- Email: [email protected]
Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines
Energy xxx (2014) 1e11
Contents lists available at ScienceDirect
Energy journal homepage: www.elsevier.com/locate/energy
Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines Paul Nwachukwu Nwosu a, b, *, Mika Nuutinen a, Martti Larmi a a b
Department of Energy Technology, Aalto University School of Engineering, PO Box 14300, FIN-00076 Aalto, Finland Energy Research Centre and Department of Mechanical Engineering, 41001, University of Nigeria, Nsukka, Nigeria
a r t i c l e i n f o
a b s t r a c t
Article history: Received 30 July 2013 Received in revised form 21 March 2014 Accepted 22 March 2014 Available online xxx
This paper investigates a thermal system that absorbs waste heat from an internal combustion (IC) engine in order to raise the temperature of a working fluid to a saturated state using thermosyphonic flow, non-intrusive of the engine operations. The absorbed heat is rejected to an enclosed space, suitable for in-transit drying. The thermal system comprises a cross-flow heat exchanger connected to a radiator which preheats the working fluid from an insulated (storage) tank. The preheated fluid flows through a radiant heat absorber which absorbs radiant heat from the exhaust manifold. To ensure that the system efficiently performs, a temperature differential is maintained by the heated space while the fluid is cyclically delivered to the tank. The system’s operations are described using a novel flow cycle, and the results indicate a significant heat recovery potential. Ó 2014 Elsevier Ltd. All rights reserved.
Keywords: Thermosyphon Radiant Waste heat absorber Preheater Internal combustion engine
1. Introduction Typically an internal combustion engine converts less than 42% of the chemical exergy available in the fuel into mechanical energy with the remainder converted to thermal energy which is lost to the environment: there are two pathways for utilizing the chemical exergy of the fuel [1], the first involves minimizing exergy destruction in the combustion process while the second involves tapping the exhaust exergy to obtain further improvements in the thermal efficiency of the engine. Besides the heat energy converted into useful work to deliver mechanical power in IC engines, there is a considerable amount of heat losses in the engines which affect the fuel utilization efficiency. A good number of waste heat recovery devices operate on the basis of the Rankine cycle [1,2]. A, methodical approach for employing a waste heat recovery device in truck vehicles based on the Rankine cycle dates back to the early 1970s.A research investigation conducted by Parimal and Doyle [2], Dibella et al. [3] and Doyle et al. [4], on an Organic Rankine Cycle System (ORCS) coupled to a Mack truck diesel engine resulted in an improvement in the
* Corresponding author. Energy Research Centre and Department of Mechanical Engineering, University of Nigeria, Nsukka, Nigeria. Tel.: þ234 (0)8138564432. E-mail addresses: [email protected], [email protected] (P.N. Nwosu).
brake specific fuel consumption (BSFC). Recently, Ghazikaki et al. [5] investigated the effect of exhaust cooling system on exergy recovery in a direct ignition diesel engine, and obtained a BSFC reduction of 5e15% in different load and speed conditions. Similar research programs have been undertaken with some improvements in several ORCS [6,7]. Khatita et al. [8] conducted parametric analysis and optimization study on an ORC system for power generation, with optimal conditions for operation. In addition, Briggs et al. [9] conducted a study on the effects of turbogeneration on an electric hybrid bus, and developed a one-dimensional simulation model, which resulted to considerable reduction in the fuel consumption over a drive cycle. With the exception of turbocompounding, most existing solutions for the recovery of exhaust heat losses utilize a heat exchanger to extract the heat [1,6], such heat exchangers must have areas that match the thermal duty, and this significantly affect vehicle weight. Some other designs are essentially suitable for industrial use [1,7], while others for thermoelectric applications [10]. These units are however often bulky and do not scale with small engine compartments to warrant their use. Considerable improvement in diesel engine’s BSFC can be achieved by utilization of the exhaust energy [6]. Heat engines, due to high combustion temperature and pressure, can be adapted to efficient energy technologies. As research interest into waste energy and scavenging technologies gathers momentum as a result
http://dx.doi.org/10.1016/j.energy.2014.03.091 0360-5442/Ó 2014 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
2
P.N. Nwosu et al. / Energy xxx (2014) 1e11
Nomenclature A C Cp Cr De Ef Et f F FB gc h hþ Isur k L _ m _f m N Nu Pr q Q Qu vt qrad qrad R Re T
qT T0 Up U Up V U DP
r rb rt d
area (m2) heat capacity rate (J/s K) specific heat capacity (J/kg K) heat capacity ratio equivalent diameter (m) total energy in fuel (J) total energy absorbed by cold fluid (J) fin fanning friction factor buoyant force (J) proportionality constant (kg m/N s2) convective heat transfer coefficient (W/m2 K) vertical length of the tank (m) radiation intensity (W/m2) thermal conductivity (W/m K) flow length (m) mass flow rate (kg/s) mass flow rate of fuel (kg/s) number of passes Nusselt number Prandtl number heat flux (W/m2) heat transfer rate (W) useful power (W) tank volume (m3) radiation heat flux (W/m2) average value of qrad (W/m2) resistance (m2 K/W) Reynolds number (m) temperature (K) dimensionless temperature reference temperature overall conductance (W/m2 K) overall conductance (W/m2 K) overall conductance for the absorber (W/m2 K) velocity (m/s) fluid volume in tank (m3) pressure drop (Pa) density (kg/m3) density of fluid at bottom of tank (kg/m3) density of fluid at top of tank (kg/m3) thickness (m)
of rising fuel costs, specific designs of heat recovery systems which are relatively simple and non-intrusive are rarely being investigated. Globally, there are a billion cars [11] with a conservative estimate of about 200 GW of heat losses. These losses can be harnessed to meet some useful thermal applications. Since most of the energy in the fuel is lost as heat, capturing these losses to heat a working fluid to a saturated state (for in-transit drying in food-purveying vehicles, and commercial hot water applications in remote locations where grid power is unavailable) can provide an increase in the fuel conversion efficiency, as well as mitigate environmental impacts arising from the emission of greenhouse gases which arguably can impact the pattern of fuel consumption globally, as well as the environment. In this work, a novel approach for tapping waste heat from an IC engine on the basis of thermosyphonic flow mechanism is studied. It differs from other heat recovery units in its non-intrusive operational design, retrofittability and placement.
m mw s u* yþ G
fus s n g
dynamic viscosity (Ns/m2) the dynamic viscosity (wall quantity) (Ns/m2) StefaneBoltzmann constant (W/m2 K4) dimensionless gas speed (m/s) dimensionless distance source in energy equation dimensionless unsteady temperature dimensionless time gas speed (m/s) adiabatic coefficient
Subscripts c cold f fin me mean εpt particle emissivity h hot H height W width L length min minimum max maximum i h, c in inlet out out a ambient ex exhaust t tube Tot total rad radiation N free stream m manifold p plate rad radiation abs absorber pre preheater c1 cover 1 c2 cover 2 s scale w wall t tube pt particle tot total
The system comprises a cross-flow heat exchanger connected to a radiator for preheating the working fluid from an insulated (storage) tank. The preheated fluid flows through a radiant heat absorber which absorbs radiant heat from the exhaust manifold. The absorbed heat can be utilized in a variety of thermal applications, thereby increasing the engine energy efficiency. The system and its operations e for which a patent is pending e are subsequently described. 2. Methodology of design Descriptively, the present design consists of a two-stage heating process. In the first, waste heat from the engine jacket is used to preheat the working fluid employing a cross-flow heat exchanger (1) connected to a radiator (2), (Fig. 1). Successively, the preheated fluid is passed through a specially designed radiant heat absorber (RHA) (3), consisting of a selectively coated absorber, transparent
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
P.N. Nwosu et al. / Energy xxx (2014) 1e11
3
glazing, protective gauze, and an insulated casing, in order to minimize convective and radiative losses. The model of the RHA is illustrated in Fig. 2. For smooth operations the RHA is coupled to the part of the exhaust manifold adjacent to the engine block, where the temperature of the exhaust gas is maximum. To ensure that the system operates efficiently, a temperature differential is maintained by a heat rejection enclosure (or heated space) (4), this enables the dissipation of excess heat from the fluid downstream, so as to prevent undesired pressure build-up in the insulated tank (5), located upstream of the enclosure. The system operations are described using a novel flow cycle, which consists of an insulated piping (6), six tap valves, which can to isolate flow at any given time, in addition to a pressure relief-valve (V7), which is used to prevent excessive pressure build-up in the tank. 3. Theory 3.1. Preheater In high temperature conditions, considerable amount of heat is mainly lost by radiation in IC engines compared to other modes of heat transfer [12]. The exhaust manifold and the water jackets dissipate significant amount of heat which far much outweigh other non-recoverable losses. In an attempt to entrap the radiant and convective heat which mainly occurs in the vicinity of the exhaust manifold and considerably in the coolant, classical models are employed. The water jackets extract heat from the cylinder walls after fuel combustion, and it is typically connected to a radiator. The inlet and outlet temperatures of the radiator are in the range of 70e95 C and 30e55 C [13], respectively. Notwithstanding that these temperatures are of a low thermal potential, the resulting flux can be tapped for the preheating phase of the heat recovery unit thereby improving the heat extraction performance of the coolant, and in principle reducing the pumping power. The hot fluid (i.e. coolant)
Fig. 2. (a) A CAD model of the radiant heat absorber; and (b) a cut-away view of the model.
Fig. 1. Flow loop of the thermosyphonic flow waste heat recovery system.
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
4
P.N. Nwosu et al. / Energy xxx (2014) 1e11
from the engine passes through a series of heat transfer surfaces of the exchanger and exits to the radiator, inlet hose thus delivering low temperature coolant to the radiator. Cold fluid from the tank flows through the exchanger simultaneously, where it absorbs the heat and then flows to the RHA. Particular detail in the design of the RHA with respect to size and weight are required due to the nature of the intended application where such factors are typically major design constraints, considering that space is a premium in the engine compartment. Thus, the RHA is sized in a manner as to improve heat transfer without significantly adding to the engine weight if a coupled radiator were to be redesigned. Since the size of a radiator connected to the exchanger will potentially get smaller with decreasing thermal duty. The consideration is such that no additional weight results while waste heat is recovered. The physical model and nomenclature of the exchanger plate is shown in Fig. 3. The exchanger comprises an array highly conductive metal plates with a low pressure drop potential. The exchanger is sized employing classical relations of the NTU-effectiveness methodology. This involves determining the heat transfer rate from the hot fluid to the cold fluid on the basis of the exchanger effectiveness, and the temperature difference between the hot and cold fluids at the inlet. The following idealizations are the driving theoretical standpoints for the energy balances and the rate equations i. The exchanger operates in a steady state condition, i.e., temperatures are independent of time, and there is a stable flow rate of both fluids. ii. The exchanger is adiabatic, hence heat losses to the surroundings are negligible. iii. There is no phase change in the fluid streams traversing the exchanger. iv. Specific heats are constant throughout the exchanger. v. Temperature of each fluid is uniform over every crosssectional area of the heat exchange surface. Defining the overall thermal resistance of the exchanger as Ro, then
Ro ¼ Rh þ Rs;h þ Rw þ Rs;c þ Rc
1 1 1 ¼ þ Um A ðhh At Þh hc A f
At ¼ nt ð2$tH $tL þ tW $tH Þ
(3b)
Af ¼ nf ð2$fH $fL þ fW $fH Þ
(3c)
nf is the number of fins, nt is the number of tubes, tH is the tube height, tW is the tube width, tL is the tube length, fH is the fin height, fW is the fin width and fL is the fin length. The convective heat transfer coefficients for the hot and cold fluids, hh and hc, depend on the thermal-fluid properties, velocity and the flow channel geometry. The convective heat transfer coefficient of the hot fluid hh can be obtained employing the DittuseBoelter equation [15] for Re > 5000
Nuh ¼
hh DH;h 1=3 ¼ 0:023Re0:8 h Prh kh
c
(2)
(4a)
where
ReD;h ¼
rh Vh DH;h mh
(4b)
in which case the thermal-fluid properties are evaluated at bulk fluid temperatures. A similar correlation by Sleicher and Rouse [16] gives
Nuh ¼ 5 þ 0:0015Reah Prhb
(1)
1
(3a) c
where
a ¼ 0:88
where h, c, s, w and d denote hot, cold, scale, wall, and the plate thickness, respectively. The above resistances are expressed in terms of individual mean heat transfer coefficients as
d 1 1 1 1 ¼ þ þ þ þ ho hs A ðho hh At Þh ðAw kw Þh ðho hs AÞc Um A ho hc Af
where m denotes the mean value. For prime surfaces (i.e. surfaces with no fins), ho ¼ 1, the thickness of wall and the scale width are typically negligible, then the above yields
(5)
0:24 1 and b ¼ þ 0:5e0:6 Pr 4 þ Pr 3
subject to the constraints
0:1 < Prs < 100:6 and 104 < ReD < 106 where a denotes the film temperature, and b is the surface temperature. Buoyancy-driven flows are largely dependent on density differences. The convective heat transfer coefficient of the cold fluid hc based on the Nusselt number for fully developed laminar flow in a duct of aspect ratio of 0.25 is [12]
Nuc ¼
hc DH;c ¼ 4:439 kc
(6)
and for an aspect ratio of 1/8 [12],the equation gives
Nuc ¼
hc DH;c ¼ 5:597 kc
(7)
For flow conditions in which Re > 5000, the DittuseBoelter equation [13] applies to both fluids The heat duty of the cross-flow exchanger is given by [12]
Qex ¼ xCmin DT Fig. 3. The cross-flow heat exchanger plate.
(8)
where
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
P.N. Nwosu et al. / Energy xxx (2014) 1e11
x ¼
1 exp½ NTUð1 Cr Þ 1 Cr exp½ NTUð1 Cr Þ
(9)
and
DT ¼ Th;in Tc;in
(10)
Cmin Cmax
(11)
Cr ¼
The exchanger effectiveness is a function of the number of transfer units (NTU) and heat capacity ratio Cr. The number of transfer unit is a ratio of the overall conductance UA to the smaller heat transfer capacity rate Cmin, which is a fluid dimensionless residence time, or a modified station number [12]. The heat capacity rates of both fluids can be obtained from the product of the flow rate and the specific heat capacity, given by
Ci ¼ ri Vf ;i Cp;i
(12)
where subscript i denotes min or max and applies to the minimum or maximum of the heat capacity rates of both fluids. The terminal temperatures of both fluids can be predicted by [14]
Tc;out ¼ Tc;in xDTmax
(13)
Th;out ¼ Th;in xDTmax
(14)
where DTmax ¼ Th;in Tc;in . Defining the following nondimensional variables
qh;out ¼
Th;out ; Ta
qc;out ¼
Tc;out ; Ta
qc;in ¼ !
q ¼
Tc;in Ta
DTmax Ta
(15)
(16)
Eqs. (15) and (16) results to
!
qc;out ¼ qc;in x q
!
qh;out ¼ qh;in x q
DT Qex ;g ¼ ; hc Ac Ta hc Ac Ta
G ¼ Vh =Ah
where N is the number of passes, ri is the density of the fluid at inlet, f is the fanning friction factor, De is the equivalent diameter of flow channel, gc is a proportionality constant in Newton’s 2nd law, L is the flow length, ri and ro are the densities of the fluid at inlet and outlet, respectively. 3.2. Radiant heat absorber As hot gases are expelled from the cylinder, the temperature peaks at the junction of the manifold pipes and is subject to a variety drive terrain conditions [18] (Fig. 4). The distribution of the temperature contours shows that in all the cases, maximum temperatures occur at the junction of the manifold tubes, thereby influencing any design and placement considerations with respect to the radiant heat absorber. In terms of losses, a typical gasoline engine releases about 21% of the chemical energy of the fuel at the exhaust alone [19]. This increases to 44% at peak power; this is also dependent on engine size, as well as torque-speed conditions. Light duty passenger (LDP) vehicles operate at exhaust gas temperatures ranging from 500 C to 900 C, while heavy duty (HD) vehicles operate within the range of 500 C to 650 C. The range of waste heat losses is 46e120 kW in light duty 4-cylinder spark ignition engines, and 9e48 kW for the cooling system [16]. On the average, a third of the generated energy is wasted through the exhaust at most common load and speed conditions, this increase to 44% at peak power [6]. The mechanism of heat transfer from the exhaust gas to the pipewall consists of turbulent convective and radiative heat transfer. Although the range of the Reynolds number is 103e5 104 in the exhaust pipe, the flow is actually turbulent [9]. The exhaust valves and the unsteady, flow pulsation do not aid the transition to the laminar flow region. The exhaust flow often enters the region of Re < 2300, especially in the exhaust manifold runners. The heat transfer coefficient due to the turbulent flow can be modeled by NusselteReynolds correlations [9] similar to those of turbulent pipe flows
(18)
where he is the heat transfer coefficient, L is the characteristic length, ke is the gas thermal conductivity, and a is the turbulent pipe flow condition coefficient. The following correlation proposed by Gnielinksi [19] for calculating convective heat transfer from the exhaust gas to the pipe wall takes into account the effects of the wall roughness, employing a friction factor, F:
(19)
(20)
Notably, pressure drop is an important design consideration for heat exchangers. If pressure drop is considerable, flow is affected with a potential for flow stagnation. Hence, variables which can keep to the barest minimum the pressure drop performance of the exchanger are of important considerations. The correlation for pressured drop for plate heat exchangers is given by [17]
DP ¼
1:5G2 N 4fLG2 1 1 1 G2 þ þ ro ri gc 2gc ri 2gc De r me
(21a)
ð1=rÞme ¼ 1 2ð1=ri þ 1=ro Þ
Nue ¼
(21b)
(22)
F 8 ðRee
1000ÞPre qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1:07 þ 12:7 8f Pr 2=3 1
(23)
subject to the constraint
104 < Ree < 5 106 and for Re < 104
Nue ¼
F e Pre 8 Re rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1:07 þ 12:7
where
he L ¼ aðReÞ0:8 ke
Nue ¼
Combining Eqs. (10), (13) and (18) yields
! Q ¼ xCr Cmax g
(21c)
(17)
Defining also
! Q ¼
5
f 8
2=3 Pre 1
(24)
the correlation for calculating convective heat transfer coefficient for a moving vehicle (from the manifold to the ambient) is derived using the combined heat transfer coefficient [9]
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
6
P.N. Nwosu et al. / Energy xxx (2014) 1e11
Fig. 4. Contour plots of temperature distribution in an exhaust manifold in variable conditions [18].
Nue;b
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ 0:3 Nu2laminar þ Nu2turbulent
(25a)
qw ¼
2:1ln yþ þ 2:5 0 s g1 Z g kr u* r 1 0 @ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi þ 0:082 kT0 mw r0 pðs qT Þ 0 pffiffiffiffiffiffiffiffiffiffiffiffiffi!!1 s qT A dfus ð0; qT ÞdqT 1 exp 3 dqT
rCp u* TlnðT=Tw Þ 2:1yþ þ 33:4 Gn u*
subject to the constraint
10 < Re < 107 where
Nulaminar
pffiffiffiffiffiffiffiffipffiffiffiffiffi 3 ¼ 0:664 Re3 Pr
Nuturbulent
0:037Re0:8 Pr ¼ 1 þ 2443Re0:1 Pr2=3 1
where
(25b)
(25c)
Radiative heat transfer is relatively significant in the area adjoining the manifold, due to the presence of hot radiating particles in the gases. At temperatures lower than 500 C, radiation from the hot gases is relatively small and can be neglected [9]. But for temperatures exceeding that threshold, the radiation flux from the particulate matter is substantial, and can be expressed by [19]
4 qrad ¼ εpt sTpt
ð27aÞ
(26)
In diesel engines, the order of magnitude of qrad is given by qrad ¼ 0:2qtotal , where qtotal is the total cycle cumulative flux. At peak valve, qrad;max z0:4qtotal;max , where qtotal;max is the total maximum peak flux value. The order of magnitude of qrad is about approximately 1e3 MW/m2 for SI engines, and w10 MW/m2 for diesel engines, which occurs around the cylinder head [16]. The correlation proposed by Han and Reitz [20] for estimating the cylinder wall heat flux takes into consideration various operational variables and parameters:
u* ¼
qffiffiffiffiffiffiffiffiffiffiffiffi 1=2 kCm
(27b)
where u* is the dimensionless gas speed, yþ is the dimensionless distance, fus is the dimensionless unsteady temperature, s is the dimensionless time, n is the gas speed, T0 is the reference temperature, T and qT are temperatures and dimensionless temperatures, G is the source term in the energy equation, and mw is the dynamic viscosity (wall quantity). The placement of the absorber is critical to thermal performance. The radiant heat absorber is coupled to the manifold at the vicinity where the temperature is a maximum so as to maximize its heat absorption potential. The control volume is schematically idealized in Fig. 5, illustrating the mechanism of heat transfer. The impinging radiation strikes the absorber (coated with a selective coating of high absorptivity and negligible emissivity) through a glass envelope. The total heat transfer rate from the manifold to the surrounding is given by
4 Isur ¼ εm sAm Tm Ta4 þ ha Am ðTm TN Þ
(28)
where εm is the emissivity of the manifold surface, s is the Stefan– Boltzmann constant, Am is the manifold heat exchange area, ha is the convective heat transfer coefficient of ambient air, Tm is the temperature of the manifold, Ta is the ambient temperature and TN
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
P.N. Nwosu et al. / Energy xxx (2014) 1e11
is the free stream temperature. The above relation comprises convective and radiative contributions. Given that in high temperature conditions, radiative heat transfer is the dominant mode of heat transfer [21], assuming this condition and neglecting convection, the above equation yields
4 Ir;sur ¼ εm sAm Tm Ta4
(29)
Considering the above equation, the total useful heat gain by the working fluid(which is the net heat gain of the absorber)is written as
Qtot;abs ¼ saIsur Up Aabs ðTabs Ta Þ
(30)
where Up is the overall loss conductance which accounts for the heat loss components between the absorber and the glass envelope in addition to that of the glass envelope and the surrounding, Aabs is the absorptive surface area of the absorber. Up is a measure of the thermal resistance the absorber components pose to the flow of heat. Ignoring the conductive resistance of the glass, for a double glazed absorber, Up is expressed by
Up ¼
1 1 hc1
(31)
þ h1 þ h1 c2
7
xc ¼
Tc;in Tc;outabs T T 3 saεm sAm ; xp ¼ abs ; f ¼ a ; qw;in ¼ _ p Ta Ta Ta mC
xe ¼
Up Aabs saha Am Tm ;c ¼ ;g ¼ _ p _ p Ta mC mC
(35)
where f is the dimensionless radiant heat, c is the dimensionless absorber materials property, and g is the dimensionless manifold emissive area. Assuming Ta z TN, combining Eqs. (28) (30), (34) and (35), consequently
xc ¼ qw;in þ f x4e 1 þ gðxe 1Þ c xp 1
(36)
Since radiation is pre-dominant in high temperatures for this case, neglecting the convection term, the above equation becomes
xc ¼ qw;in þ f x4e 1 c xp 1
(37)
To determine the energy efficiency of the engine as a result of the coupled waste heat recovery system, comprising the preheater and the absorber, the relation which expresses the total energy absorbed by the fluid is
Etot ¼ mCp Tc;outabs Tc;in
(38)
The energy input of the fuel is given by [22]
a
The exchanger (or the preheater) coupled to the radiator extracts heat from the hot coolant, supposing there is no phase change within the preheater e this is plausible due to the finite terminal temperature of the coolant e the overall heat gain of the working fluid is the sum of the heat gain of the fluid as it flows through the preheater and the absorber, respectively given by
_ p Tc;out Tc;in Qpre ¼ mC
(32)
and
$
Efu ¼ mf $LCV
(39)
where mf is the mass flow rate of the fuel and LCV is the lower calorific value, which gives an indication of the amount of energy released on combustion. The energy efficiency of the waste heat recovery system is
hsystem ¼
Etot Efu
(40)
Consequently, the energy efficiency of the engine with the coupled waste heat recovery system is
_ p Tc;outabs Tc;in Qabs ¼ mC abs
(33)
where Tc;inabs and Tc;outabs are the inlet and outlet temperatures of the working fluid which flows through the absorber. Tc,in and Tc,out are the inlet and outlet temperatures of the working fluid which flows through the preheater. The total heat gain of the working fluid, assuming negligible path losses with Tc;out zTc;inabs is the summation of the last two energy rate equations
_ p Tc;outabs Tc;in Qtot;abs ¼ mC
(34)
Defining also the following dimensionless variables
heff ¼ hengine þ hsystem
(41)
heff is the typical engine efficiency, and assuming steady state conditions, the heat absorbed by the working fluid rapidly heats up the fluid in the tank, charging the tank. The charging rate depends on the heat input and the path losses. With increased heat input, the fluid in the tank becomes thermally stratified, and this is sometimes used as a measure of its operational performance [23]. As the fluid becomes stratified, the density gradient results to differential buoyant forces, which cause convection by gravity. The buoyancy force FB resulting from the difference in density and pressure is defined by [23]
Fig. 5. Schematic diagram of the control volume of the radiant heat absorber.
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
8
P.N. Nwosu et al. / Energy xxx (2014) 1e11
Fig. 6. Effects of nondimensional fin width on the terminal temperatures of the participating fluids for a range of qmax values.
Fig. 8. Effects of nondimensional manifold surface temperatures on the cold fluid terminal temperatures for varying c values.
FB ¼ gðrb rt Þhþ
tank, (UA)t is the overall conductance of the tank wall. Optimally, the insulated tank is mounted at least a unit foot above the absorber [23], so that buoyant forces are aided by the height differential in order to avoid flow stagnation in the piping.
(42)
where rb is the density of the fluid at the bottom of the tank, rt is the density of the fluid at the top of the tank, hþ is the vertical length between the top and bottom of the tank, and g is the gravitational force. An energy balance on a well-mixed storage tank that is rapidly charged by the absorbed heat is defined by
rUCp dTt dt
¼ Qu Ls ðUAÞt ðTt Ta Þ
(43)
where Tt is the tank temperature, Qu is the useful heat gain of the tank, Ls is the extracted load of the tank, U is the fluid volume in the
Fig. 7. Effects of nondimensional fin width on the fluid terminal temperatures for varying Cmax/Cmin values.
4. Results and discussion In Fig. 6, the thermal performance of the preheater, i.e. the exchanger, indicated by a variation in the nondimensional (terminal) temperatures of the participant fluids qi is analyzed, vis-à-vis the nondimensional fin width fw/tw and the maximum temperature
Fig. 9. Effects of nondimensional parameter f on the cold fluid terminal temperatures xc for varying xe values.
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
P.N. Nwosu et al. / Energy xxx (2014) 1e11
9
difference of the fluids qmax. As the nondimensional fin width increases, the nondimensional cold fluid terminal temperatures decrease correspondingly, indicating that flow channel dimensions are critical to thermal performance of the exchanger. With very low fW/tW values, qi increases markedly, particularly for the cold fluid. On the contrary, the hot fluid terminal temperature is lowest when fW/tW is a minimum. Inferably, as the fin width decreases the cold fluid becomes increasingly in contact with the heated plates, thus heat is transferred more rapidly in this case than in a case involving increasing fin width dimensions. Ostensibly, the terminal temperatures of both fluids are affected by changes in the fin width. The value of qi is a maximum when qmax ¼ 0.2681 for the cold fluid, and as the nondimensional fin width decreases, the terminal fluid temperatures are also affected by small differences in qmax parameter. The significance of this is that the amount of heat transferred from the hot fluid to the cold fluid increases and decreases respectively with a decrease and increase in the fin width dimensions. Fig 7 illustrates the evolution of the nondimensional terminal fluid temperatures of both fluids qi with the nondimensional fin width fW/tW in relation to the heat capacity ratios Cmax/Cmin. The nondimensional fin width gives an indication of the dimensions of the flow passage of the cold fluid in relation to the tube width tw, which is the hot fluid flow passage variable. The heat capacity ratios are essentially ratios of the specific heats of both working fluids. As the nondimensional fin width increases from a minimum value of about 2 to 28, the nondimensional (cold) fluid terminal temperature decreases, correspondingly. At a value of fW/tW ¼ 2, the value of qi for the cold fluid is about 1.22 at Cmax/Cmin ¼ 0.8794. Relatively for varying values of Cmax/Cmin, we observe a near opposite trend for the hot fluid. What can be inferred thus is that as the fin width decreases, the cold fluid is more in contact with the heated plates, and heat is thus transferred more rapidly than for increasing fin width dimensions which implies reduced contact between the plates and the fluid. For low values of Cmax/Cmin, the cold fluid terminal temperature increases irrespective of fW/tW values. Hence, cold fluids with relatively high heat capacities will extract more heat from the exchanger plates than the fluids with lower heat capacity rate for prescribed hot fluid properties. A major design
concern of any waste heat recovery system is to select working fluids which can extract the maximum (possible) amount of heat from the system while minimizing irreversibility. From the figure it can be adduced that the waste heat recovery system would have high heat extraction rates with low values of Cmax/Cmin. Notably, the cold fluid terminal temperatures drop with increasing fW/tW values for a given Cmax/Cmin value, indicating that the flow passage dimensions and the specific heat capacities of both fluids can considerably influence heat transfer in the exchanger. Regardless of changes in the Cmax/Cmin parameter, there is negligible difference in qi with high values of fW/tW. In Fig. 8, the nondimensional cold fluid terminal temperature xc is plotted against the nondimensional exhaust manifold temperature xe in relation to the variable c. c gives an indication of the relative change in the absorber parameters (the overall conductance and the absorptive area of the radiant heat absorber). As xe increases from 1 to 6, parameter xc increases non-linearly. The values of xc are a maximum at xe ¼ 6 and c ¼ 0.0, showing that increasing manifold temperatures can significantly enhance the cold fluid terminal temperatures. Conversely, the obtained values of xc are a minimum at xe ¼ 1.0 and c ¼ 2.0. Thus, if c tends to zero, there is a prospect of high terminal cold fluid temperatures since the parameter is coupled to thermal resistance within the absorber, which is based on the values of the heat transfer coefficients (of the absorber- to-glass, and glass-to-ambient). Typically low c values entail reduced convective and emissive losses from the absorber and improved heat absorption by the working fluid. But there is a limit as to values which Up can attain (hence c) in typical absorbers, owing to the finite conductivity of the materials. It is plausible to select materials and production processes which can potentially reduce parameter c values, by ensuring air tight properties of the glass envelope, as well as the use of a selective glazing with desired properties. Also, the two cases involving radiation and convection, and the lone radiation mode are analyzed. It is notable from the plot that the twin modes of radiation and convection, ‘with convection’, considerably improved heat transfer relative to the lone radiation mode, ‘without convection’. Therefore in moving vehicles, if the hot fluid stream around the manifold and the underhood, which is typically turbulent is directed toward the absorber, heat transfer will be improved. Effects of variation of xc, the nondimensional terminal fluid temperature, with f and parameter xe is shown in Fig. 9. As f
Fig. 10. Effects of tube dimensions on pressure drop.
Fig. 11. Thermal efficiency vs. flow rate.
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
10
P.N. Nwosu et al. / Energy xxx (2014) 1e11
increases from 0.005 to 0.015, xc increases correspondingly, indicating the effects of ambient temperature, the absorber material and the flow prescription on the cold fluid terminal temperatures. For increasing xe values, the dependence of xc on f is apparently strong. Clearly, there exists a linear relationship between xc and f for increasing xe values. The effects of f on xc becomes weak for low values of parameter xe. Since f is a lumped variable, high values of f for a given absorber dimension, flow prescription and ambient condition, indicate the effects of increasing absorptivity of the coating relative to emitted radiation from the manifold, thus giving an indication of the net absorbed radiation. Consequently, increasing f values entail increasing coating absorptivity values, and thus improved absorption. Absorptive coatings have been used to enhance heat transfer in energetic systems, especially in cases where heat transfer occurs mainly by radiation. Therefore for operational conditions characterized by increasing manifold and ambient temperatures in relation to increased absorptive area and high coating absorptivity values, heat transfer will improve due to the fact that high ambient temperatures potentially lower the prospect of low manifold surface temperatures thereby improving exergy. Another important design consideration is pressure drop. The power required to move the coolant across the exchanger plates is dependent on the frictional pressure drop which is often a major operating expense. The plate fin exchanger considered here has the lowest pressure drop performance, compared to other exchanger designs [25]. In Fig. 10, the pressure drop variable DP is plotted against the tube width tW e which represents the width of the (coolant) flow channel e in relation to the ratio of the tube height and length tH/tL. Importantly, these variables give an indication of the heat exchange surface dimensions. When tw increases the pressure drop decreases, for this particular case, a significant portion of the fluid mass is probably not in contact with the heated plates thus there is lower frictional drag. However for decreasing tube widths, the pressure drop increases markedly since most of the fluid mass is in contact with the exchanger plates hence leading to increased frictional drag and pressure drop. As tH/tL decreases, the pressure drop is affected as a result. Low tH/tL values give rise to increased pressure drop. Notably, the smaller the plate dimensions, the more the pressure drop due to the increased drag although heat transfer is improved as a result. Typically, a trade-off in the thermal and pressure drop functions is usually obtained by optimization [24], so as to offset any gain in one of the functions that could adversely affect the other. The inter-dependence between pressured drop, thermal performance and flow channel dimensions is critical to the overall thermal performance, hence by optimization an enhancement in the heat transfer as well as pressure drop performance would be realized. In Fig. 11, it is seen that the heat extraction efficiency heff of the waste heat recovery system decreases with increasing volumetric flow rate of the cold fluid Vc. It is assumed that the working fluid flows through the preheater and the absorber at the same flow rate, given that the fluid velocity is driven by thermosyphonic flow. As the flow rate increases, the efficiency decreases. The efficiency is about 30% at a flow rate of 0.1 m3/s, while the highest value of heff observed in the plot is about 67%, at a value of Vc ¼ 0.02 m3/s due to the fact that the heat extraction efficiency improves with flow velocity and the convective heat transfer coefficient of the working fluid, in line with Newton’s law of cooling [14]. Lower values of Vc imply a higher heat extraction potential, giving rise to increased efficiencies. Although the flow velocity is driven by the thermosyphonic flow mechanism, it is possible to influence the flow rate by introducing bends in the piping without appreciably impacting pressure drop in the flow loop. However, excessive bends in the piping would reduce the flow rate and potentially lead to increased
pressure build up. A good alternative is to optimize the efficiency on the basis of the flow velocity and pressure drop without penalizing the thermal performance. 5. Conclusion A thermosyphonic waste heat recovery system for IC engines is investigated. Heat transfer modeling of temperature and the operational variables showed that this design has the potential to improve thermal efficiency in IC engines while delivering high temperature working fluid. The system can be fabricated into a variety of sizes to fit the engine compartment, as well as conform to vehicular weight specifications. The technology is non-intrusive of the engine operations and allows for a modular design as space becomes a premium in the engine compartment: a good design of any vehicle engine waste heat recovery system must be lightweight, and adapt to the engine space provisions. It is recommended that appropriate waste heat recovery technologies which are, lightweight and non-intrusive of the engine operations be given practical considerations as a feasible solution for improved energy efficiency. Acknowledgment The efforts of Prof Risto Lahdelma of the Energy Technology Department of Aalto University are acknowledged. References [1] Tommi P, Seepo N, Pekka R. Waste heat recovery-bottoming cycle alternatives. In: Proceedings of the University of Vaasa; 2012. pp. 1e15. Reports 175. [2] Parimal PS, Doyle EF. Compounding the tuck diesel engine with an organic Rankine cycle system; 1976. SAE Paper 760343. [3] Dibella FA, Di Nanno LR, Koplow MD. Laboratory and on-highway testing of diesel organic Rankine compound long-haul vehicle; 1983. SAE Paper 830122. [4] Doyle E, Di Nanno LR, Kramer S. Installation of a dieseleorganic rankine compound engine in a class 8 truck for a single-vehicle test; 1976. SAE Paper 790646. [5] Mohsen G, Mohammad H, Davood D, Mofid G, Behravana Ali, Gholamreza S. Exergy recovery from the exhaust cooling in a DI diesel engine for BSFC reduction purposes. Energy 2014;65:44e51. [6] Teng H, Regner G, Cowland C. Waste heat recovery of heavy duty diesel engines by organic rankine cycle. Part II: working fluids for WHR-ORC; 2005. SAE 01-0543. [7] Wang T, Zhang Y, Peng Z, Shu G. A review of researches on thermal exhaust heat recovery with rankine cycle. Renew Sustain Energy Rev 2008;15:2862e71. [8] Mohammed A, Tamer S, Fatma H, Ibrahim M. Power generation using waste heat recovery by organic Rankine cycle in oil and gas sector in Egypt: a case study. Energy 2014;64:462e72. [9] Ian B, Geoffrey M, Stephen S, Roy D. Whole-vehicle modelling of exhaust energy recovery on a diesel electric hybrid bus. Energy 2014;65:172e81. [10] Kumar CR, Sonthalai A, Goel R. Experimental study on waste heat recover from an internal combustion engine using thermoelectric technology. Therm Sci 2011;15(4):1011e22. [11] Sousanis J. World vehicle population tops 1 billion units. WardsAuto; Aug. 15, 2011. http://wardsauto.com/ar/world_vehicle_population_110815. [12] Kandylas IP, Stamatelos AM. Engine exhaust system design based on heat transfer computation. Energy Convers Manage 1999;40:1055e72. [13] Vasu V, Krishna RK, Kumar ACS. Thermal design analysis of compact heat exchanger using nanofluids. Int J Nanomanufact 2008;2(3):271e88. [14] Kern F, Bohn SM. Principles of heat transfer. New York: Harper and Row; 1992. [15] Dittus FW, Boelter LMK. Heat transfer in automobile radiators of the tubular type2. University of California, Berkeley Publications in Engineering; 1930. p. 443. [16] Sleicher CA, Rouse MW. A conventional correlation for heat transfer to constant and variable property fluids in turbulent pipe flow. Int J Heat Transfer 1975;18:677e83. [17] Shah RK, Sekulic DP. Fundamentals of heat exchanger design. Hoboken, NJ: John Wiley and Sons Inc.; 2003. [18] Krenn C, Hofer G, Prosi M, Samhaber C, Pessl G. Thermal simulation of an exhaust manifold in a high performance diesel engine. In: Star European User Conference, March 17-18th, 2008, London; 2008. [19] Gnielinski V. New equations for heat and mass transfer in turbulent pipe and channel flow. Int Chem Eng 1976;16:359e68.
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
P.N. Nwosu et al. / Energy xxx (2014) 1e11 [20] Han Z, Reitz RD. Temperature wall function formulation for variable-density turbulent flows with application to engine convective heat transfer modelling. Int J Heat Mass Transfer 1997;40(3):613e25. [21] Gray WA, Kilham JK. Heat transfer from flames. London: Elek Science; 1976. [22] Bourhis G, Leduc P. Energy and exergy balances for modern diesel and gasoline engines. Oil Gas Sci Technol Rev 2010;65(No. 1):39e46. IFP. [23] C., Currie J, Muneer T. Appl Energy 2009;86:1465e9.
11
[24] Nwosu PN. Employing exergy-optimized pin fins in the design of an absorber in a solar air heater. Energy 2010;35:571e5. [25] Mavridou S, Mavropoulos GC, Bouris D, Hountalas DT, Bergeles G. Comparative study of a diesel exhaust gas heat exchanger for truck applications with conventional and state of the art heat transfer enhancement. Appl Therm Eng 2010;30:935e47.
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
Contents lists available at ScienceDirect
Energy journal homepage: www.elsevier.com/locate/energy
Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines Paul Nwachukwu Nwosu a, b, *, Mika Nuutinen a, Martti Larmi a a b
Department of Energy Technology, Aalto University School of Engineering, PO Box 14300, FIN-00076 Aalto, Finland Energy Research Centre and Department of Mechanical Engineering, 41001, University of Nigeria, Nsukka, Nigeria
a r t i c l e i n f o
a b s t r a c t
Article history: Received 30 July 2013 Received in revised form 21 March 2014 Accepted 22 March 2014 Available online xxx
This paper investigates a thermal system that absorbs waste heat from an internal combustion (IC) engine in order to raise the temperature of a working fluid to a saturated state using thermosyphonic flow, non-intrusive of the engine operations. The absorbed heat is rejected to an enclosed space, suitable for in-transit drying. The thermal system comprises a cross-flow heat exchanger connected to a radiator which preheats the working fluid from an insulated (storage) tank. The preheated fluid flows through a radiant heat absorber which absorbs radiant heat from the exhaust manifold. To ensure that the system efficiently performs, a temperature differential is maintained by the heated space while the fluid is cyclically delivered to the tank. The system’s operations are described using a novel flow cycle, and the results indicate a significant heat recovery potential. Ó 2014 Elsevier Ltd. All rights reserved.
Keywords: Thermosyphon Radiant Waste heat absorber Preheater Internal combustion engine
1. Introduction Typically an internal combustion engine converts less than 42% of the chemical exergy available in the fuel into mechanical energy with the remainder converted to thermal energy which is lost to the environment: there are two pathways for utilizing the chemical exergy of the fuel [1], the first involves minimizing exergy destruction in the combustion process while the second involves tapping the exhaust exergy to obtain further improvements in the thermal efficiency of the engine. Besides the heat energy converted into useful work to deliver mechanical power in IC engines, there is a considerable amount of heat losses in the engines which affect the fuel utilization efficiency. A good number of waste heat recovery devices operate on the basis of the Rankine cycle [1,2]. A, methodical approach for employing a waste heat recovery device in truck vehicles based on the Rankine cycle dates back to the early 1970s.A research investigation conducted by Parimal and Doyle [2], Dibella et al. [3] and Doyle et al. [4], on an Organic Rankine Cycle System (ORCS) coupled to a Mack truck diesel engine resulted in an improvement in the
* Corresponding author. Energy Research Centre and Department of Mechanical Engineering, University of Nigeria, Nsukka, Nigeria. Tel.: þ234 (0)8138564432. E-mail addresses: [email protected], [email protected] (P.N. Nwosu).
brake specific fuel consumption (BSFC). Recently, Ghazikaki et al. [5] investigated the effect of exhaust cooling system on exergy recovery in a direct ignition diesel engine, and obtained a BSFC reduction of 5e15% in different load and speed conditions. Similar research programs have been undertaken with some improvements in several ORCS [6,7]. Khatita et al. [8] conducted parametric analysis and optimization study on an ORC system for power generation, with optimal conditions for operation. In addition, Briggs et al. [9] conducted a study on the effects of turbogeneration on an electric hybrid bus, and developed a one-dimensional simulation model, which resulted to considerable reduction in the fuel consumption over a drive cycle. With the exception of turbocompounding, most existing solutions for the recovery of exhaust heat losses utilize a heat exchanger to extract the heat [1,6], such heat exchangers must have areas that match the thermal duty, and this significantly affect vehicle weight. Some other designs are essentially suitable for industrial use [1,7], while others for thermoelectric applications [10]. These units are however often bulky and do not scale with small engine compartments to warrant their use. Considerable improvement in diesel engine’s BSFC can be achieved by utilization of the exhaust energy [6]. Heat engines, due to high combustion temperature and pressure, can be adapted to efficient energy technologies. As research interest into waste energy and scavenging technologies gathers momentum as a result
http://dx.doi.org/10.1016/j.energy.2014.03.091 0360-5442/Ó 2014 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
2
P.N. Nwosu et al. / Energy xxx (2014) 1e11
Nomenclature A C Cp Cr De Ef Et f F FB gc h hþ Isur k L _ m _f m N Nu Pr q Q Qu vt qrad qrad R Re T
qT T0 Up U Up V U DP
r rb rt d
area (m2) heat capacity rate (J/s K) specific heat capacity (J/kg K) heat capacity ratio equivalent diameter (m) total energy in fuel (J) total energy absorbed by cold fluid (J) fin fanning friction factor buoyant force (J) proportionality constant (kg m/N s2) convective heat transfer coefficient (W/m2 K) vertical length of the tank (m) radiation intensity (W/m2) thermal conductivity (W/m K) flow length (m) mass flow rate (kg/s) mass flow rate of fuel (kg/s) number of passes Nusselt number Prandtl number heat flux (W/m2) heat transfer rate (W) useful power (W) tank volume (m3) radiation heat flux (W/m2) average value of qrad (W/m2) resistance (m2 K/W) Reynolds number (m) temperature (K) dimensionless temperature reference temperature overall conductance (W/m2 K) overall conductance (W/m2 K) overall conductance for the absorber (W/m2 K) velocity (m/s) fluid volume in tank (m3) pressure drop (Pa) density (kg/m3) density of fluid at bottom of tank (kg/m3) density of fluid at top of tank (kg/m3) thickness (m)
of rising fuel costs, specific designs of heat recovery systems which are relatively simple and non-intrusive are rarely being investigated. Globally, there are a billion cars [11] with a conservative estimate of about 200 GW of heat losses. These losses can be harnessed to meet some useful thermal applications. Since most of the energy in the fuel is lost as heat, capturing these losses to heat a working fluid to a saturated state (for in-transit drying in food-purveying vehicles, and commercial hot water applications in remote locations where grid power is unavailable) can provide an increase in the fuel conversion efficiency, as well as mitigate environmental impacts arising from the emission of greenhouse gases which arguably can impact the pattern of fuel consumption globally, as well as the environment. In this work, a novel approach for tapping waste heat from an IC engine on the basis of thermosyphonic flow mechanism is studied. It differs from other heat recovery units in its non-intrusive operational design, retrofittability and placement.
m mw s u* yþ G
fus s n g
dynamic viscosity (Ns/m2) the dynamic viscosity (wall quantity) (Ns/m2) StefaneBoltzmann constant (W/m2 K4) dimensionless gas speed (m/s) dimensionless distance source in energy equation dimensionless unsteady temperature dimensionless time gas speed (m/s) adiabatic coefficient
Subscripts c cold f fin me mean εpt particle emissivity h hot H height W width L length min minimum max maximum i h, c in inlet out out a ambient ex exhaust t tube Tot total rad radiation N free stream m manifold p plate rad radiation abs absorber pre preheater c1 cover 1 c2 cover 2 s scale w wall t tube pt particle tot total
The system comprises a cross-flow heat exchanger connected to a radiator for preheating the working fluid from an insulated (storage) tank. The preheated fluid flows through a radiant heat absorber which absorbs radiant heat from the exhaust manifold. The absorbed heat can be utilized in a variety of thermal applications, thereby increasing the engine energy efficiency. The system and its operations e for which a patent is pending e are subsequently described. 2. Methodology of design Descriptively, the present design consists of a two-stage heating process. In the first, waste heat from the engine jacket is used to preheat the working fluid employing a cross-flow heat exchanger (1) connected to a radiator (2), (Fig. 1). Successively, the preheated fluid is passed through a specially designed radiant heat absorber (RHA) (3), consisting of a selectively coated absorber, transparent
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
P.N. Nwosu et al. / Energy xxx (2014) 1e11
3
glazing, protective gauze, and an insulated casing, in order to minimize convective and radiative losses. The model of the RHA is illustrated in Fig. 2. For smooth operations the RHA is coupled to the part of the exhaust manifold adjacent to the engine block, where the temperature of the exhaust gas is maximum. To ensure that the system operates efficiently, a temperature differential is maintained by a heat rejection enclosure (or heated space) (4), this enables the dissipation of excess heat from the fluid downstream, so as to prevent undesired pressure build-up in the insulated tank (5), located upstream of the enclosure. The system operations are described using a novel flow cycle, which consists of an insulated piping (6), six tap valves, which can to isolate flow at any given time, in addition to a pressure relief-valve (V7), which is used to prevent excessive pressure build-up in the tank. 3. Theory 3.1. Preheater In high temperature conditions, considerable amount of heat is mainly lost by radiation in IC engines compared to other modes of heat transfer [12]. The exhaust manifold and the water jackets dissipate significant amount of heat which far much outweigh other non-recoverable losses. In an attempt to entrap the radiant and convective heat which mainly occurs in the vicinity of the exhaust manifold and considerably in the coolant, classical models are employed. The water jackets extract heat from the cylinder walls after fuel combustion, and it is typically connected to a radiator. The inlet and outlet temperatures of the radiator are in the range of 70e95 C and 30e55 C [13], respectively. Notwithstanding that these temperatures are of a low thermal potential, the resulting flux can be tapped for the preheating phase of the heat recovery unit thereby improving the heat extraction performance of the coolant, and in principle reducing the pumping power. The hot fluid (i.e. coolant)
Fig. 2. (a) A CAD model of the radiant heat absorber; and (b) a cut-away view of the model.
Fig. 1. Flow loop of the thermosyphonic flow waste heat recovery system.
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
4
P.N. Nwosu et al. / Energy xxx (2014) 1e11
from the engine passes through a series of heat transfer surfaces of the exchanger and exits to the radiator, inlet hose thus delivering low temperature coolant to the radiator. Cold fluid from the tank flows through the exchanger simultaneously, where it absorbs the heat and then flows to the RHA. Particular detail in the design of the RHA with respect to size and weight are required due to the nature of the intended application where such factors are typically major design constraints, considering that space is a premium in the engine compartment. Thus, the RHA is sized in a manner as to improve heat transfer without significantly adding to the engine weight if a coupled radiator were to be redesigned. Since the size of a radiator connected to the exchanger will potentially get smaller with decreasing thermal duty. The consideration is such that no additional weight results while waste heat is recovered. The physical model and nomenclature of the exchanger plate is shown in Fig. 3. The exchanger comprises an array highly conductive metal plates with a low pressure drop potential. The exchanger is sized employing classical relations of the NTU-effectiveness methodology. This involves determining the heat transfer rate from the hot fluid to the cold fluid on the basis of the exchanger effectiveness, and the temperature difference between the hot and cold fluids at the inlet. The following idealizations are the driving theoretical standpoints for the energy balances and the rate equations i. The exchanger operates in a steady state condition, i.e., temperatures are independent of time, and there is a stable flow rate of both fluids. ii. The exchanger is adiabatic, hence heat losses to the surroundings are negligible. iii. There is no phase change in the fluid streams traversing the exchanger. iv. Specific heats are constant throughout the exchanger. v. Temperature of each fluid is uniform over every crosssectional area of the heat exchange surface. Defining the overall thermal resistance of the exchanger as Ro, then
Ro ¼ Rh þ Rs;h þ Rw þ Rs;c þ Rc
1 1 1 ¼ þ Um A ðhh At Þh hc A f
At ¼ nt ð2$tH $tL þ tW $tH Þ
(3b)
Af ¼ nf ð2$fH $fL þ fW $fH Þ
(3c)
nf is the number of fins, nt is the number of tubes, tH is the tube height, tW is the tube width, tL is the tube length, fH is the fin height, fW is the fin width and fL is the fin length. The convective heat transfer coefficients for the hot and cold fluids, hh and hc, depend on the thermal-fluid properties, velocity and the flow channel geometry. The convective heat transfer coefficient of the hot fluid hh can be obtained employing the DittuseBoelter equation [15] for Re > 5000
Nuh ¼
hh DH;h 1=3 ¼ 0:023Re0:8 h Prh kh
c
(2)
(4a)
where
ReD;h ¼
rh Vh DH;h mh
(4b)
in which case the thermal-fluid properties are evaluated at bulk fluid temperatures. A similar correlation by Sleicher and Rouse [16] gives
Nuh ¼ 5 þ 0:0015Reah Prhb
(1)
1
(3a) c
where
a ¼ 0:88
where h, c, s, w and d denote hot, cold, scale, wall, and the plate thickness, respectively. The above resistances are expressed in terms of individual mean heat transfer coefficients as
d 1 1 1 1 ¼ þ þ þ þ ho hs A ðho hh At Þh ðAw kw Þh ðho hs AÞc Um A ho hc Af
where m denotes the mean value. For prime surfaces (i.e. surfaces with no fins), ho ¼ 1, the thickness of wall and the scale width are typically negligible, then the above yields
(5)
0:24 1 and b ¼ þ 0:5e0:6 Pr 4 þ Pr 3
subject to the constraints
0:1 < Prs < 100:6 and 104 < ReD < 106 where a denotes the film temperature, and b is the surface temperature. Buoyancy-driven flows are largely dependent on density differences. The convective heat transfer coefficient of the cold fluid hc based on the Nusselt number for fully developed laminar flow in a duct of aspect ratio of 0.25 is [12]
Nuc ¼
hc DH;c ¼ 4:439 kc
(6)
and for an aspect ratio of 1/8 [12],the equation gives
Nuc ¼
hc DH;c ¼ 5:597 kc
(7)
For flow conditions in which Re > 5000, the DittuseBoelter equation [13] applies to both fluids The heat duty of the cross-flow exchanger is given by [12]
Qex ¼ xCmin DT Fig. 3. The cross-flow heat exchanger plate.
(8)
where
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
P.N. Nwosu et al. / Energy xxx (2014) 1e11
x ¼
1 exp½ NTUð1 Cr Þ 1 Cr exp½ NTUð1 Cr Þ
(9)
and
DT ¼ Th;in Tc;in
(10)
Cmin Cmax
(11)
Cr ¼
The exchanger effectiveness is a function of the number of transfer units (NTU) and heat capacity ratio Cr. The number of transfer unit is a ratio of the overall conductance UA to the smaller heat transfer capacity rate Cmin, which is a fluid dimensionless residence time, or a modified station number [12]. The heat capacity rates of both fluids can be obtained from the product of the flow rate and the specific heat capacity, given by
Ci ¼ ri Vf ;i Cp;i
(12)
where subscript i denotes min or max and applies to the minimum or maximum of the heat capacity rates of both fluids. The terminal temperatures of both fluids can be predicted by [14]
Tc;out ¼ Tc;in xDTmax
(13)
Th;out ¼ Th;in xDTmax
(14)
where DTmax ¼ Th;in Tc;in . Defining the following nondimensional variables
qh;out ¼
Th;out ; Ta
qc;out ¼
Tc;out ; Ta
qc;in ¼ !
q ¼
Tc;in Ta
DTmax Ta
(15)
(16)
Eqs. (15) and (16) results to
!
qc;out ¼ qc;in x q
!
qh;out ¼ qh;in x q
DT Qex ;g ¼ ; hc Ac Ta hc Ac Ta
G ¼ Vh =Ah
where N is the number of passes, ri is the density of the fluid at inlet, f is the fanning friction factor, De is the equivalent diameter of flow channel, gc is a proportionality constant in Newton’s 2nd law, L is the flow length, ri and ro are the densities of the fluid at inlet and outlet, respectively. 3.2. Radiant heat absorber As hot gases are expelled from the cylinder, the temperature peaks at the junction of the manifold pipes and is subject to a variety drive terrain conditions [18] (Fig. 4). The distribution of the temperature contours shows that in all the cases, maximum temperatures occur at the junction of the manifold tubes, thereby influencing any design and placement considerations with respect to the radiant heat absorber. In terms of losses, a typical gasoline engine releases about 21% of the chemical energy of the fuel at the exhaust alone [19]. This increases to 44% at peak power; this is also dependent on engine size, as well as torque-speed conditions. Light duty passenger (LDP) vehicles operate at exhaust gas temperatures ranging from 500 C to 900 C, while heavy duty (HD) vehicles operate within the range of 500 C to 650 C. The range of waste heat losses is 46e120 kW in light duty 4-cylinder spark ignition engines, and 9e48 kW for the cooling system [16]. On the average, a third of the generated energy is wasted through the exhaust at most common load and speed conditions, this increase to 44% at peak power [6]. The mechanism of heat transfer from the exhaust gas to the pipewall consists of turbulent convective and radiative heat transfer. Although the range of the Reynolds number is 103e5 104 in the exhaust pipe, the flow is actually turbulent [9]. The exhaust valves and the unsteady, flow pulsation do not aid the transition to the laminar flow region. The exhaust flow often enters the region of Re < 2300, especially in the exhaust manifold runners. The heat transfer coefficient due to the turbulent flow can be modeled by NusselteReynolds correlations [9] similar to those of turbulent pipe flows
(18)
where he is the heat transfer coefficient, L is the characteristic length, ke is the gas thermal conductivity, and a is the turbulent pipe flow condition coefficient. The following correlation proposed by Gnielinksi [19] for calculating convective heat transfer from the exhaust gas to the pipe wall takes into account the effects of the wall roughness, employing a friction factor, F:
(19)
(20)
Notably, pressure drop is an important design consideration for heat exchangers. If pressure drop is considerable, flow is affected with a potential for flow stagnation. Hence, variables which can keep to the barest minimum the pressure drop performance of the exchanger are of important considerations. The correlation for pressured drop for plate heat exchangers is given by [17]
DP ¼
1:5G2 N 4fLG2 1 1 1 G2 þ þ ro ri gc 2gc ri 2gc De r me
(21a)
ð1=rÞme ¼ 1 2ð1=ri þ 1=ro Þ
Nue ¼
(21b)
(22)
F 8 ðRee
1000ÞPre qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1:07 þ 12:7 8f Pr 2=3 1
(23)
subject to the constraint
104 < Ree < 5 106 and for Re < 104
Nue ¼
F e Pre 8 Re rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1:07 þ 12:7
where
he L ¼ aðReÞ0:8 ke
Nue ¼
Combining Eqs. (10), (13) and (18) yields
! Q ¼ xCr Cmax g
(21c)
(17)
Defining also
! Q ¼
5
f 8
2=3 Pre 1
(24)
the correlation for calculating convective heat transfer coefficient for a moving vehicle (from the manifold to the ambient) is derived using the combined heat transfer coefficient [9]
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
6
P.N. Nwosu et al. / Energy xxx (2014) 1e11
Fig. 4. Contour plots of temperature distribution in an exhaust manifold in variable conditions [18].
Nue;b
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ 0:3 Nu2laminar þ Nu2turbulent
(25a)
qw ¼
2:1ln yþ þ 2:5 0 s g1 Z g kr u* r 1 0 @ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi þ 0:082 kT0 mw r0 pðs qT Þ 0 pffiffiffiffiffiffiffiffiffiffiffiffiffi!!1 s qT A dfus ð0; qT ÞdqT 1 exp 3 dqT
rCp u* TlnðT=Tw Þ 2:1yþ þ 33:4 Gn u*
subject to the constraint
10 < Re < 107 where
Nulaminar
pffiffiffiffiffiffiffiffipffiffiffiffiffi 3 ¼ 0:664 Re3 Pr
Nuturbulent
0:037Re0:8 Pr ¼ 1 þ 2443Re0:1 Pr2=3 1
where
(25b)
(25c)
Radiative heat transfer is relatively significant in the area adjoining the manifold, due to the presence of hot radiating particles in the gases. At temperatures lower than 500 C, radiation from the hot gases is relatively small and can be neglected [9]. But for temperatures exceeding that threshold, the radiation flux from the particulate matter is substantial, and can be expressed by [19]
4 qrad ¼ εpt sTpt
ð27aÞ
(26)
In diesel engines, the order of magnitude of qrad is given by qrad ¼ 0:2qtotal , where qtotal is the total cycle cumulative flux. At peak valve, qrad;max z0:4qtotal;max , where qtotal;max is the total maximum peak flux value. The order of magnitude of qrad is about approximately 1e3 MW/m2 for SI engines, and w10 MW/m2 for diesel engines, which occurs around the cylinder head [16]. The correlation proposed by Han and Reitz [20] for estimating the cylinder wall heat flux takes into consideration various operational variables and parameters:
u* ¼
qffiffiffiffiffiffiffiffiffiffiffiffi 1=2 kCm
(27b)
where u* is the dimensionless gas speed, yþ is the dimensionless distance, fus is the dimensionless unsteady temperature, s is the dimensionless time, n is the gas speed, T0 is the reference temperature, T and qT are temperatures and dimensionless temperatures, G is the source term in the energy equation, and mw is the dynamic viscosity (wall quantity). The placement of the absorber is critical to thermal performance. The radiant heat absorber is coupled to the manifold at the vicinity where the temperature is a maximum so as to maximize its heat absorption potential. The control volume is schematically idealized in Fig. 5, illustrating the mechanism of heat transfer. The impinging radiation strikes the absorber (coated with a selective coating of high absorptivity and negligible emissivity) through a glass envelope. The total heat transfer rate from the manifold to the surrounding is given by
4 Isur ¼ εm sAm Tm Ta4 þ ha Am ðTm TN Þ
(28)
where εm is the emissivity of the manifold surface, s is the Stefan– Boltzmann constant, Am is the manifold heat exchange area, ha is the convective heat transfer coefficient of ambient air, Tm is the temperature of the manifold, Ta is the ambient temperature and TN
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
P.N. Nwosu et al. / Energy xxx (2014) 1e11
is the free stream temperature. The above relation comprises convective and radiative contributions. Given that in high temperature conditions, radiative heat transfer is the dominant mode of heat transfer [21], assuming this condition and neglecting convection, the above equation yields
4 Ir;sur ¼ εm sAm Tm Ta4
(29)
Considering the above equation, the total useful heat gain by the working fluid(which is the net heat gain of the absorber)is written as
Qtot;abs ¼ saIsur Up Aabs ðTabs Ta Þ
(30)
where Up is the overall loss conductance which accounts for the heat loss components between the absorber and the glass envelope in addition to that of the glass envelope and the surrounding, Aabs is the absorptive surface area of the absorber. Up is a measure of the thermal resistance the absorber components pose to the flow of heat. Ignoring the conductive resistance of the glass, for a double glazed absorber, Up is expressed by
Up ¼
1 1 hc1
(31)
þ h1 þ h1 c2
7
xc ¼
Tc;in Tc;outabs T T 3 saεm sAm ; xp ¼ abs ; f ¼ a ; qw;in ¼ _ p Ta Ta Ta mC
xe ¼
Up Aabs saha Am Tm ;c ¼ ;g ¼ _ p _ p Ta mC mC
(35)
where f is the dimensionless radiant heat, c is the dimensionless absorber materials property, and g is the dimensionless manifold emissive area. Assuming Ta z TN, combining Eqs. (28) (30), (34) and (35), consequently
xc ¼ qw;in þ f x4e 1 þ gðxe 1Þ c xp 1
(36)
Since radiation is pre-dominant in high temperatures for this case, neglecting the convection term, the above equation becomes
xc ¼ qw;in þ f x4e 1 c xp 1
(37)
To determine the energy efficiency of the engine as a result of the coupled waste heat recovery system, comprising the preheater and the absorber, the relation which expresses the total energy absorbed by the fluid is
Etot ¼ mCp Tc;outabs Tc;in
(38)
The energy input of the fuel is given by [22]
a
The exchanger (or the preheater) coupled to the radiator extracts heat from the hot coolant, supposing there is no phase change within the preheater e this is plausible due to the finite terminal temperature of the coolant e the overall heat gain of the working fluid is the sum of the heat gain of the fluid as it flows through the preheater and the absorber, respectively given by
_ p Tc;out Tc;in Qpre ¼ mC
(32)
and
$
Efu ¼ mf $LCV
(39)
where mf is the mass flow rate of the fuel and LCV is the lower calorific value, which gives an indication of the amount of energy released on combustion. The energy efficiency of the waste heat recovery system is
hsystem ¼
Etot Efu
(40)
Consequently, the energy efficiency of the engine with the coupled waste heat recovery system is
_ p Tc;outabs Tc;in Qabs ¼ mC abs
(33)
where Tc;inabs and Tc;outabs are the inlet and outlet temperatures of the working fluid which flows through the absorber. Tc,in and Tc,out are the inlet and outlet temperatures of the working fluid which flows through the preheater. The total heat gain of the working fluid, assuming negligible path losses with Tc;out zTc;inabs is the summation of the last two energy rate equations
_ p Tc;outabs Tc;in Qtot;abs ¼ mC
(34)
Defining also the following dimensionless variables
heff ¼ hengine þ hsystem
(41)
heff is the typical engine efficiency, and assuming steady state conditions, the heat absorbed by the working fluid rapidly heats up the fluid in the tank, charging the tank. The charging rate depends on the heat input and the path losses. With increased heat input, the fluid in the tank becomes thermally stratified, and this is sometimes used as a measure of its operational performance [23]. As the fluid becomes stratified, the density gradient results to differential buoyant forces, which cause convection by gravity. The buoyancy force FB resulting from the difference in density and pressure is defined by [23]
Fig. 5. Schematic diagram of the control volume of the radiant heat absorber.
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
8
P.N. Nwosu et al. / Energy xxx (2014) 1e11
Fig. 6. Effects of nondimensional fin width on the terminal temperatures of the participating fluids for a range of qmax values.
Fig. 8. Effects of nondimensional manifold surface temperatures on the cold fluid terminal temperatures for varying c values.
FB ¼ gðrb rt Þhþ
tank, (UA)t is the overall conductance of the tank wall. Optimally, the insulated tank is mounted at least a unit foot above the absorber [23], so that buoyant forces are aided by the height differential in order to avoid flow stagnation in the piping.
(42)
where rb is the density of the fluid at the bottom of the tank, rt is the density of the fluid at the top of the tank, hþ is the vertical length between the top and bottom of the tank, and g is the gravitational force. An energy balance on a well-mixed storage tank that is rapidly charged by the absorbed heat is defined by
rUCp dTt dt
¼ Qu Ls ðUAÞt ðTt Ta Þ
(43)
where Tt is the tank temperature, Qu is the useful heat gain of the tank, Ls is the extracted load of the tank, U is the fluid volume in the
Fig. 7. Effects of nondimensional fin width on the fluid terminal temperatures for varying Cmax/Cmin values.
4. Results and discussion In Fig. 6, the thermal performance of the preheater, i.e. the exchanger, indicated by a variation in the nondimensional (terminal) temperatures of the participant fluids qi is analyzed, vis-à-vis the nondimensional fin width fw/tw and the maximum temperature
Fig. 9. Effects of nondimensional parameter f on the cold fluid terminal temperatures xc for varying xe values.
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
P.N. Nwosu et al. / Energy xxx (2014) 1e11
9
difference of the fluids qmax. As the nondimensional fin width increases, the nondimensional cold fluid terminal temperatures decrease correspondingly, indicating that flow channel dimensions are critical to thermal performance of the exchanger. With very low fW/tW values, qi increases markedly, particularly for the cold fluid. On the contrary, the hot fluid terminal temperature is lowest when fW/tW is a minimum. Inferably, as the fin width decreases the cold fluid becomes increasingly in contact with the heated plates, thus heat is transferred more rapidly in this case than in a case involving increasing fin width dimensions. Ostensibly, the terminal temperatures of both fluids are affected by changes in the fin width. The value of qi is a maximum when qmax ¼ 0.2681 for the cold fluid, and as the nondimensional fin width decreases, the terminal fluid temperatures are also affected by small differences in qmax parameter. The significance of this is that the amount of heat transferred from the hot fluid to the cold fluid increases and decreases respectively with a decrease and increase in the fin width dimensions. Fig 7 illustrates the evolution of the nondimensional terminal fluid temperatures of both fluids qi with the nondimensional fin width fW/tW in relation to the heat capacity ratios Cmax/Cmin. The nondimensional fin width gives an indication of the dimensions of the flow passage of the cold fluid in relation to the tube width tw, which is the hot fluid flow passage variable. The heat capacity ratios are essentially ratios of the specific heats of both working fluids. As the nondimensional fin width increases from a minimum value of about 2 to 28, the nondimensional (cold) fluid terminal temperature decreases, correspondingly. At a value of fW/tW ¼ 2, the value of qi for the cold fluid is about 1.22 at Cmax/Cmin ¼ 0.8794. Relatively for varying values of Cmax/Cmin, we observe a near opposite trend for the hot fluid. What can be inferred thus is that as the fin width decreases, the cold fluid is more in contact with the heated plates, and heat is thus transferred more rapidly than for increasing fin width dimensions which implies reduced contact between the plates and the fluid. For low values of Cmax/Cmin, the cold fluid terminal temperature increases irrespective of fW/tW values. Hence, cold fluids with relatively high heat capacities will extract more heat from the exchanger plates than the fluids with lower heat capacity rate for prescribed hot fluid properties. A major design
concern of any waste heat recovery system is to select working fluids which can extract the maximum (possible) amount of heat from the system while minimizing irreversibility. From the figure it can be adduced that the waste heat recovery system would have high heat extraction rates with low values of Cmax/Cmin. Notably, the cold fluid terminal temperatures drop with increasing fW/tW values for a given Cmax/Cmin value, indicating that the flow passage dimensions and the specific heat capacities of both fluids can considerably influence heat transfer in the exchanger. Regardless of changes in the Cmax/Cmin parameter, there is negligible difference in qi with high values of fW/tW. In Fig. 8, the nondimensional cold fluid terminal temperature xc is plotted against the nondimensional exhaust manifold temperature xe in relation to the variable c. c gives an indication of the relative change in the absorber parameters (the overall conductance and the absorptive area of the radiant heat absorber). As xe increases from 1 to 6, parameter xc increases non-linearly. The values of xc are a maximum at xe ¼ 6 and c ¼ 0.0, showing that increasing manifold temperatures can significantly enhance the cold fluid terminal temperatures. Conversely, the obtained values of xc are a minimum at xe ¼ 1.0 and c ¼ 2.0. Thus, if c tends to zero, there is a prospect of high terminal cold fluid temperatures since the parameter is coupled to thermal resistance within the absorber, which is based on the values of the heat transfer coefficients (of the absorber- to-glass, and glass-to-ambient). Typically low c values entail reduced convective and emissive losses from the absorber and improved heat absorption by the working fluid. But there is a limit as to values which Up can attain (hence c) in typical absorbers, owing to the finite conductivity of the materials. It is plausible to select materials and production processes which can potentially reduce parameter c values, by ensuring air tight properties of the glass envelope, as well as the use of a selective glazing with desired properties. Also, the two cases involving radiation and convection, and the lone radiation mode are analyzed. It is notable from the plot that the twin modes of radiation and convection, ‘with convection’, considerably improved heat transfer relative to the lone radiation mode, ‘without convection’. Therefore in moving vehicles, if the hot fluid stream around the manifold and the underhood, which is typically turbulent is directed toward the absorber, heat transfer will be improved. Effects of variation of xc, the nondimensional terminal fluid temperature, with f and parameter xe is shown in Fig. 9. As f
Fig. 10. Effects of tube dimensions on pressure drop.
Fig. 11. Thermal efficiency vs. flow rate.
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
10
P.N. Nwosu et al. / Energy xxx (2014) 1e11
increases from 0.005 to 0.015, xc increases correspondingly, indicating the effects of ambient temperature, the absorber material and the flow prescription on the cold fluid terminal temperatures. For increasing xe values, the dependence of xc on f is apparently strong. Clearly, there exists a linear relationship between xc and f for increasing xe values. The effects of f on xc becomes weak for low values of parameter xe. Since f is a lumped variable, high values of f for a given absorber dimension, flow prescription and ambient condition, indicate the effects of increasing absorptivity of the coating relative to emitted radiation from the manifold, thus giving an indication of the net absorbed radiation. Consequently, increasing f values entail increasing coating absorptivity values, and thus improved absorption. Absorptive coatings have been used to enhance heat transfer in energetic systems, especially in cases where heat transfer occurs mainly by radiation. Therefore for operational conditions characterized by increasing manifold and ambient temperatures in relation to increased absorptive area and high coating absorptivity values, heat transfer will improve due to the fact that high ambient temperatures potentially lower the prospect of low manifold surface temperatures thereby improving exergy. Another important design consideration is pressure drop. The power required to move the coolant across the exchanger plates is dependent on the frictional pressure drop which is often a major operating expense. The plate fin exchanger considered here has the lowest pressure drop performance, compared to other exchanger designs [25]. In Fig. 10, the pressure drop variable DP is plotted against the tube width tW e which represents the width of the (coolant) flow channel e in relation to the ratio of the tube height and length tH/tL. Importantly, these variables give an indication of the heat exchange surface dimensions. When tw increases the pressure drop decreases, for this particular case, a significant portion of the fluid mass is probably not in contact with the heated plates thus there is lower frictional drag. However for decreasing tube widths, the pressure drop increases markedly since most of the fluid mass is in contact with the exchanger plates hence leading to increased frictional drag and pressure drop. As tH/tL decreases, the pressure drop is affected as a result. Low tH/tL values give rise to increased pressure drop. Notably, the smaller the plate dimensions, the more the pressure drop due to the increased drag although heat transfer is improved as a result. Typically, a trade-off in the thermal and pressure drop functions is usually obtained by optimization [24], so as to offset any gain in one of the functions that could adversely affect the other. The inter-dependence between pressured drop, thermal performance and flow channel dimensions is critical to the overall thermal performance, hence by optimization an enhancement in the heat transfer as well as pressure drop performance would be realized. In Fig. 11, it is seen that the heat extraction efficiency heff of the waste heat recovery system decreases with increasing volumetric flow rate of the cold fluid Vc. It is assumed that the working fluid flows through the preheater and the absorber at the same flow rate, given that the fluid velocity is driven by thermosyphonic flow. As the flow rate increases, the efficiency decreases. The efficiency is about 30% at a flow rate of 0.1 m3/s, while the highest value of heff observed in the plot is about 67%, at a value of Vc ¼ 0.02 m3/s due to the fact that the heat extraction efficiency improves with flow velocity and the convective heat transfer coefficient of the working fluid, in line with Newton’s law of cooling [14]. Lower values of Vc imply a higher heat extraction potential, giving rise to increased efficiencies. Although the flow velocity is driven by the thermosyphonic flow mechanism, it is possible to influence the flow rate by introducing bends in the piping without appreciably impacting pressure drop in the flow loop. However, excessive bends in the piping would reduce the flow rate and potentially lead to increased
pressure build up. A good alternative is to optimize the efficiency on the basis of the flow velocity and pressure drop without penalizing the thermal performance. 5. Conclusion A thermosyphonic waste heat recovery system for IC engines is investigated. Heat transfer modeling of temperature and the operational variables showed that this design has the potential to improve thermal efficiency in IC engines while delivering high temperature working fluid. The system can be fabricated into a variety of sizes to fit the engine compartment, as well as conform to vehicular weight specifications. The technology is non-intrusive of the engine operations and allows for a modular design as space becomes a premium in the engine compartment: a good design of any vehicle engine waste heat recovery system must be lightweight, and adapt to the engine space provisions. It is recommended that appropriate waste heat recovery technologies which are, lightweight and non-intrusive of the engine operations be given practical considerations as a feasible solution for improved energy efficiency. Acknowledgment The efforts of Prof Risto Lahdelma of the Energy Technology Department of Aalto University are acknowledged. References [1] Tommi P, Seepo N, Pekka R. Waste heat recovery-bottoming cycle alternatives. In: Proceedings of the University of Vaasa; 2012. pp. 1e15. Reports 175. [2] Parimal PS, Doyle EF. Compounding the tuck diesel engine with an organic Rankine cycle system; 1976. SAE Paper 760343. [3] Dibella FA, Di Nanno LR, Koplow MD. Laboratory and on-highway testing of diesel organic Rankine compound long-haul vehicle; 1983. SAE Paper 830122. [4] Doyle E, Di Nanno LR, Kramer S. Installation of a dieseleorganic rankine compound engine in a class 8 truck for a single-vehicle test; 1976. SAE Paper 790646. [5] Mohsen G, Mohammad H, Davood D, Mofid G, Behravana Ali, Gholamreza S. Exergy recovery from the exhaust cooling in a DI diesel engine for BSFC reduction purposes. Energy 2014;65:44e51. [6] Teng H, Regner G, Cowland C. Waste heat recovery of heavy duty diesel engines by organic rankine cycle. Part II: working fluids for WHR-ORC; 2005. SAE 01-0543. [7] Wang T, Zhang Y, Peng Z, Shu G. A review of researches on thermal exhaust heat recovery with rankine cycle. Renew Sustain Energy Rev 2008;15:2862e71. [8] Mohammed A, Tamer S, Fatma H, Ibrahim M. Power generation using waste heat recovery by organic Rankine cycle in oil and gas sector in Egypt: a case study. Energy 2014;64:462e72. [9] Ian B, Geoffrey M, Stephen S, Roy D. Whole-vehicle modelling of exhaust energy recovery on a diesel electric hybrid bus. Energy 2014;65:172e81. [10] Kumar CR, Sonthalai A, Goel R. Experimental study on waste heat recover from an internal combustion engine using thermoelectric technology. Therm Sci 2011;15(4):1011e22. [11] Sousanis J. World vehicle population tops 1 billion units. WardsAuto; Aug. 15, 2011. http://wardsauto.com/ar/world_vehicle_population_110815. [12] Kandylas IP, Stamatelos AM. Engine exhaust system design based on heat transfer computation. Energy Convers Manage 1999;40:1055e72. [13] Vasu V, Krishna RK, Kumar ACS. Thermal design analysis of compact heat exchanger using nanofluids. Int J Nanomanufact 2008;2(3):271e88. [14] Kern F, Bohn SM. Principles of heat transfer. New York: Harper and Row; 1992. [15] Dittus FW, Boelter LMK. Heat transfer in automobile radiators of the tubular type2. University of California, Berkeley Publications in Engineering; 1930. p. 443. [16] Sleicher CA, Rouse MW. A conventional correlation for heat transfer to constant and variable property fluids in turbulent pipe flow. Int J Heat Transfer 1975;18:677e83. [17] Shah RK, Sekulic DP. Fundamentals of heat exchanger design. Hoboken, NJ: John Wiley and Sons Inc.; 2003. [18] Krenn C, Hofer G, Prosi M, Samhaber C, Pessl G. Thermal simulation of an exhaust manifold in a high performance diesel engine. In: Star European User Conference, March 17-18th, 2008, London; 2008. [19] Gnielinski V. New equations for heat and mass transfer in turbulent pipe and channel flow. Int Chem Eng 1976;16:359e68.
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091
P.N. Nwosu et al. / Energy xxx (2014) 1e11 [20] Han Z, Reitz RD. Temperature wall function formulation for variable-density turbulent flows with application to engine convective heat transfer modelling. Int J Heat Mass Transfer 1997;40(3):613e25. [21] Gray WA, Kilham JK. Heat transfer from flames. London: Elek Science; 1976. [22] Bourhis G, Leduc P. Energy and exergy balances for modern diesel and gasoline engines. Oil Gas Sci Technol Rev 2010;65(No. 1):39e46. IFP. [23] C., Currie J, Muneer T. Appl Energy 2009;86:1465e9.
11
[24] Nwosu PN. Employing exergy-optimized pin fins in the design of an absorber in a solar air heater. Energy 2010;35:571e5. [25] Mavridou S, Mavropoulos GC, Bouris D, Hountalas DT, Bergeles G. Comparative study of a diesel exhaust gas heat exchanger for truck applications with conventional and state of the art heat transfer enhancement. Appl Therm Eng 2010;30:935e47.
Please cite this article in press as: Nwosu PN, et al., Thermal modeling of a novel thermosyphonic waste heat absorption system for internal combustion engines, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.03.091