| Heat Transfer |
| Leon R. Glicksman |
| © 1991, 1997, 2004, 2005, 2010 |
|
|
| 1)Introduction |
|
|
| Heat transfer deals with the rate of heat transfer between |
| different bodies. While thermodynamics deals with the magnitude of |
| heat exchanged in a process, heat transfer is necessary to determine |
| the time required for a process or alternatively the size of a surface |
| necessary to achieve a certain total rate of heat transfer. |
|
|
| Heat transfer analysis permits a calculation of the heat loss |
| from a building surface to the surroundings for a given building size, |
| window area and wall design, e.g. the level of insulation in the wall |
| cavity. The comfort conditions for occupants in a room is determined |
| by a balance of heat transfer from the person to the air surrounding |
| him or her as well as the heat transfer to the walls of the interior. |
| The size and cost of a heat exchanger is also determined by |
| considering the heat transfer between the fluid streams in the |
| exchanger. |
|
|
| In other fields, heat transfer plays a key role as well. The |
| design of integrated microprocessors which contain very closely spaced |
| elements, each with a finite amount of heat generation, is limited by |
| the requirement for adequate cooling so that the operating temperature |
| of the electronic components is not exceeded. Reentry of the space |
| shuttle in the earth's atmosphere must be carefully programmed so that |
| temperature extremes due to air friction are confined to the |
| insulating tiles on the shuttle's surface. |
|
|
| Modes of Heat Transfer |
|
|
| Following thermodynamics, heat transfer is that energy transfer |
|
|
| which takes place between two bodies by virtue of a temperature |
| difference between the bodies. From the second law considerations it |
| can be demonstrated that there is always a net positive energy |
| transfer from the body at a high temperature to a second body at a |
| lower temperature. Following the definition of heat, there are only |
| two physical mechanisms for heat transfer: (1) electromagnetic waves |
| produced by virtue of the temperature of a body, referred to as |
| thermal radiation heat transfer and (2) atomic or molecular motion in |
| a medium between the bodies exchanging energy, referred to as |
| conduction heat transfer. |
|
|
| Sometimes conduction heat transfer takes place during the |
|
|
| change of phase and is referred to as boiling or condensation |
| heat transfer. Conduction heat transfer can also take place in |
| the presence of fluid motion, which is called convection heat |
| transfer. |
|
|
| The rate of heat transfer between two bodies is |
|
|
| proportional to the temperature difference between the bodies |
| and in some cases the temperature level of the bodies as well. |
| In many instances the heat transfer process is analogous to the |
| rate of transfer which appears in other fields. The analogy |
| between heat transfer and DC electrical current flow will be |
| used to illustrate some of the simpler heat transfer processes. |
| Similarly, it can be shown that the rate of transfer of mass in |
| an evaporation process follows a process very similar to that |
| for heat transfer. |
|
|
| 2)Conduction Heat Transfer |
|
|
| In a homogenous body which experiences a temperature |
|
|
| gradient the rate of heat transfer due to microscopic motions is |
| conduction heat transfer. In a gas the gas molecules in the |
| higher temperature portion of the gas will have a higher kinetic |
| energy. As the molecules of the gas randomly move through the |
| gas volume there is a net energy transfer |
| from the high temperature portion to the |
| low temperature zones. In a solid, the |
| energy transfer from high to low |
| temperature may be due to the migration of |
| electrons or the vibration of the |
| molecular bonds. |
|
|
| Viewed as a macroscopic phenomena, |
| the rate of heat transfer by conduction |
| represented by the symbol q or Q is found |
| to be directly proportional to the product |
| of the local temperature gradient and the |
| cross-sectional area available for heat transfer, |
|
|
| Fig 2.1 One dimensional |
| conduction |
|
|
| (2.1) |
|
|
| ~qAgradT |
|
|
| In the case of one-dimensional heat transfer normal to a plane |
| slab, figure 1, the conduction heat transfer can be given by |
| Fourier's Equation, |
|
|
| (2.2) |
|
|
| The constant k is known as the thermal conductivity. q has |
|
|
| the dimensions of BTU/hr or Watts and k has the dimensions of |
| BTU/hrft F or W/m K. |
|
|
| The thermal conductivity defined by equation 2.2 is a |
| thermophysical property of the material. If the composition and |
| thermodynamic state is known then the thermal conductivity can |
| be found. |
|
|
| Table 2.1 lists the thermal conductivity of common solids, |
| liquids and gases at normal temperatures. Note that these |
| values span many orders of magnitude with electrically |
| conductors having the highest thermal conductivity and high |
| molecular weight gases generally having the lowest thermal |
| conductivity. |
|
|
| Consider a slab with a steady |
|
|
| conduction heat transfer across it in the x |
| direction, fig. 2, with the temperature |
| equal to T1 and T2 at the surfaces |
| corresponding to x equal to 0 and L, |
| respectively. Then q is a constant and |
| equation 2 can be integrated to give, |
|
|
| Fig. 2,2 Conduction Through |
| a Plane Wall |
|
|
| (2.3) |
|
|
| For this case the temperature varies linearly across the width |
| of the slab. One can consider an analogy between the solution |
| for steady conduction and for steady D.C. electric current flow, |
| Ohm's Law, |
|
|
| dTqkAdx12TTqkAL |
|
|
| (2.4) |
|
|
| Table 2.1 |
| Thermal Conductivity of |
| Common Materials |
|
|
| k(BTU/hr ft F) |
|
|
| (W/mK) |
|
|
| Solids |
|
|
| Copper |
| Aluminum |
| Steel |
| Brick,common |
| Concrete |
| Glass |
| Glass fiber insulation |
| Ice |
| Plastic |
| Wood |
|
|
| 219 |
| 119 |
| 25 |
| 0.2 - 0.1 |
| 0.5 - 0.8 |
| 0.5 |
| .03 1.3 |
| 0.1 |
| 0.1 - 0.2 |
|
|
| 378 |
| 206 |
| 43 |
| 0.17 - 0.34 |
| 0.87 - 1.38 |
| 0.87 |
| 0.05 |
| 2.2 |
| 0.17 |
| 0.17 - 0.34 |
|
|
| Liquids |
|
|
| Ammonia |
| Refrigerant-12 |
| Light Oil |
| Water |
| Mercury |
|
|
| Gases |
|
|
| 0.3 |
| 0.04 |
| 0.08 |
| 0.34 |
| 5 |
|
|
| 0.5 |
| 0.07 |
| 0.14 |
| 0.59 |
| 8.7 |
|
|
| Air,dry |
| Carbon Dioxide |
| Helium |
| Hydrogen |
| Water Vapor (Steam) |
|
|
| 0.015 |
| 0.009 |
| 0.09 |
| 0.11 |
| 0.015(at 212 F)0.026 (at 100 |
|
|
| 0.026 |
| 0.016 |
| 0.16 |
| 0.19 |
|
|
| C) |
|
|
| Refrigerant-11 |
|
|
| 0.005 |
|
|
| 0.009 |
|
|
| 1.0 (BTU/hr ft F)= 1.73 (W/m C) |
| ________________________________________________________________ |
|
|
| 21VVIR |
|
|
| The rate of heat transfer q is analogous to the current flow I, |
| the potential difference V is analogous to T and the balance of |
| equation 2.3 is analogous to the resistance. The term thermal |
| resistance is used; for eqn. 2.3 the thermal resistance is L/kA. |
|
|
| Fig. 2.3 Steady State Heat Transfer Through a Composite Wall |
|
|
| Consider the case of steady heat transfer through a composite |
| wall as shown in figure 2.3. Each element of the wall has the |
| same heat transfer rate q through it and for each an equation |
| similar to equation 2.3 can be written. At steady state with no |
| change in internal energy with time, no work, and no mass flows |
| through each of the elements, the rate of heat transfer into and |
| out of each wall element must be the same. For the wall board, |
|
|
| This can be rewritten as, |
|
|
| For the insulation |
|
|
| (2.5) |
|
|
| (2.6) |
|
|
| 12WBWBkAqTTL12WBWBLTTqkA |
|
|
| Similar equations can be written for the plywood, T3 - T4 and |
| and the siding, T4 - T5. When these equations are summed up the |
| intermediate temperatures T2, T3 and T4 cancel and the resulting |
| equation becomes |
|
|
| (2.7) |
|
|
| or |
|
|
| (2.8) |
|
|
| (2.9) |
|
|
| The electrical analogy for this case is resisters in series as |
| shown in figure 2.4. |
|
|
| Fig. 2.4 Electric Analogy to Steady State Heat Transfer |
|
|
| Then the overall |
| solution can be |
| easily written as, |
|
|
| which is identical to equation 2.9. |
|
|
| (2.10) |
|
|
| For a planar wall, the cross sectional area of each element is the |
|
|
| 23IILTTqkA15WBSIPWBIPSLLLLTTqkAkAkAkA15TTqLkA(2.10)TqR |
|
|
| same so that equation 9 can be rewritten as |
|
|
| R-value of that material. Note the R-value is independent of the |
| surface area A while the thermal resistance RT includes the surface |
| area. For the plywood, a typical R-value for a one inch thickness is |
|
|
| Each of the terms in the denominator is sometimes referred to as the |
|
|
| Note in US building practice units and dimensions are still in the |
| imperial system. |
| residential wall constructed with 2 by 4 studs, |
|
|
| If the insulation is three and one half inches thick in a typical |
|
|
| For the composite wall in figure 2.3, the R-value of the insulation |
| dominates all of the terms in equation 2.11 |
|
|
| (2.11) |
|
|
| (2.12) |
|
|
| (2.13) |
|
|
| 155(2.11)WBSIPWBIPSTTAqLLLLkkkk21/120.8(2.12)0.1PhrftFLkBTU23.5/12112.13)0.028IhrftFLkBTU |
|
|
| 8 |
|
|
| Convection Heat Transfer, Introduction |
|
|
| temperature on the inside wall |
| surface, T1, and the outside |
| siding surface T5 are not |
| generally known. Rather the |
| interior room air temperature, |
| Ti, and the exterior air |
| temperature, Te, are the known |
| quantities. Consider a |
| wintertime condition, when the |
| building is at a higher |
| temperature then the exterior |
| air. The temperature through the |
| built-up wall continuously de- |
| creases from the inside wall at |
| T1 to the outside surface at T5. |
| This is shown in figure 5. The |
| outside surface temperature T5 is higher than the exterior air |
| temperature Te. In the air layer close to the building surface |
| the air is in motion parallel to the surface. There is heat |
| transfer by conduction from the building surface through this |
| air layer. Because there is also energy transfer by the motion |
| of the fluid the temperature through the air layer does not vary |
| linearly. Rather, there is a large temperature gradient near |
| the surface which decreases further from the surface until the |
| temperature reaches the constant air temperature Te. The layer |
| over which the temperature change occurs is thin, typically one |
| quarter of an inch or less. |
|
|
| Fig. 2.5 Temperature Distribution with Convection |
| at the Surfaces |
|
|
| energy transfer by fluid motion is called convection heat |
| transfer. The rate of heat transfer is proportional to the |
| surface area and the temperature difference between the surface |
| and the uniform air temperature outside of the thin surface or |
| boundary layer, |
|
|
| The expression is changed to an equality and in the process a |
| new quantity, h, the heat transfer coefficient is defined, |
|
|
| In a typical wall construction, as shown in figure 2.3, the |
|
|
| The process of conduction heat transfer through the air combined with |
|
|
| (2.14) |
|
|
| 5~()(2.14)eqATT |
|
|
| 9 |
|
|
| where h has the units of BTU/hr ft2 F or W/m2 . Equation 2.15 |
| is of no use until some way to calculate h is established. |
|
|
| Generally, the heat transfer coefficient, h, is a function of |
| the fluid properties, the fluid velocity, the surface geometry |
| and sometimes the temperature level. A more detailed discussion |
| of convection will be given later. For now it is sufficient to |
| observe that h increases as the air velocity increases and it |
| increases with fluids of higher thermal conductivity. |
|
|
| There are two general forms of convection. When the air |
|
|
| motion is set up by buoyancy effects due to the applied |
| temperature difference between the surface and the fluid, e.g. |
| the air flow over a hot `radiator', the flow is natural or free |
| convection. When the flow is due to an external source, e.g. |
| the wind, a fan or by the motion of the surface, the flow is |
| forced convection. |
|
|
| Rohsenow has presented a table which gives good estimates |
| of the order of magnitude of h for convection heat transfer as |
| well as boiling and condensation. It is reproduced in table |
| 2.2. |
|
|
| Table 2.2 |
| Convection Heat Transfer Coefficients |
|
|
| BTU/hr ft2 |
|
|
| Gases, Natural Convection 0.5-50 |
| Gases, Forced Convection 2-50 |
| Liquids, forced Convection 30-1000 |
| Boiling Liquids 200-50,000 |
| Phase Change 500-5,000 |
|
|
| Now returning to the concept of thermal resistance, from |
| equation 15 the equivalent thermal resistance, RT for convective |
| heat transfer is 1/hA. |
|
|
| (2.15) |
|
|
| Generally, the heat transfer coefficient, h, is a function of the. |
|
|
| (2.15)surfacefluidfarfromsurfaceqhATT |
|
|
| Consider a single glazed window. What is the increase in the energy |
|
|
| 10 |
|
|
| Example |
| For a single glazed window |
| what is the increase in |
| thermal efficiency if the |
| glass is made of plastic with |
| k = 0.1 BTU/hr ft F instead |
| of glass with a k = 0.5 |
| BTU/hr ft F? Assume that |
| the radiation heat transfer |
| remains the same. |
|
|
| Fig. 2.6 Heat transfer through window |
|
|
| Fig. 2.6 Heat transfer through window |
|
|
| SOLUTION |
|
|
| In this case convection heat |
| transfer from the inside air at |
| Ti to the glass surface acts in |
| series with conduction through |
| Figure2.7Electrical Analogy |
| the glass and convection to the |
| outside air. The equivalent electrical circuit is shown on |
| figure 7 and the steady state heat transfer, neglecting |
| radiation, becomes |
|
|
| Using table 2 the magnitudes of hi and he are |
|
|
| (2.16) |
|
|
| 1//1/1//1/insideairexteriorairinsideairexteriorairigeigeTTTTAqhALkAhAhLkh |
|
|
| 11 |
|
|
| (2.17) |
|
|
| For the glass, assuming it is 1/8 inch thick, |
|
|
| (2.18) |
|
|
| Changing to plastic decreases kg to 0.1 and increases L/kg to |
| l/10 but it will only change the overall value of q, given by |
| equation 16 by less than 10 percent. |
|
|
| The overall heat transfer for composite systems such as |
| figure 2.3 or figure 2.6, represented by equations 2.9 and 2.16, |
| respectively is sometimes rewritten in terms of an overall heat |
| transfer coefficient U defined as |
|
|
| (2.19) |
| Although U has the same units as h, U can involve a combination |
| of conduction and convection heat transfer and is not physically |
| meaningful although it may be helpful for estimate purposes. A |
| number of handbooks like the ASHRAE Handbook of Fundamentals |
| list values of U for typical built up wall and roof |
| construction. These values of U include convection heat |
| transfer on the inside and outside for an assumed wind velocity |
| and interior air circulation conditions. |
|
|
| Two-dimensional Heat Transfer |
|
|
| 12121111110.333ieBTUhhrftFBTUhhrftF121/81/1210.550gLBTUkhrftF()ieqUATT |
|
|
| 12 |
|
|
| Fig. 2.8 |
|
|
| Most walls are not uniform across their entire surface |
| area. Wood framing using 2 by 4's has studs spaced at regular |
| intervals in the wall cavity, fig 2.8. Clearly the heat |
| transfer through the studs is higher than the heat transfer |
| through |
| containing |
| insulation. |
|
|
| cross-sectional |
|
|
| equivalent |
|
|
| area |
|
|
| an |
|
|
| The heat transfer through the wall cavity is due to two |
| parallel conduction paths, one through the studs and the other |
| through the insulation. If the lateral resistance is very |
| large, i.e., the lateral conductivity of the wallboard and the |
| plywood approaches zero, then the overall heat transfer can be |
| modeled as two separate parallel heat flow paths from Ti to Te, |
| shown in fig. 2.9. |
|
|
| The heat transfer through the studs is |
|
|
| and through the insulation, |
|
|
| (2.20) |
|
|
| 1/(/)(/)(/)1/iestudstudiWBSPeTTAqhLkLkLkh |
|
|
| 13 |
|
|
| (2.21) |
|
|
| The total heat transfer rate is the sum of equations 2.20 and |
| 2.21. The true value of the two-dimensional heat transfer lies |
| between these two limiting cases of very small lateral |
| conductivity and very large lateral conductivity. |
| The electrical analogy of this, in one limit, is shown in figure |
| 2.9 with the insulation and the studs in parallel. |
|
|
| It is important to note that this approximation for the two- |
| dimensional case represents one limiting case, the temperature |
| on the inside of the wall board where it contacts the stud will |
| be different from the temperature of the wall board in contact |
|
|
| Fig 2.9 Limiting Case of Small Lateral Conductivity in Wall Board and Plywood |
| with the insulation. Similarly the temperature of the plywood |
| will differ in the lateral direction between the stud and the |
| insulation. These temperature differences across the wall board |
| will cause heat to flow laterally into the stud, fig. 2.10. |
| Similarly heat will flow |
| laterally out from the stud |
| through the plywood. These |
| lateral effects will enhance |
| the total heat transfer |
| through the stud. This will |
| increase the overall heat |
| transfer over the values |
| calculated from the sum of |
| equations 2.20 and 2.21. |
|
|
| Fig. 2.10 |
| lateral heat Flow in the vicinity of the stud |
|
|
| 1/(/)(/)(/)1/ieinsulationinsulationiWBinsulationPeTTAqhLkLkLkh |
|
|
| The exact analysis of these two-dimensional effects is |
| difficult. We can look at the other extreme case. If the wall |
| board and plywood were replaced by metal sheets, something |
| sometimes seen in metal walled buildings, then the resistance to |
| lateral conduction through them would be reduced. In the |
| limiting case w e could assume that the wall board and plywood |
| would have a very high conductivity and the temperature would be |
| uniform in the lateral direction across these elements. The |
| electrical analogy for this case is shown on fig. 2.11. The |
| solution shown in figure 10 is only valid when the lateral |
| resistance is small, i.e., the y direction conductivity of the |
| wall board and the plywood approaches infinity. |
|
|
| In this case, the heat transfer through the studs and insulation |
| acts in parallel between T2 and T3, the sheet rock and plywood |
| temperatures, respectively. The heat transfer through the two |
| elements is summed, |
|
|
| and defining an equivalent resistance, |
|
|
| (2.22) |
|
|
| (2.23) |
|
|
| Fig 2.11 Limiting Case for |
| Small Lateral Resistance to Heat Transfer in Sheathing |
|
|
| 2323(/)(/)insulationstudTTTTqLkALkA2323(/)(/)insulationstudequivalentTTTTqLkALkAR |
|
|
| The resistance, Requivalent,is then added in series to the remaining |
| resistance shown in figure 2.11. This extreme case will result |
| in a much larger overall calculated heat transfer than the other |
| extreme shown in figure 2.9. The true value lies between the two |
| and must be evaluated by analysis or practical judgment. |
| Transient Heat Transfer |
|
|
| Conditions for Uniform Temperature |
|
|
| the |
|
|
| provide |
|
|
| changes, |
|
|
| significant |
|
|
| When the temperature of the |
| interior |
| building |
| structure may be sufficiently massive |
| to |
| thermal |
| storage. This technique is used in |
| naturally ventilated buildings, using |
| night cooling in the summer to reduce |
| the structural temperature. During the |
| hot daytime the cool structure will |
| interior |
| help |
| temperature |
| comfortable |
| range. However, if the structure is |
| very |
| a |
| substantial time interval before the |
| temperature change is felt throughout the thickness of the |
| structure. |
|
|
| maintain |
| within |
|
|
| the |
| a |
|
|
| require |
|
|
| Fig. 2.12 |
|
|
| thick, |
|
|
| may |
|
|
| to |
|
|
| it |
|
|
| Consider the case of a plane homogenous slab of thickness |
| 2L, fig. 2.12. The y and z dimensions are large compared to L |
| so that the heat transfer can be assumed to be one dimensional, |
| in the x direction only. Initially the slab is at a uniform |
| temperature T0 equal to the exterior air temperature. At time |
| zero the air temperature suddenly increases to a new temperature |
| level Te. Both sides of the slab are in contact with the air and |
| both sides have convective heat transfer coefficient h. At |
| short times, the surface temperature of the slab will increase |
| due to the heat transfer from the air. Some of the energy |
| transferred to the slab will be used to raise the internal |
| energy of the material near the |
| surface. The balance of the heat |
| transfer will be transferred to the |
| next layer inside the slab where the |
| same process occurs. The temperature |
| distribution within the slab at some |
|
|
| Fig. 2.13 |
|
|
| intermediate time is shown in figure 2.13. When the conduction |
| of the slab is large and the thickness is small, it is expected |
| that the temperature differences from the surface to the center |
| of the slab will be small at all times. A better criterion is to |
| compare the resistance to conduction, L/kslab,to the convective |
| resistance at the surface, 1/h. When 1/h is much larger than |
| L/kslab it is expected that the temperature of the slab will be |
| uniform throughout it thickness. A handy criterion is hL/kslab < |
| 1/6 for uniform temperature. The term hL/k which is |
| dimensionless is known as the Biot number. |
|
|
| When hl/k < 1/6, the slab temperature T only varies with time. |
| Assuming the slab initially is at temperature T0 and at time t it |
| contacts air at a different temperature T the energy equation |
| becomes, |
|
|
| (2.24) |
|
|
| where A is the entire surface area of the slab which is in |
| contact with air at a constant temperature T . When there isn't |
| any change of phase, the energy change is given as Mc T, where M |
| is the total mass of the slab. Equation 2.24 becomes, |
|
|
| Since T is a constant this can be rearranged to read, |
|
|
| (2.25) |
|
|
| (2.26) |
|
|
| with the initial condition for the slab that |
| at t = 0, T = T0 |
| The solution of equation 31 is |
|
|
| (2.27) |
|
|
| where is the thermal time constant of the slab with convective |
| heat transfer coefficient h and |
|
|
| ()dEqhATTdT()dTMchATTdt()()dTThAdtTTMc00()exp()exphAttTTTTTTMc |
|
|
| (2.28) |
|
|
| Example |
| A two inch thick steel structural beam has natural convection |
| heat transfer over one surface; the opposite side is insulated. |
| Find its time constant. |
|
|
| Solution |
|
|
| First it must be determined if the steel can be assumed |
| uniform in temperature across its width. Since only one side |
| has convective heat transfer, referring to figure 2.13, L in |
| this case should be the full width of the steel beam. The Biot |
| number is |
|
|
| (2.29) |
|
|
| where h is estimated for natural convection from table 2.2. The |
| assumption of uniform temperature is clearly justified. Note |
| that if we were considering a two inch thick concrete section |
| with a conductivity of about 1 BTU/hr ft F the Biot number |
| would be 1/3 and the assumption of uniform temperature through |
| the concrete would be questionable. |
|
|
| The thermal time constant for the steel beam is, |
|
|
| The time constant is a function of the steel properties and the |
| heat transfer coefficient at its surface. If air was blowing |
| over the surface at a high velocity the time constant would be |
| substantially reduced. |
|
|
| MchA2(2/12)112060hLk400(0.1)(2/12)3.52McLAcLchrshAhAh |
|
|
| 3. Convection Heat Transfer |
|
|
| Figure 3.1 Flow over a heated plate |
| Figure 3.1 Flow over a heated plate |
|
|
| Convective heat transfer is conduction |
|
|
| though a gas or liquid augmented by fluid motion. |
| In these notes an introduction to the physics |
| governing convection will be given along with |
| some results for several different conditions. |
| Consider the case of air at a uniform |
|
|
| temperature Ta blown by a fan along a flat |
| surface which is heated to a uniform temperature |
| Ts, Figure 3.1. Say, this is cool air in an air- |
| conditioned room flowing over the surface of a |
| window heated by solar radiation. The element |
| of air closest to the heated surface has a |
| temperature increase as it starts to move up the |
| plate. Elements further away, at a larger y |
| coordinate still are at Ta. As the element |
| becomes hotter it moves up and is replaced with |
| another element at Ta and the process is |
| repeated. Viewed from the point of view of the |
| room as a whole, room air at Ta approaches the |
|
|
| plate and a given flow of air leaves the plate at an elevated temperature somewhere between Ta |
| and Ts. Thus there is a net energy transfer from the surface of the window to the room air. |
|
|
| The convective heat transfer coefficient is defined as, |
|
|
| (3.1) |
|
|
| To get an estimate of how the rate of heat transfer is influenced |
| by the parameters of the problems we have to look more closely |
| at the layer of air moving along the plate surface. To make the |
| explanation clear we will assume all of the elements of air are |
| moving along the plate surface at uniform velocity and in straight |
| lines. This is an instance of laminar flow. At any location (x,y) |
| fixed relative to the stationary plate the temperature remains |
| constant with time and there is a steady conduction heat transfer |
| from the surface to the air in the y direction, normal to the plate |
| surface. As the air continues up the plate the elements close to |
| the plate increase their temperature. At the same time the |
| elements further away start to rise in temperature due to |
| conduction from the hotter elements at the plate surface..The |
| further along the air moves in the x direction the more elements |
| further from the plate surface feel the conduction heat transfer |
| and rise in temperature as shown in figure 3.2.The maximum y |
|
|
| 1.2 1.2 |
|
|
| saqhATT |
|
|
| distance at which the thermal effects are felt, at any location x, will be specified as , the thermal |
| boundary layer thickness. |
|
|
| At any position x the variation of air temperature normal to the plate surface, the y axis is shown |
| on figure 3.3. The rate of heat transfer from the plate surface is determined by the conduction |
| into the air at y=0. This can be calculated from, |
|
|
| As a good first estimate we can use, |
|
|
| (3.2) |
|
|
| (3.3) |
|
|
| where the constant C should be of order of magnitude unity. Now substituting this into the |
| definition of the convective heat transfer coefficient, |
|
|
| (3.4) |
|
|
| Thus the heat transfer coefficient is proportional to the |
| thermal conductivity of air and inversely proportional to the |
| size of the thermal boundary layer thickness. |
|
|
| We can use the estimate for h to qualitatively predict how the |
| heat transfer coefficient will vary with the main parameters |
|
|
| First, if the velocity of the air is increased over the plate |
| surface then each fluid element spends a shorter time in |
| contact with, or close to the plate. The fluid element |
| temperatures don’t increase as much. From figure 3.1 and 3.2, in such an instance the thermal |
| boundary layer, , will be smaller at a given x value. Thus an increase in air velocity should result |
| in an increase in the heat transfer coefficient and the rate of convective heat transfer. This result |
| is summarized in table 3.1 |
|
|
| Fig. 3.3 Air temperature in |
| Fig. 3.3 Air temperature in |
| thermal boundary layer |
| thermal boundary layer |
|
|
| Consider now an increase in the air density, say, by an increase in the air pressure, assuming all |
| other air properties and the air velocity remains the same. For the same magnitude of heat |
| transfer the temperature increase of the element will be less since the element has a larger mass. |
|
|
| 0ayTqkysaaTTqCk()()()asaasasakTTkqhTTTT |
|
|
| A smaller temperature increase will reduce the thermal boundary layer thickness. Therefore, an |
| increase in the density will increase the heat transfer coefficient, one reason that liquid water has |
| a higher convective heat transfer rate than air. An increase in the specific heat of the fluid flowing |
| over the plate has the same behavior as an increase in density. |
|
|
| As the plate length is increased the air will flow over the plate for a longer time. Heat transfer will |
| penetrate a further distance from the plate surface. The thermal boundary layer thickness will |
| grow larger. When averaged over the entire plate length , the average thermal boundary layer |
| thickness will become larger and the average heat transfer coefficient will be smaller. Note, the |
| total heat transfer will be higher for the longer plate but the heat transfer per unit area will be |
| smaller. To augment the heat transfer when possible, designers will break a long surface up into a |
| series of smaller surfaces. This can be done by physically separating sections of the plate or by |
| placing an array of ribs at right angles to the flow to break up the boundary layer and restart it. |
|
|
| Table 3.1 Key Parameters Influence on Convective Heat Transfer |
|
|
| Parameter |
|
|
| Velocity increase |
|
|
| Density increase |
|
|
| Specific heat increase |
|
|
| , thermal boundary layer |
| thickness |
|
|
| h, convective heat transfer |
| coefficient |
|
|
| decreases |
|
|
| decreases |
|
|
| decreases |
|
|
| h increases |
|
|
| h increases |
|
|
| h increases |
|
|
| Plate length L increase |
|
|
| averaged over L increases |
|
|
| h averaged over L decreases |
|
|
| Thermal conductivity ka |
| increase |
|
|
| Transition to turbulent flow |
| from laminar flow |
|
|
| increase |
|
|
| h increases |
|
|
| decreases |
|
|
| h increases |
|
|
| When the thermal conductivity of the fluid passing over the plate is increased, y using a higher |
| conductivity gas or liquid there are two elements in play. The thermal boundary layer thickness |
| will increase because of augmented means of heat transfer through the fluid. Remember that h is |
| proportional to the ratio of conductivity to boundary layer thickness. In this case k increases faster |
| than and the heat transfer coefficient increases. It should be expected that when we increase |
| the mechanism for heat transfer, in this case the molecular conductivity, that the rate of |
| convection will increase. |
|
|
| Turbulent Flow |
|
|
| At low velocity, the fluid flows in very smooth paths about parallel to the plate surface. As |
|
|
| the velocity is increased a point is reached where the fluid motion is much more chaotic |
| characterized by eddies in the flow near the plate surface. This is termed turbulent flow. |
| Turbulent flow over a flat plate is found to occur when the Reynolds number |
| 300,000. x is used to indicate the distance from the leading edge; the front of the flat plate can |
| have laminar flow while the rear experiences turbulent flow. The distinction between laminar and |
|
|
| Vx/ exceeds |
|
|
| turbulent flow is important because the eddies in the turbulent flow tend to bring fluid at the |
| ambient temperature Ta much closer to the heated plate surface. In effect the eddies reduce the |
| distance for conduction heat transfer and can markedly increase the heat transfer coefficient, |
| sometimes by an order of magnitude or more. |
|
|
| At can be seen that the convective heat transfer is a function of the fluid properties such as |
| density and conductivity, the flow conditions and the surface geometry. Given below are a few |
| dimensional expressions that can be used for specific cases. The constants in the equations |
| already include the fluid properties. |
|
|
| Laminar flow expressions for h |
| Using air properties at room temperature h for laminar flow over flat plates can be found as |
|
|
| while for water, |
|
|
| (3.5) |
|
|
| (3.6) |
|
|
| In this form h is in BTU/hrft2F, V is in ft/sec and L is in ft. Note, water gives a much higher heat |
| transfer coefficient than air because it has a much higher thermal conductivity as well as a higher |
| density and specific heat. |
|
|
| Turbulent Flow |
|
|
| At low velocity, the fluid flows in very smooth paths about parallel to the plate surface. As |
|
|
| the velocity is increased a point is reached where the fluid motion is much more chaotic |
| characterized by eddies in the flow near the plate surface. This is termed turbulent flow. |
| Turbulent flow over a flat plate is found to occur at higher velocities and longer plate lengths. Also |
| flowing liquids will reach turbulent flow at lower velocities than gases. |
|
|
| For air near room temperature turbulent flow is reached when the product of plate length and air |
| velocity exceeds, |
|
|
| (3.7) |
|
|
| 0.50.71VhL0.512.7VhL50ftVLfts |
|
|
| For water |
|
|
| 22 |
|
|
| (3.8) |
|
|
| (3.9) |
|
|
| Again, all of the parameters are in Imperial units. |
|
|
| Flow Inside Tubes |
|
|
| The other important flow geometry is gas or liquid flow inside tubes. A similar |
| development exists for the convective heat transfer with the exception that h is defined based |
| on the mean temperature TM of the fluid within the tube at the location in question. For a |
| section within the tube of axial length between x and x+ x, |
|
|
| (3.10) |
|
|
| where TM is the mean fluid temperature at x. |
|
|
| Almost all practical cases of tube flow, the flow is turbulent. Exceptions are flows |
|
|
| through very small tube diameters or the flow of viscous fluids such as oil. For turbulent tube |
| at room temperature, the relationship becomes, |
|
|
| while for water we get, |
|
|
| where V is given in ft/sec and D is in feet. |
|
|
| (3.11) |
|
|
| (3.12) |
|
|
| 0.80.20.55VhL0.80.23.921.2ftVLftsVhLSMqhDxTT0.80.20.34VhD0.80.213VhD |
|
|
| Natural or Free Convection |
|
|
| In natural convection the fluid motion is solely due to |
| buoyancy effects caused by the heating or cooling |
| process of the fluid. Natural convection flow over |
| vertical flat plates has similar physical considerations |
| and analogous expressions to those used for forced |
| convection, see Figure 3.4. However, in this case |
| the fluid velocity V is not set by external fans or by |
| the motion of the heated body. The air density varies |
| with 1/T. The air close to the plate is at a higher |
|
|
| Figure 3.4 natural convection on a |
| Figure 3.4 natural convection on a |
| vertical heated plate |
| vertical heated plate |
|
|
| temperature and has a lower mass. This results in a net upward buoyancy force on these |
| elements that accelerates them. We would expect the velocity and the heat transfer coefficient |
| is a function on the temperature difference Ts-Ta. |
|
|
| Just as forced flow, laminar flow exists at low velocities. The exact expression for natural |
| convection of air over a vertical plate at room temperature is |
|
|
| For turbulent flow, |
|
|
| (3.13) |
|
|
| (3.14) |
|
|
| Natural Convection in Enclosed Spaces |
|
|
| In an open wall cavity between interior and exterior walls air circulation will take place |
| from the hot to the cold wall. The air will rise along the hot wall move horizontally to the cold |
| wall at the upper end of the wall cavity (and at other vertical locations as well). The warm air will |
| then flow down the cold wall. This process will result in energy transfer from the hot to the cold |
|
|
| 331/410000.29oTLFftThL31/310000.21TLhT |
|
|
| wall. The overall heat transfer can be represented by a heat transfer coefficient defined in terms |
| of the two wall temperatures, |
|
|
| (3.15) |
|
|
| Figure 3.5 shows measured results for eight foot high walls. hc is a function of the |
| spacing between the walls and , the temperature difference from the hot to the cold wall. At a |
| small spacing and/or a small temperature difference the buoyancy effects are minimal and hc is |
| simply the ratio of air conductivity to wall spacing. As the spacing is increased, hc reaches a |
| constant, typically a spacing between 1/2 and 3/4 inch is optimum. Radiation heat transfer |
| across the cavity must be added to the convection. If there is infiltration of outside air into the |
| cavity the energy transfer may be increased considerably. |
|
|
| ) |
| F |
| G |
| E |
| D |
| 2 |
|
|
| ( |
|
|
| T |
| F |
| - |
| R |
| H |
| U |
| T |
| B |
|
|
| / |
|
|
| , |
|
|
| 0 |
| 6 |
|
|
| ) |
|
|
| c |
| h |
| ( |
|
|
| 10 |
|
|
| 0.9 |
|
|
| 0.8 |
|
|
| 0.7 |
|
|
| 0.6 |
|
|
| 0.5 |
|
|
| 0.4 |
|
|
| 0.3 |
|
|
| 0.2 |
|
|
| 0.1 |
|
|
| 0 |
|
|
| 0 |
|
|
| q |
| 60 F |
| 40 |
| 30 |
| 20 |
| 10 |
|
|
| Conduction only |
|
|
| 0.25 0.50 0.75 1.00 |
|
|
| 1.50 |
|
|
| 2.00 |
|
|
| 2.50 |
|
|
| 3.00 |
|
|
| 3.50 |
|
|
| Air space thickness, l, inches |
|
|
| Data taken from Housing Research Paper 32, "The Thermal Insulating Value of Airspaces", |
| Housing and Home Finance Agency Director of Housing Research, Washington D.C., |
| April 1954. |
|
|
| Image by MIT OpenCourseWare. |
|
|
| ChotwallcoldwallqhATT |
|
|
| 4) Radiation Heat Transfer |
|
|
| Thermal radiation and conduction are the two fundamental mechanisms of heat transfer. |
|
|
| Bodies, such as the sun, emit electromagnetic energy by virtue of their temperature level. The |
| electromagnetic energy is exchanged between bodies at different temperatures giving rise to a |
| net heat transfer. |
|
|
| For heat transfer by radiation the electromagnetic energy generally falls in the visible |
|
|
| range, with wavelengths from 0.4 to 0.7 m, the near infrared, from 0.7 to 25 m in wavelength, or |
| the far infrared from 25 to 1000 m. Electromagnetic radiation in other wavelengths such as X- |
| Rays, and Radio waves is not thermally induced and will be excluded from consideration. |
|
|
| A black body is a body which absorbs all of the radiation incident on its surface over all |
|
|
| wavelengths of importance for heat transfer. The absorbed energy represents an energy |
| transfer to the black body which can contribute to the internal energy increase or it may be |
| transferred through the body by one or more forms of heat transfer. |
|
|
| For a black body at a uniform temperature T, the rate of radiation heat transfer emitted |
|
|
| by the body and leaving the surface, over all wavelengths, is given as |
|
|
| (4.1) |
|
|
| where T is the absolute temperature and |
| value 0.17 x 10-8 BTU/hr ft2 R4 or 5.7 x 10-8 W/m2 K4. Because of the non-linear nature of the |
| expression, radiation becomes progressively more important at higher temperatures although |
| there is considerable radiative heat transfer between bodies at room temperature. |
|
|
| is the Stefan-Boltzmann constant, which has the |
|
|
| The black body radiation is emitted over a range of wavelengths. The emitted energy at |
|
|
| a single wavelength , qre , can be defined so that |
|
|
| (4.2) |
|
|
| Figure 20 shows the distribution of qre as a function of wavelength for black bodies at three |
| temperature levels. For solar radiation, the sun has an effective black body temperature near |
| 10,400 R, a majority of the emitted energy is in the visible wavelengths and the balance of the |
| energy is in the near infrared below 3 m. In contrast, the energy emitted by a black body at |
| room temperature falls in the infrared wavelength range from 8 m to 40 m. For any black body |
| at temperature T eighty percent of the total black body emissions occurs when the product of |
| the wavelength and the absolute temperature, T, is between 4000 ( m)( R) and 17,000 |
| ( m)( R). One half of the total radiation is emitted at wavelengths equal to or below the value of |
| T equal to 7400 ( m)( R). |
|
|
| Material Properties |
|
|
| In the visible wavelengths many common building materials are not good absorbers. |
|
|
| The fraction of the black body radiation absorbed by a body at a given wavelength is given by |
| lie between zero and one. In the visible |
| the monochromatic absorptivity |
| , the values of |
| wavelengths, white paint, wood and plaster have |
| values between about 0.1 and 0.5 see |
| Table 3. In the infrared above 4 or 5 m wavelengths almost all common building materials |
| of 0.8 or higher. Plate glass which is transparent in the visible |
| except shiny metals have |
|
|
| 4emittedqAT0emittedreqqd |
|
|
| wavelength becomes a very good black body in the infrared wavelengths with |
| 0.9. |
|
|
| greater than |
|
|
| Heat Transfer Between Black Bodies |
|
|
| Consider two flat plates, A1 and A2, parallel to each other with the spacing between the |
| surfaces small compared to the width of the plates, see Figure 4.1. The air between the plates |
| can be considered transparent. |
|
|
| Fig. 4.1 Radiation between |
| Fig. 4.1 Radiation between |
| parallel black plates |
| parallel black plates |
|
|
| Each plate is a black body at a uniform temperature T1 and T2, respectively. The radiant energy |
| emitted by the first plate is |
|
|
| (4.3) |
|
|
| All of the energy emitted by plate 1 is absorbed by plate 2. Similarly all of the energy emitted by |
| plate 2, A2 |
|
|
| 4 is absorbed by plate 1. The net radiative heat transfer is |
|
|
| T2 |
|
|
| (4.4) |
|
|
| This heat transfer takes place independent of any conduction or convection heat transfer |
| assuming the temperature T1 and T2 can be maintained. Thus the total heat transfer is the sum |
| of the radiative heat transfer given by Eqn. 70 plus the convective heat transfer between the two |
| plates. |
|
|
| In many cases it is useful to linearize the relationship given by Eqn. 70. Taking the |
| mean temperature as (T1+T2)/2 and T as the temperature difference T1-T2 then T1=TM+ T/2 and |
| T2=TM - T/2 and |
|
|
| 41211qAT441112qATT |
|
|
| When T<<TM, where TM is measured in absolute temperatures, K or R, |
| |
| (4.5) |
| |
| An effective or radiative heat transfer coefficient can be defined so that |
| |
| (4.6) |
| |
| (4.7) |
| |
| Table 4.1 gives values of hr for different values of mean temperature. These values should be |
| compared to the values given for convection in Table 2. At room temperatures heat transfer |
| |
| Table 4.1 hr for radiation at mean temperature T |
| ___________________________________________________________________________ |
| |
| Mean Temperature oF |
| |
| hr (BTU/hr ft2 F) |
| |
| hr(W/m2K) |
| |
| 0 |
| 60 |
| 200 |
| 500 |
| 1500 |
| |
| 0.7 |
| 1.0 |
| 2.0 |
| 6.0 |
| 51 |
| |
| 3.8 |
| 5.7 |
| 11 |
| 34 |
| 290 |
| |
| ___________________________________________________________________________ |
| |
| by radiation is comparable in magnitude to natural convection. For heat transfer between two |
| glass sheets that make up a double glazed window, infrared radiation is important. To reduce |
| the radiation heat transfer one of the glass layers is coated with a very thin metallic layer. For |
| residential buildings this layer allows visible solar radiation to pass but reflects most of the |
| infrared radiation. The layer is referred to as a low e coating. The net infrared radiation can be |
| expressed as, |
| |
| (4.8) |
| |
| where is the emissivity of the glass covered by the metallic layer. Typically can be less than |
| 0.1, substantially reducing the net radiation transfer. |
| |
| For radiation between walls and ceiling or between a person and surrounding surfaces which |
| are each black bodies at different temperatures the geometry of the configuration must be |
| included. For example the radiation heat tranfer between a person and a large glazed wall can |
| be expressed as |
| |
| 2442121212224MTTTTTTTT44312124MTTTTT3124rrrMqhATTwherehT441112qATT |
| |
| (4.9) |
| |
| 28 |
| |
| where F12 is the configuration factor between surface 1, the person, and the glazed surface. F |
| represents the fraction of radiation form the entire surface of the person that is directed toward the |
| glazed wall. When the person is very close to a large glazed surface extending from floor to ceiling, the |
| value of F will be close to ½. On the other hand when the person moves further from the window F will |
| decrease. If the person sits in the center of a cubic room, F from the person to the glazed wall drops to |
| 1/6. |
| |
| 44111212qAFTT |
| |
| MIT OpenCourseWare |
| http://ocw.mit.edu |
| |
| 4.42J / 1.044J / 2.45J Fundamentals of Energy in Buildings |
| Fall 2010 |
| |
| For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms. |
| |
| |