for the dispersive elements such as nanopaticles or flexible hairy fins extending from the channel plates are considered. Energy equations for different fluid regions are dimensionalized and solved analytically and numerically. The boundary arrangement and the exponential distribution for the dispersive elements are found to produce enhancements in heat transfer compared to the case with a uniform distribution for the dispersive elements. The presence of the dispersive elements in the core region does not affect the heat transfer rate. Moreover, the maximum Nusselt number for analyzed distributions of the dispersive elements are found to be 21% higher than that with uniformly distributed dispersive elements for a uniform flow. On the other hand, the parabolic velocity profile is found to produce a maximum Nusselt number that is 12% higher than that with uniformly distributed dispersive elements for the boundary arrangement. The distribution of the dispersive elements that maximizes the heat transfer is governed by the flow and thermal conditions plus the properties of the dispersive elements. Results in this work point towards preparation of super nanofluids or super dispersive media with enhanced cooling characteristics.
In some embodiments, the super dispersive media comprises at least one nanoparticle which may be metallic or carbon based and includes nanotubes and flexible nanostrings known in the art. In preferred embodiments, the devices of the present invention comprise a coolant and super dispersive media in the microchannels, preferably the super dispersive media comprises at least one metallic nanoparticle, at least one carbon nanoparticle, at least one nanotube, at least one flexible nanostring, or a combination thereof.
In some embodiments, the super dispersive media is non-uniformly distributed in the volumetric space of the microchannel. In some embodiments, the super dispersive media is minimally distributed in the volumetric space of microchannel regions having least transverse convection heat transfer. In other words, the concentration of the super dispersive media is minimal in the volumetric space of microchannel regions having least transverse convection heat transfer. In some embodiments, the super dispersive media is maximum in the volumetric space of microchannel regions having maximum transverse convection heat transfer.
The following Table 9 provides the various symbols and meanings used in this section:
f
nf
p
The heat flux of VLSI microelectronic components can reach up to 1000 kW/m2. As such, many methods are proposed to eliminate excess of heating associated with the operation of these components. One of these methods is to utilize two-layered microchannels. See Vafai & Zhu (1999) Int. J. Heat Mass Transfer 42:2287-2297. Two phase flow are utilized for cooling which was found to be capable of removing maximum heat fluxes generated by electronic packages yet the system may become unstable near certain operating conditions. See Bowers & Mudawar (1994) ASME J. Electronic Packaging 116:290-305. The use of porous blocks inside channels was found to be efficient in eliminating the excess of heat. See Vafai & Huang (1994) ASME J. Heat Transfer 116:604-613; Huang & Vafai (1994) AIAA J. Thermophysics and Heat Transfer 8:563-573; and Hadim, A. (1994) ASME J. Heat Transfer 116:465-472. However, the porous medium creates a substantial increase in the pressure drop inside the cooling device. Recently, Khaled and Vafai demonstrated that expandable systems can provide an efficient method for enhancing the cooling rate. See Khaled & Vafai (2003) ASME J. Heat Transfer 125:916-925. The performance of expandable systems and other cooling systems can be further improved when nanofluids are used as their coolants. See Khaled & Vafai (2003) ASME J. Heat Transfer 125:916-925; Khaled & Vafai (2002) Numerical Heat Transfer, Part A, 42:549-564; Khanafer, K., et al. (2003) Int. J. Heat Mass Transfer 46:3639-3653; and Vafai & Khaled (2004) Int. J. Heat Mass Transfer 47:743-755.
Nanofluids are mixtures of a pure fluid with a small volume of suspensions of ultrafine particles such as copper nanoparticles or nantubes. They were found to possess a large effective thermal conductivity. For example, the effective thermal conductivity of nanofluids could reach 1.5 times that of the pure fluid when the volume fraction of the copper nanoparticles is 0.003. See Eastman, J. A., et al. (2001) Applied Physics Letters 78:718-720. This enhancement is expected to be further enhanced as the flow speed increases resulting in an increase in the mixing effects associated with the Brownian motion of the nanoparticles. This mixing effect is referred in literature as the thermal dispersion effect. See Xuan & Roetzel (2000) Int. J. Heat Mass Transfer 43:3701-3707. Other aspects of dispersion effects can be found in some of the recent works. See Chang, P. Y., et al. (2004) Numerical Heat Transfer, 45:791-809; Hancu, S., et al. (2002) Int. J. Heat Mass Transfer 45:2707-2718; Kuznetsov, A. V., et al. (2002) Numerical Heat Transfer 42:365-383; Gunn, D. J. (2004) Int. J. Heat Mass Transfer 48:2861-2875; and Metzger, T., et al. (2004) Int. J. Heat Mass Transfer 47:3341-3353. Li and Xuan (Li & Xuan (2002) Science in China (Series E) 45:408-416) reported an increase of 60% in the convective heat transfer inside a channel filled with a nanofluid, having 3% volume fraction for copper nanoparticles, compared to its operation with the pure fluid. This significant increase indicates that thermal dispersion is the main mechanism for heat transfer inside convective flows. The challenge is to find new ways to improve the performance of the cooling systems.
In this work, a method for enhancing the heat transfer characteristics through the use of nanofluids with proper thermal dispersion properties is proposed and analyzed. This can be accomplished by having a proper distribution for the ultrafine particles. Physically, the distribution of the ultrafine particles can be controlled using different methods: (i) having nanoparticles with different sizes or physical properties, (ii) applying appropriate magnetic forces along with using magnetized nanoparticles, (iii) applying appropriate centrifugal forces, and (iv) applying appropriate electrostatic forces along with using electrically charged nanoparticles. Different distribution for the nanoparticles can be obtained using any combination of the above methods.
For example, denser nanoparticles such as copper nanoparticles or those with a larger size tend to suspend at lower altitudes in coolants. However, nanoparticles with lower density such as carbon nanoparticles or those having a lower size tend to swim at higher altitudes within denser liquids such as aqueous solutions and liquid metals. As such, non-homogenous thermal dispersion properties can be attained. Centrifugal effects tend to produce concentrated thermal dispersion properties near at least one of the boundaries. On the other hand, non-homogenous thermal dispersion properties inside the coolant can be obtained by attaching to the plates of the cooling device flexible thin fins like hair with appropriate lengths. The Brownian motion of the suspended hairy medium will increase the thermal dispersion properties mainly near the plates of the cooling device and it can be used with a proper suspension system to obtain any required thermal dispersion properties.
Heat transfer enhancements are analyzed inside a channel filled with a coolant having different thermal dispersion properties. Different arrangements for the nanoparticles or the dispersive elements are considered in this work. The nanoparticles or the dispersive elements are considered to be uniformly distributed near the center of the channel for one of the arrangements. In another arrangement, they are uniformly distributed near the channel plates. Exponential or parabolic distributions for the dispersive elements are also analyzed in this work. The energy equations for the corresponding fluid regions are non-dimensionalized. Solutions for the Nusselt number and the temperature are obtained analytically for special cases and numerically for general cases. They are utilized to determine the appropriate distribution for the dispersive elements that will result in the maximum heat transfer with the same total number of nanopartilces or the dispersive elements.
Consider a flow inside a two dimensional channel with a height 2h and a length B. The x-axis is aligned along the centerline of the channel while the y-axis is in the traverse direction as shown in
wherein T, ρ, cp and k are the temperature, effective fluid density, fluid specific heat and thermal conductivity, respectively. The velocity field u in the channel is taken to be fully developed. The volume of the dispersive elements is very small such that the velocity profile is parabolic.
wherein um is the mean flow speed.
For nanofluids or in the thermally dispersed region, the parameter ρcp will be (ρcp)nf and it is equal to:
(ρcp)nf=(1−φ)(ρcp)f+φ(ρcp)p Eq. 168
wherein the subscript nf, f, and p denote the nanofluid or the dispersive region, pure fluid, and the particles, respectively. The parameter φ is the nanoparticles or the dispersive elements volume fraction which represents the ratio of the nanoparticles or the dispersive elements volume to the total volume. A nanofluid composed of pure water and copper nanoparticles suspensions with 2% volume fraction has a value of (ρcp)nf equal to 99% that for the pure water which is almost the same as the thermal capacity of the pure fluid.
The ultrafine suspensions such as nanoparticles, nanotubes or any dispersive elements in the fluid plays an important role in heat transfer inside the channel as their Brownian motions tend to increase fluid mixing. This enhances the heat transfer. The correlations presented in the work of Li and Xuan (Li & Xuan (2002) Science in China (Series E) 45:408-416) for Nusselt numbers in laminar or turbulent flows show that the heat transfer is enhanced in the presence of nanoparticles and it increases as the nanoparticles volume fraction, the diameter of the nanoparticles or the flow speed increase. Xuan and Roetzel (Xuan & Roetzel (2000) Int. J. Heat Mass Transfer 43:3701-3707) suggest (consistent with the dispersion model given in Amiri and Vafai (Amiri & Vafai (1994) Int. J. Heat Mass Transfer 37:939-954)) the following linearalized model for the effective thermal conductivity of the nanofluid:
k=k
o
=C*(ρcp)nfφhu Eq. 169
wherein C* is a constant depending on the diameter of the nanoparticle and its surface geometry.
Physically, Equation 169 is a first approximation for the thermal conductivity of the nanofluid that linearly relates it to thermal capacitance of the flowing nanoparticles or flowing dispersive elements. The constant ko represents the effective thermal conductivity of the nanofluid or the dispersive region under stagnant conditions, at u=0. This constant can be predicted for nanofluids from the formula suggested by Wasp (Wasp, F. J. (1977) Solid-Liquid Slurry Pipeline Transportation, Trans. Tech. Berlin) which has the following form:
wherein kp and kf are the thermal conductivity of the nanoparticles and the pure fluid, respectively.
According to Equation 170, a two percent volume fraction of ultrafine copper particles produces 8 percent increase in ko when compared to the thermal conductivity of the pure fluid. On the other hand, the experimental results illustrated in the work of Li and Xuan (Li & Xuan (2002) Science in China (Series E) 45:408-416) shows that the presence of suspended copper nanoparticles with 2 percent volume fraction produced about 60% increase in the convective heat transfer coefficient compared to pure fluid. See Table 10 as follows:
This indicates that thermal dispersion is the main mechanism for enhancing heat transfer inside channels filled with nanofluids under convective conditions. Non-dimensionalizing Equation 166 with the following dimensionless variables:
leads to the following dimensionless energy equation:
wherein q, T1 and Pe are the heat flux at the channel's plates, the inlet temperature and the Peclet number (Pe=(ρcpumh)/kf), respectively. It is assumed that the heat flux is constant and equal at both plates.
For simplicity, the term k/kf will be rearranged in the following form:
wherein Pef=(ρcp)fumh/kf.
A portion of the fluid's volume are considered in part of this work to be subjected to thermal dispersion effects due the suspensions of nanoparticles or any dispersive elements while the other portion contains only the pure fluid. The most obvious way to obtain specific distributions for thermal dispersive elements is to have conductive hairy fins extending from the channel plates or from a carefully designed fixed or flexible structure placed in the channel. The volume of this structure is small enough such that the parabolic assumption for the velocity profile is still valid. Also, non-homogenous thermal dispersion properties can be achieved by having nanoparicles with different densities or different sizes. Heavier nanoparticles or dispersive elements tend to swim closer to the lower plate due to gravitational forces while lighter nanoparticles or dispersive elements tend to swim closer to the upper force due to buoyancy forces. The dispersive elements such as nanoparticles can be further concentrated near the channel's plates by having these particles magnetized along with applying appropriate magnetic fields. As such, the difference in the thermal dispersive properties of the nanofluid can be achieved. Appropriate thermal dispersive properties can be obtained by utilizing the different methods discussed in the introduction section.
The dimensionless energy equation for the part involving thermal dispersion is:
whilein the energy equation for the volume containing the pure fluid is:
Different distributions for the nanoparticles of the dispersive elements will be analyzed in this work. In one of these distributions, the region that is active with thermal dispersion effects is considered to be a rectangular region of height 2l around the channel's centerline as shown in
The boundary conditions for the central arrangement are:
while the boundary conditions for the second arrangement (boundary arrangement) are:
wherein Λ=l/h. Other distributions for the dispersive elements will be considered later such as the parabolic distribution and the exponential distribution.
For thermal fully developed conditions, axial gradient of the temperature reaches a constant value equal to dT/dx. That is, the heat flux is equal to:
wherein φcf is the ratio of the volume comprising thermal dispersion effects to the total channel volume. (um)f is the average velocity in the fluid phase while (um)nf is the average velocity in the nanofluid or the region containing the thermal dispersive elements.
As such, Equation 174 and Equation 175 reduce to:
Since (ρcp)nf does not vary significantly when the volume fraction of the ultrafine particles or the dispersive elements is less than 4% as used in the literature, A and G are almost equal to unity
Consider a uniform flow inside the channel such that U=1. Equation 179 and Equation 180 reduce to:
The solution to Equation 181a and Equation 181b for the central arrangement of the dispersive elements is Equation 182a:
wherein θW is the plate temperature at a given section X. The parameter θm is the mean bulk temperature. It is defined as:
As such, the fully developed value for the Nusselt number is:
wherein hc is the convective heat transfer coefficient at the channel's plate.
For the second type of arrangements for the thermal dispersion region. The solution for Equation 181a and Equation 181b is Equation 185a:
The corresponding fully developed value for Nusselt number for this case is:
The total number of dispersive elements is considered to be fixed for each distribution. As such, the volume fraction of the dispersive element for the central or the boundary arrangements is related to their thickness according to the following relation:
wherein φo is the volume fraction of the dispersive elements when they are uniformly filling the whole channel volume. Utilizing Equation 187, the parameter E utilized in Equation 179 and Equation 180 can be expressed according to the following:
wherein Eo is named as the thermal dispersion parameter.
Practically, it is difficult to have the dispersive elements concentrated in a region while the other region is a pure fluid. As such, two other distributions for the dispersive elements are considered in this work. They are the exponential and the parabolic distributions as illustrated in the following:
Note that the average volume fraction for each distribution is φo irrespective to values of De and Dp. One of the objectives of our work is to obtain the values of Dc and De and Λ that produces maximum heat transfer inside the channel.
The excess in Nusselt number κ is defined as the ratio of the maximum Nusselt number that can be obtained by having a certain volume fraction distribution (Nund) to the Nusselt number corresponding to a uniform distribution of the dispersive elements (Nuud). It is expressed as follows:
It can be shown that Equation 191 exhibits a local maximum or minimum value at specific thermal dispersion parameter *o
critical for the boundary arrangement. This is related to the dimensionless thickness of the dispersive region through the following relation:
Equation 174 and Equation 175 were descritized using three points centeral differencing in the Y direction while backward differencing was utilized for the temperature gradient in the X-direction. The resulting tri-diagonal system of algebraic equations at X=ΔX was then solved using the well established Thomas algorithm. See Blottner, F. G. (1970) AIAA J. 8:193-205. The same procedure was repeated for the consecutive X-values until X reached the value of B/h. Equation 179 and Equation 180 were also descritized using three points central differencing and solved using Thomas algorithm.
At high temperatures, the main gas expands while the secondary gas volume shrinks. Since the main gas possesses lower thermal conductivity, the effective resistance of the device increases causing an enhancement in the insulating properties at large operating temperatures. Additional enhancement can be obtained by utilizing a series of the gas compartments as shown in
Enhancements in heat transfer can be investigated inside channels filled with a fluid having different thermal dispersive properties. Different distributions for dispersive elements such as nanoparticles or flexible hairy tubes extending from the channel plates are considered. The dispersive elements are considered to be uniformly distributed in the central region, near the boundaries, having an exponential distribution and having a parabolic distribution.
The boundary arrangement and the exponential distribution of the dispersive elements have been shown to produce substantial enhancements in heat transfer compared to the case when the dispersive elements are uniformly distributed. The presence of the dispersive elements in a core region does not produce any significant change in the heat transfer. The maximum excess in Nusselt number has been found to be 1.21 utilizing the boundary arrangement for the volume fraction with uniform flow while the parabolic velocity profile produced a maximum excess in Nusselt number equal to 1.12. The volume fraction distribution that maximizes the heat transfer is generally governed by the flow and thermal conditions as well as the properties of dispersive elements. This demonstrates that super nanofluids or super dispersive media can be prepared by controlling the thermal dispersion properties inside the fluid.
In general, an apparatus can be configured which includes a first gas compartment comprising a first substrate and a second substrate. In an embodiment, such an apparatus can also include a second gas compartment comprising the second substrate and a third substrate. In such an embodiment of such an apparatus, the first substrate can include a face in contact with at least one hot medium and having another face in contact with a main gas and having fixed supports. Additionally, in such an apparatus, the second substrate can include a face in contact with the secondary gas and another face in contact with a main gas. Additionally, the third substrate can include a face in contact with the secondary gas and having another face in contact with at least one cold medium.
Such an apparatus can also include a first at least one flexible seal attached to one end of the second substrate and to the opposing end of the first substrate to form at least one closed enclosure comprising the main gas so that the second substrate moves when the main gas undergoes volumetric thermal expansion. Such an apparatus can also include a second at least one flexible seal attached to other end of the second substrate and to the opposing end of the third substrate to form at least one closed enclosure comprising the secondary gas so that this enclosure contracts when the main gas undergoes volumetric thermal expansion. Additionally, the second gas compartment can be vented such that the secondary gas pressure remains constant and allows a maximum volumetric thermal expansion of the main gas.
In another embodiment of such an apparatus, the secondary gas compartment can further include a plurality of elastic balloons filled with the main gas. In yet another embodiment of such an apparatus, the third substrate is not fixed and the second at least one flexible seal can be attached to one end of the first substrate and to the opposing end of the third substrate. In still another embodiment, the main gas can possess a thermal conductivity that is smaller than that of the secondary gas so that effective thermal insulation properties of said apparatus are improved when the main gas undergoes a volumetric thermal expansion. In other embodiments, the flexible seal can comprise a closed-cell form seal that can sustain high temperatures. An additional enhancement in insulating property can be obtained utilizing a series of the first and second gas compartments.
To the extent necessary to understand or complete the disclosure herein, all publications, patents, and patent applications mentioned herein are expressly incorporated by reference therein to the same extent as though each were individually so incorporated.
Having thus described exemplary embodiments of the present invention, it should be noted by those skilled in the art that the within disclosures are exemplary only and that various other alternatives, adaptations, and modifications may be made within the scope of the present invention. Accordingly, the present invention is not limited to the specific embodiments as illustrated herein, but is only limited by the following claims.
This patent is a Continuation-in-Part of U.S. patent application Ser. No. 12/793,826, entitled “Methods and Devices Comprising Flexible Seals, Flexible Microchannels, or Both for Modulating or Controlling Flow and Heat” filed on Jun. 4, 2010, which in turn is a continuation of U.S. patent application Ser. No. 11/184,932, entitled “Methods and Devices Comprising Flexible Seals, Flexible Microchannels, or Both For Modulating or Controlling Flow and Heat,” which was filed on Jul. 20, 2005, which issued as U.S. Pat. No. 7,770,809 on Aug. 10, 2010, the disclosure of which is incorporated herein by reference in its entirety. U.S. patent application Ser. No, 11/184,932 in turn is a continuation-in-part of U.S. patent application Ser. No. 10/840,303, filed 7 May 2004, which issued as U.S. Pat. No. 7,654,468 on Feb. 2, 2010, which claims the benefit of U.S. Provisional Patent Application No. 60/470,850 filed 16 May 2003, which are herein incorporated by reference in their entirety.
| Number | Date | Country | |
|---|---|---|---|
| 60470850 | May 2003 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 11184932 | Jul 2005 | US |
| Child | 12793826 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 12793826 | Jun 2010 | US |
| Child | 13030319 | US | |
| Parent | 10840303 | May 2004 | US |
| Child | 11184932 | US |
