- Email: [email protected]
Analysis of natural and forced convection heat losses from a thermal manikin: Comparative assessment of the static and dynamic postures
J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
Contents lists available at ScienceDirect
Journal of Wind Engineering and Industrial Aerodynamics journal homepage: www.elsevier.com/locate/jweia
Analysis of natural and forced convection heat losses from a thermal manikin: Comparative assessment of the static and dynamic postures A. Virgílio M. Oliveira a,b,n, Adélio R. Gaspar a, Sara C. Francisco a, Divo A. Quintela a a
ADAI, LAETA, Department of Mechanical Engineering, University of Coimbra, Pólo II, 3030-788 Coimbra, Portugal Coimbra Institute of Engineering, Polytechnic Institute of Coimbra, Department of Mechanical Engineering, Rua Pedro Nunes, Quinta da Nora, 3030-199 Coimbra, Portugal
b
art ic l e i nf o
a b s t r a c t
Article history: Received 30 January 2014 Received in revised form 20 May 2014 Accepted 29 June 2014
The present experimental work is dedicated to the analysis of the effect of walking movements and air velocity on the convective heat transfer coefficients (hconv) of the human body. A wind tunnel and an articulated thermal manikin of the Pernille type with sixteen body segments were used. Beyond the standing posture (static condition), a step rate of 45 steps/min was selected, corresponding to a walking speed of 0.51 m/s (dynamic condition). The free stream air velocity was varied from 0 to about 10 m/s. The experimental conditions were thus varied from natural to forced convection. The convection coefficients for the different body segments and the whole body were determined for each air velocity giving details about the differences between them. In the case of the whole body, for the standing static reference posture and free convection, the mean value of hconv was equal to 3.5 W m 2 1C 1. In the dynamic condition the corresponding hconv value was 4.5 W m 2 1C 1. In forced convection, the highest values correspond to the highest wind speed and were equal to 22.4 W m 2 1C 1 for the static condition and 23.0 W m 2 1C 1 for the dynamic posture. & 2014 Elsevier Ltd. All rights reserved.
Keywords: Convective heat transfer coefficients Human body Thermal manikin
1. Introduction The Kyoto Protocol and all the foregoing initiatives have been underlining the need for the reduction of energy consumption related with global climate changes. Such goal has several effects on industry, transportation and built environment, among others. In a certain sense, the EU Directive 2010/31/EU on the energy performance of buildings is a recognition that further efforts are needed in order to balance between building quality, energy efficiency and comfort requirements. With this in mind, and whenever we deal with human occupancy, it seems acceptable that improved analysis of heat transfer phenomena around human body may be of interest for such matters. The present contribution is directed to this particular goal. According to the latest experimental and numerical developments characterized by an increasing need for detail, this study is focused on the experimental evaluation of convective heat transfer coefficients for the whole
n Corresponding author at: Coimbra Institute of Engineering, Polytechnic Institute of Coimbra, Department of Mechanical Engineering, Rua Pedro Nunes, Quinta da Nora, 3030-199 Coimbra, Portugal. Tel.: þ 351239790330; fax: þ 351239790331. E-mail address: [email protected] (A.V.M. Oliveira). URLS: http://www.isec.pt (A.V.M. Oliveira), http://www.dem.uc.pt (A.R. Gaspar), http://www.dem.uc.pt (S.C. Francisco), http://www.dem.uc.pt (D.A. Quintela).
http://dx.doi.org/10.1016/j.jweia.2014.06.019 0167-6105/& 2014 Elsevier Ltd. All rights reserved.
human body and the body parts. Discussion about the convective trends in natural, mixed and forced regimes is also foreseen. When heat transfer coefficients are considered, it must be recognized that most data available in the literature refer to the whole body and some of the first studies took place in the thirties by Büttner (1934) and Gagge (1937). Later on, the whole body natural convection heat transfer coefficients were characterized as equal to 5.1 W m 2 1C 1 by Colin and Houdas (1967), 4.0 W m 2 1C 1 by Seppänen et al. (1972) and 3.1 W m 2 1C 1 by Mitchell (1974). Afterwards, de Dear et al. (1997) obtained values of 3.4 and 3.3 W m 2 1C 1 for a standing and sitting thermal manikin, respectively, and Omori et al. (2004) proposed a value of 3.3 W m 2 1C 1 for the standing posture. With the development of thermal manikins, the heat losses from each part of the body have also been addressed and a set of data is now available. Oguro et al. (2002a) and Quintela et al. (2004) proposed expressions to calculate hconv in natural convection as a function of the difference between the skin and the air temperatures ðT sk T a Þ. While Oguro et al. (2002a) considered a manikin dressed and nude for both standing and sitting postures, in the work of Quintela et al. (2004) the experimental determination of convective and radiative heat transfer coefficients in standing, sitting and lying postures were considered. Nilsson (2004) assessed the repeatability of heat transfer coefficients by considering five independent calibrations performed from
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
1995 to 1998 and the results showed that the mean standard deviation was 8% for all zones. Kurazumi et al. (2008) have also investigated the convective and radiative heat transfer coefficients, focusing their research on the convective heat transfer area of the human body. Under natural convection conditions, the mannequin was placed in seven postures and the radiative heat transfer coefficient was determined for each posture and empirical formulas were proposed for the convective heat transfer coefficient of the entire human body, driven by the difference between the air and mean skin temperatures. The effect of body motion was first considered in the work of Chang and Gonzalez (1989) using the sublimation technique with circular naphthalene disks affixed to five different segments of one articulated manikin. The static posture and four walking gaits (20, 40, 68 and 80 steps/min) were studied. The results for the heat transfer coefficient by convection have an intrinsic local nature and should not be compared with more recent thermal manikins where such coefficients are averaged through the surface of different parts. More recently, the effect of walking on the convective heat transfer coefficients was considered by Oliveira et al. (2012). Beyond the standing static posture, three step rates were considered (20, 30 and 45 steps/min) and the convective heat transfer coefficients for the whole body and for the different body segments were determined for each step rate. In the case of the whole body, for the standing static reference posture, a correlation for natural convection was also presented. In the case of forced convection, a relevant study was developed by de Dear et al. (1997) through a wide range of wind speeds, varying from still air to 5.0 m/s. Both standing and sitting postures were investigated, as there were eight different azimuth angles. Another interesting work was carried out by Silva and Coelho (2002), considering three flow incidences (front, back and side) and two postures (seated and standing), and Oguro et al. (2002b) considered the same postures with the manikin nude and clothed facing upwind and downstream flows. Ono et al. (2008) measured the convective heat transfer coefficient for the human body in an outdoor environment by means of a thermal manikin placed in a wind tunnel complemented with a computational fluid dynamics (CFD) analysis. Different wind velocities were considered and the mean convective heat transfer coefficients for the whole body were 6.6, 10.0 and 16.3 W m 2 1C 1, for 0.5, 1.0 and 2.0 m/s, respectively. In a different perspective, Defraeye et al. (2011) investigated the convective heat transfer for cyclists. The convective heat transfer coefficients of the whole body and of the 19 body parts of a cyclist were determined. Wind speed correlations based on power-law equations were proposed and the results have shown that the power-law exponent does not differ significantly for the individual body segments for all positions tested. More recently, Luo et al. (2014) considered five moving speeds of a manikin (0.2, 0.5, 0.8, 1.1 and 1.3 m/s) carried out with a motor on a 10 m-length-rail, and four temperature differences (4, 8, 12 and 16 1C) by changing the heating power of the manikin. They concluded that to explain the details of the differences between the moving condition and the wind tunnel condition, which may be caused by the flow field, further analysis of the flow field should be foreseen. In a complementary perspective, extensive research has been done in the assessment of the thermal insulation of clothing, a field directly related to heat transfer by convection and radiation. When the analysis of human typical postures is considered and restricting the focus to experimental conditions close to the ones of the present study, i.e., considering the walking effect and high air velocity, the studies of Anttonen (2000); Nilsson et al. (2000); Havenith and Nilsson (2004); Anttonen et al. (2004) and Oliveira et al. (2008, 2011) are relevant contributions. In the present work the convective heat transfer coefficients are addressed in a complementary perspective with novel approaches.
67
Body motion is taken into account – a rather seldom feature considered in previous studies besides the work of Chang and Gonzalez (1989) – in both natural and forced convection, but the actual methodology as well as the thermal manikin differs significantly from the available in the eighties and followed by Chang and Gonzalez (1989). In addition, a wider range of air velocity is considered, up to 10 m/s. Such air velocity values are also not easy to find in other studies, and this distinct characteristic is presented in this paper. Several combinations of tests were performed and comparisons between static (i.e. standing still) and dynamic (i.e. walking at 45 steps/min) conditions are made. The main purpose is to measure convective heat transfer coefficients for all the conditions tested with the manikin operating under the thermal comfort regulation mode, a condition that represents another different approach and also not often addressed, besides the work of Oguro et al. (2002a) and Oliveira et al. (2012).
2. Methods 2.1. Heat transfer coefficients The exchange of heat between a solid and the environment can occur by convection (C), radiation (R) and conduction (K). In the case of the human body, the heat transfer by conduction is limited to the body parts in contact with solid surfaces, which are generally restricted to a few, namely to the feet, and thus usually neglected. Therefore, the heat is transferred mainly by convection and radiation. The study of convective heat transfer between the human body and the environment represents a great challenge. In fact, the heat transfer by convection depends on various factors and exact analytical solutions are actually not available. Hence, the heat transfer by convection is generally calculated making use of Newton's law of cooling (Fanger, 1972) C ¼ hconv ðT s T a Þ
ð1Þ
where hconv is the convective heat transfer coefficient [W m 2 1C 1] and T s and T a represent the mean temperatures of the body surface and the environment [1C]. For the dressed human body, the heat transfer by convection is calculated in an analogous way from (Fanger, 1972; Parsons, 2003) C ¼ f cl hconv ðT cl T a Þ
ð2Þ
where f cl is the clothing area factor and T cl is the mean clothing surface temperature [1C]. In the case of radiation, the heat transfer for the dressed human body is calculated by an expression similar to Newton's law (Fanger, 1972; Parsons, 2003) R ¼ f cl hrad ðT cl T r Þ
ð3Þ 2
1
where hrad is the radiative heat transfer coefficient [W m 1C ] and T r is the mean radiant temperature [1C]. Therefore, in studies concerning the human body heat exchanges with the environment, the overall sensible heat released by convection and radiation, Q_ s [W m 2], can be calculated from the following equation: Q_ s ¼ f cl hconv ðT cl T a Þ þ f cl hrad ðT cl T r Þ
ð4Þ
for the case of a clothed body (Fanger, 1972; Parsons, 2003), or from Q_ s ¼ hconv ðT sk T a Þ þ hrad ðT sk T r Þ
ð5Þ
for the naked human body (Fanger, 1972; Parsons, 2003), where T sk is the mean skin temperature [1C].
68
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
As a first approach, the radiative heat transfer coefficient for the whole body, hrad [W m 2 1C 1], can be calculated from the expression (ASHRAE, 2001)
equations with the following general form:
" #3 Arad T cl þ T r 273:2 þ hrad ¼ 4 σ ε ADuBois 2
Tables 2 and 3 list values and correlations proposed by several authors for natural and forced convection, respectively. The analysis of such values and correlations must be done with caution since the information about the experimental procedures in which the values and equations are based is rarely available. For instance, data about posture, environmental conditions of the tests and air velocity limits play an important role in the differences that are found and so conclusive comparisons are difficult to achieve.
ð6Þ
where σ is the Stephan–Boltzman constant (5.67 10 8 W m 2 K 4), ε is the average body surface emissivity, Arad is the effective radiation area of the human body, ADuBois is the Du Bois and Du Bois (1916) body surface area. The emissivity of the body can be assumed as 0.95, as already adopted by other manikin owners (de Dear et al., 1997; Quintela et al., 2004; Fukazawa et al., 2009), and the effective radiation area factor Arad/ADuBois for the standing posture is generally accepted as 0.73 (Fanger, 1972). The literature provides hrad values for the whole body and for the different body sections obtained with thermal manikins (de Dear et al., 1997; Ichihara et al., 1995). Nevertheless, preference was given to results obtained during tests carried out by Quintela et al. (2004) with the same manikin as used in this study (see Table 1). In the present study the evaluation of the convective heat transfer coefficient for the case of a nude human body is one of the main goals. For this reason, Eq. (5) can be expressed in terms of hconv by hconv ¼
Q_ s hrad ðT sk T r Þ
ð7Þ
ðT sk T a Þ
If attention is now given to the body parts, the convective heat transfer coefficient for each part can be computed from the following: hconv;i ¼
Q_ s;i hrad;i ðT sk;i T r Þ ðT sk;i T a Þ
ð8Þ
where the subscript i (i¼1,…,16) stands for individual reference. The analysis of both natural and forced convective heat transfer coefficients is considered in the literature by different approaches. In the case of natural convection, single values or correlations based on the difference between the clothing/skin and the air/ operative temperatures are the most common. On the other hand, for forced convection, the literature usually suggests regression
hconv ¼ C vna
ð9Þ
3. Measurement procedures 3.1. Wind tunnel The measurements were carried out in a wind tunnel at the Industrial Aerodynamics Laboratory. The wind tunnel is a closed circuit with an opened squared test chamber (2 2 m2) and 5 m long where the free stream air velocity can be varied from 0 to 18 m s 1 (Ferreira and Oliveira, 2009). Fig. 1a shows part of the test chamber and the downstream opening for flow recirculation and Fig. 1b shows the opening of the incoming air. The mean air velocities analysed in the present work ranged from 0 to 10 m s 1 (0 and approximately 0.25, 0.5, 0.75, 1.0, 2.5, 5 and 10 m s 1), thus gathering both indoors and outdoors conditions. 3.2. Thermal manikin A Pernille type thermal manikin named “Maria” (Fig. 1) was placed at the centre of the test chamber. The manikin is articulated and divided into 16 parts independently controlled by a computer according to the following relation between dry heat losses and skin temperature of the human body under thermal comfort conditions (Fanger, 1972; Madsen, 1976): T sk ¼ 36:4 0:054 Q_ s
ð13Þ
The manikin is made of a fibreglass armed polyester shell covered with a thin nickel wire wound around all the body to ensure heating and temperature measurement. It has joints at the
Table 1 Values of hrad [W m 2 1C 1] proposed by Ichihara et al. (1995), de Dear et al. (1997) and Quintela et al. (2004). Mannequin section
1 – Left foot 2 – Right foot 3 – Left leg 4 – Right leg 5 – Left thigh 6 – Right thigh 7 – Pelvic region 8 – Head 9 – Left hand 10 – Right hand 11 – Left arm 12 – Right arm 13 – Left upper arm 14 – Right upper arm 15 – Chest 16 – Back Total
Ichihara et al. (1995) (ε ¼ 0.96)
de Dear et al. (1997)
Quintela et al. (2004)
(ε ¼ 0.95)
(ε ¼ 0.95)
Standing
Sitting
Standing
Sitting
Standing
Lying
7.3
4.2 4.2 5.3 5.3 4.3 4.3 4.2 4.1 4.1 4.1 4.9 4.9 5.2 5.2 4.5 4.4 4.5
3.9 3.9 5.3 5.3 4.3 4.3 4.2 4.1 4.1 4.1 4.9 4.9 5.2 5.2 4.5 4.4 4.5
4.7 4.9 4.7 4.6 4.6 4.2 4.4 5.7 4 4.6 4 4.4 4.7 4.4 3.5 3.9 4.4
5 5 5 5.1 4.8 4.4 4.5 5.7 4.3 4.1 4.5 4.1 4.9 4.2 4.5 4.2 4.6
5.2 5.2 4.2 4.7 4.5 4 3.9 5.6 4.3 4.1 4.1 3.9 4.8 4.1 4 3.5 4.3
4.8 4.2 3.9 4.3 3.7 3.9 4.0 3.8 3.6
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
69
Table 2 Correlations and values of hconv for natural convection. hconv [W m 2 1C 1]
va [m s 1]
Posture conditions
hconv ¼ 2:38 ðt cl t a Þ0:25
va o 0.15
Sitting and standing
5.1 4.0 3.1
va o 0.15 va o 0.15 va o 0.2 65o M o 175M [W m 2] va o 0.1 va o 0.1 va o 0.1
Reclining position Standing Sitting Sedentary office work Sitting Standing Sitting
hconv ¼ 2:48 ðt sk t o Þ0:18
va o 0.1 va o 0.1 va o 0.15
Standing Standing Lying
hconv ¼ 1:94 ðt sk t o Þ0:23
va o 0.15
Sitting
hconv ¼ 2:02 ðt sk t o Þ0:24
va o 0.15
Standing
hconv ¼ 1:007 ðt sk t a Þ0:406
va o 0.2
hconv ¼ 1:16 ðM 50Þ0:39 3.3 3.4 hconv ¼ 0:78 ðt sk t o Þ
0:59
hconv ¼ 1:21 ðt sk t o Þ0:43 3.3
Reference Nielsen and Pedersen (1953) adopted by Fanger (1972) Colin and Houdas (1967) Seppänen et al. (1972) Mitchell (1974) Gagge et al. (1976) de Dear et al. (1997) Oguro et al. (2002a) Omori et al. (2004) Quintela et al. (2004)
Standing. exposed to atmosphere
Kurazumi et al. (2008)
Standing. floor contact Chair sitting. exposed to atmosphere
hconv ¼ 1:183 ðt sk t a Þ0:347 hconv ¼ 1:175 ðt sk t a Þ0:351
Chair sitting. contact with seat. Chair back and floor
hconv ¼ 1:222 ðt sk t a Þ0:299
Cross-legged sitting. floor contact
hconv ¼ 1:271 ðt sk t a Þ0:355 hconv ¼ 1:002 ðt sk t a Þ0:409
Legs-out sitting. floor contact
hconv ¼ 0:881 ðt sk t a Þ0:368
Supine. Floor contact
Table 3 Correlations of hconv for forced convection. hconv [W m 2 1C 1]
Reference
hconv [W m 2 1C 1]
Reference Mochida et al. (1999)
hconv ¼ 6:3 v0:5 a
Büttner (1934)
Seated manikin hconv ¼ 2:82 þ 6:35 v0:848 a
hconv ¼ 12:13 v0:5 a
Winslow et al. (1939)
Standing manikin hconv ¼ 10:14 v0:431 a
hconv ¼ 2:67 þ 8:72 v0:67 a
Colin and Houdas (1967)
Seated manikin hconv ¼ 10:1 v0:61 a
hconv ¼ 8:6 v0:531 a
Gagge et al. (1969)
Standing manikin hconv ¼ 10:4 v0:56 a
hconv ¼ 6:51 v0:391 a pffiffiffiffiffi hconv ¼ 8:3 va
Nishi and Gagge (1970)
Both seated and standing postures hconv ¼ 10:3 v0:6 a
Kerslake (1972)
hconv ¼ 14:8 v0:69 a hconv ¼ 3:5 þ 5:2 va with va r 1 m=s
Seppänen et al. (1972)
Nude and clothed manikin hconv ¼ 3:36 þ 6:86 v0:92 a hconv ¼ 3:9 va þ 3:6 (n) Standing manikin – Front flow
Missenard (1973)
hconv ¼ 4:0 va þ 3:6 (n) Standing manikin – Side flow
hconv ¼ 8:7 v0:6 a with va 4 1 m=s
Missenard (1973)
hconv ¼ 4:2 va þ 3:5 (n) Standing manikin – Back flow
hconv ¼ 8:3 v0:6 a
Mitchell (1974)
(n) Nude manikin, Upstream flow hconv ¼ 9:31 v0:6 a
hconv ¼ 5:7 ðM 0:85Þ0:39 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi hconv ¼ 3 270 v2a þ 23
Gagge et al. (1976)
(n) Nude manikin, Downstream flow hconv ¼ 9:41 v0:61 a
Mochida (1977)
(n) Clothed manikin, Upstream flow hconv ¼ 13:36 v0:6 a
hconv ¼ 15:4 v0:63 Standing manikin a
Ichihara et al. (1995)
hconv ¼ 12:38 v0:65 (n) Clothed manikin, Downstream flow a
de Dear et al. (1997)
Kuwabara et al. (2001) Silva and Coelho (2002)
Oguro et al. (2002b)
(n) – Standing.
shoulders, hips and knees. They are all made of a circular cut in such a way that standing as well as sitting postures are quite natural. 3.3. Walking device A mechanism with the capability to simulate realistic walking movements was developed to reproduce the effect of body motion. The supporting frame of the manikin has wheels with brakes in order to facilitate the transportation. The manikin is hung from the head and the feet are kept 0.15 m away from the floor. The other necessary fixing point is achieved with a waist belt. The power transmission lies on a pneumatic system with a unique cylinder, which applies a swinging movement to legs and arms by fixing supports to wrists and knees. This action can be
controlled in order to simulate different step rates. The only movable parts are then the arms and legs. To provide control of the walking movement the mechanism has an electro valve and four autonomous flow control valves. The cylinder is connected to a central flat bar that distributes the movement to each limb by spherical connections. The lower and upper limbs are then connected to the same central bar ensuring that they all move at the same speed. The cylinder, the electro valves and the flow control valves are all from Festo and the spherical connections are type Pilloballs from IKO Nippon Thompson. The mechanical components were kept to a minimum in order to make the apparatus as simple and reliable as possible. The design and construction of the walking device was made at the Industrial Aerodynamics Laboratory.
70
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
of the manikin. The mean value of the last ten minutes of each test was considered for reference. Finally, the environmental parameters in the Laboratory, namely the air temperature (ta) and the relative humidity (rh), were monitored with the data logger Testo 175-H2 (refª 0563 1758) positioned away from the test area. 3.5. Test conditions The mean skin temperature ðt sk Þ and the heat flux ðQ_ s Þ of the manikin were continuously monitored, enabling an easy check of the steady state situation. Data acquisition was started after achieving this condition and lasted for 20 min, from which just the last 10 min were considered for analysis. The hconv calculation is thus performed based on t sk and Q_ s averages of these last 10 min. During the acquisition period, the heat flux and the skin temperature of each body part were recorded every minute. Two days were necessary to perform the sixteen experiments that began with the static posture. For this posture, the air temperature ranged between 12.9 and 13.7 1C and the difference between the air and operative temperatures did not exceed 0.21C. For the dynamic condition ta varied between 13.0 and 13.8 1C and the highest difference between ta and to was also 0.2 1C. The va STD values were calculated for both static and dynamic conditions and the higher value of each test was equal to 0.0381 and 0.0484 m s 1, respectively. The higher vertical temperature differences usually occur between the head and ankle levels, with ta and tg differences less than 0.5 1C. The radiant asymmetry never exceeded 0.2 1C.
4. Results and discussion
Fig. 1. Thermal manikin in the wind tunnel.
Two main sections are considered next for analysis, one focusing the whole body and the other one dealing with the local results of each body part. The discussions concerning the body parts are supported by two different approaches, pointing out their own differences and showing the variation of the convective heat transfer coefficients with air velocity.
3.4. Physical parameters 4.1. Convective heat transfer coefficients – whole body The physical parameters of the environment were measured according to ISO, 7726 (1998) requirements and using equipment from Brüel & Kjær, namely the heat stress monitor, type 1219, the thermal comfort metre, type 1213, and the indoor climate analyser, type 1212, and from Testo, namely the hot sphere sensor (refª 0635 1049) and the data loggers 445 (refª 0560 4450) and 175-H2 (refª 0563 1758). The vertical temperature difference was controlled with the heat stress monitor, type 1219, by measuring the air (ta) and globe (tg) temperatures at three heights corresponding to the head, abdomen and ankle levels of the manikin. The air (ta) and globe (tg) temperatures also support the estimation of the mean radiant temperature ðt r Þ according to the expression suggested in ISO 7726 (1998) for natural convection. The radiant asymmetry was measured with a plane radiant temperature sensor (refª MM0036) connected to the thermal comfort metre, type 1213, and the operative temperature (to) was measure with a to sensor (refª MM0023) connected to the indoor climate analyser, type 1212. The air velocity (va) was measured with a frequency of 1/10 Hz and with a hot sphere sensor Testo (refª 0635 1049) connected to the data logger Testo 445 (refª 0560 4450). In order to avoid any flow disturbance that might affect the manikin, the sensor was positioned upstream, just after the entry section of the wind tunnel test chamber, approximately at the height of the abdomen
In free convection, the estimated convective heat transfer coefficient for the standing static condition was 3.5 W m 2 1C 1. This value agrees with those found in the literature and is particularly close to the one proposed by de Dear et al. (1997) (3.4 W m 2 1C 1). In the dynamic condition the calculated hconv value was 4.5 W m 2 1C 1. This test with the manikin walking on the spot does not correspond to a pure natural convection situation, even though the experiment without an air flow was analysed together with the natural convection test. It must be underlined that this result depends on the way how motion is created. As the manikin is walking on the spot there is no changing of position in space so there is no convection from displacement. Hence, convection is due to the balancing of the body caused by walking, i.e., by arms and legs movements since the torso and the head are practically still. These areas (head, chest and back) represent almost 31% of the body surface area and approximately 42% if the pelvic region is also considered, which means that the whole body result is affected by those body parts. The experimental conditions that were used are thus more close to gym activities like walking on a treadmill. Regarding the forced convection results, the highest values correspond to the highest wind speed and were equal to 22.4 and 23.0 W m 2 1C 1, respectively for the static and dynamic
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
Table 4 Convective heat transfer coefficients and standard deviations. Air velocity [m s 1]
0 0.22 0.51 0.75 0.96 2.48 4.88 9.43
Table 5 Convective heat transfer coefficients for the 16 body parts of the thermal manikin.
Static tests (standing still)
Dynamic tests (45 steps/min)
Schematic
Manikin
hconv [W m 2 1C 1]
Convective heat transfer coefficient hconv [W m 2 1C 1]
Standard deviation STD [W m 2 1C 1]
Convective heat Standard transfer deviation STD coefficient hconv [W m 2 1C 1] 2 1 [W m 1C ]
representation
Body part
Static Dynamic de Dear et al. (1997)
3.5 3.5 5.7 6.2 6.7 11.3 16.4 22.4
0.035 0.037 0.009 0.022 0.005 0.052 0.053 0.069
4.5 4.8 6.0 6.7 7.4 11.4 17.0 23.0
0.015 0.08 0.038 0.029 0.004 0.058 0.782 0.324
8
14 12 10 6
Whole body
4
25
hconv = 8.17va0.43
2
hconv [W/m .ºC]
20
71
2
1 – Left foot 2 – Right foot 3 – Left leg 4 – Right leg 15 13 5 – Left thigh 6 – Right thigh 7 – Pelvic region 7 11 8 – Head 9 – Left hand 10 – Right hand 9 11 – Left arm 12 – Right arm 5 13 – Left upper arm 14 – Right upper 3 arm 15 – Chest 1 16 – Back Whole body
4.4 4.4 3.4 3.9 3.9 3.9 2.8 4.3 5.3 4.1 3.7 3.6 3.2
7.3 7.3 6.8 6.6 4.6 4.7 3.2 4.2 7.2 6.9 4.5 5.1 3.4
5.1 5.1 4.1 4.1 4.1 4.1 3.4 3.6 4.1 4.1 3.7 3.7 2.9
3.4
4.0
2.9
2.3 2.8 3.5
2.7 3.1 4.5
3.0 2.9 3.4
15
10
hconv=
7.34va0.49
5
Static
Dynamic
0 0
2
4
6
8
10
va [m/s]
Fig. 2. Convective heat transfer coefficients – whole body.
conditions. Table 4 presents the averages of hconv calculations and the respective standard deviations (STD). The highest STD value is 0.069 W m 2 1C 1 for the static tests and 0.782 W m 2 1C 1 for the dynamic experiments. Table 4 shows that, generally, the STD values are higher for the dynamic condition and for the highest wind speeds, fact that can be explained by a more difficult achievement of stable conditions in the thermal manikin. Fig. 2 refers to the present results and shows that the dynamic values are always higher than the static ones. The differences between the two conditions are higher for the lower air velocities, and decrease with the air velocity. For the higher air velocities, the results even become almost equal. The data from both trials also put in evidence the air velocity dependence of hconv. The static and dynamic power regression equations presented in Fig. 2 should not be used for air velocities higher than 6 m/s due to the relative position exchange of the corresponding curves above that limit. If we compare the present results based on the equation proposed for the static condition with others of the literature, one can conclude that the present results are both over and underestimated. In fact, this was expected and also sustains the validity of the present results. Such differences have to be acknowledged when simple comparisons are performed. For instance, if the values of hconv are calculated making use of the equations proposed by Büttner (1934); Gagge et al. (1969); Nishi and Gagge (1970); Kerslake (1972); Mitchell (1974); Ichihara et al. (1995) and de Dear et al. (1997), for air velocities up to 5 m s 1 and then compared, we can conclude that the present results are higher than those proposed by Büttner (1934) and Nishi and Gagge (1970), and lower than the others, despite being close to the results of Gagge et al. (1969) and Kerslake (1972). However, if the results of de Dear et al. (1997) are considered for reference, we
conclude that de Dear's results are higher than all the others, except for the case of Ichihara et al. (1995). In this last case, for an air velocity of 5 m s 1, the value of de Dear is 25.61 W m 2 1C 1, while the one proposed by Ichihara is 42.45 W m 2 1C 1. If we look at the mean relative differences up to 1 m s 1, the conclusion is that the present results overestimate the ones obtained by Büttner (1934) and Nishi and Gagge (1970) by 14.7 and 5.8%, respectively. On the contrary, the present results underestimate the ones obtained by Gagge et al. (1969); Kerslake (1972) and Mitchell (1974) by 12.5, 11.0 and 5.5%, respectively, with higher mean relative differences obtained for Ichihara et al. (1995) (48.1%) and de Dear et al. (1997) (26.4%). Differences of the same order can be found if the comparison is now applied between these authors. For such differences, the range of variation obtained in repeatability tests of heat transfer coefficients must be considered, as clearly demonstrated by Nilsson (2004). In reproducibility trials higher values are expected, so the mean relative differences above are not as significant as a simple analysis of the results might suggest. Furthermore, all of these results are considered valid as far as the conditions on which they were derived are fulfilled. Many reasons contribute to these discrepancies, namely the ones related with the methodology and with the features of the experimental devices. Some may be difficult to assess, like the characteristics of the environment created by climate chambers and wind tunnels, while others are easier to clarify. For instance, a thermal manikin with a higher body surface area will, under the same experimental conditions, naturally lead to lower heat transfer coefficient when compared to one with a smaller area. Despite all the other different experimental procedures this is the case, for example, of the manikin used in the present study (ADuBois ¼ 1.642 m2) when compared to the one used in de Dear's tests (ADuBois ¼1.471 m2). For the present results, the differences between the static and dynamic values decrease with the air velocity and a mean relative difference of about 10% was obtained between the static and dynamic results. In addition, the value of the exponent in both equations is close to 0.5, namely for the static condition, which is the classic value suggested in the literature to represent the convective phenomena. Table 3 lists some representative correlations proposed in the literature for the case of humanoid shapes and shows that the air velocity exponent varies between 0.391 and 1.0, with the majority of the values more close to 0.6.
72
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
hconv [ W/m 2.ºC ]
Fan speed 0 r.p.m. va=0 m/s
8 head 14 right upper arm 15 12 right arm
13 left upper arm 11 left arm
10
10 right hand
Fan speed 10 r.p.m. va=0.22 m/s
9 left hand
5
16 back
4 right leg
8 head 14 right upper…15
12 right arm
16 back
5 left thigh
6 right thigh
Fan speed 17 r.p.m. va=0.51 m/s
hconv [ W/m 2.ºC ]
12 right arm 10 right hand
11 left arm 9 left hand
5
16 back
0
6 right thigh 4 right leg
12 right arm
12 right arm
16 back
5 left thigh
6 right thigh
10 right hand 16 back
11 left arm 9 left hand
6 right thigh 4 right leg
Fan speed 151 r.p.m. va=4.88 m/s
10 right hand 16 back
5 left thigh
6 right thigh
16 back
0
6 right thigh 4 right leg
11 left arm 9 left hand
9 left hand
15
15 chest 5 left thigh 3 left leg
2 right foot 1 left foot 7 pelvic region Static Dynamic
hconv [ W/m 2.ºC ] 8 head 14 right upper…45
12 right arm 10 right hand 16 back
5 left thigh
6 right thigh
3 left leg
11 left arm
30
0
15 chest
2 right foot 1 left foot 7 pelvic region Static Dynamic
13 left upper arm
4 right leg
Fan speed 301 r.p.m. va=9.43 m/s
13 left upper arm
Dynamic
8 head 14 right upper…45
Dynamic
15
3 left leg
hconv [ W/m 2.ºC ]
15 chest
hconv [ W/m 2.ºC ]
10 right hand
5 left thigh
12 right arm
3 left leg
8 head 14 right upper 45 arm 12 right arm 30
15 chest
0
Static
2 right foot 1 left foot 7 pelvic region Static
9 left hand
5
2 right foot 1 left foot 7 pelvic region
Fan speed 77 r.p.m. va=2.48 m/s
0
11 left arm
4 right leg
13 left upper arm
5
13 left upper arm
10
10 right hand
Dynamic
10
Dynamic
8 head 14 right upper arm 15
3 left leg
8 head 14 right upper arm 15
3 left leg
hconv [ W/m 2.ºC ]
15 chest
hconv [ W/m 2.ºC ]
Fan speed 31 r.p.m. va=0.96 m/s
5 left thigh
Static
2 right foot 1 left foot 7 pelvic region Static
15 chest
0
Fan speed 25 r.p.m. va=0.75 m/s
13 left upper arm
10
9 left hand
5
2 right foot 1 left foot 7 pelvic region
Dynamic
8 head 14 right upper…15
11 left arm
4 right leg
3 left leg
2 right foot 1 left foot 7 pelvic region Static
13 left upper arm
10
10 right hand
15 chest
0
6 right thigh
hconv [ W/m 2.ºC ]
30 15 0
4 right leg
13 left upper arm 11 left arm 9 left hand 15 chest 5 left thigh 3 left leg
2 right foot 1 left foot 7 pelvic region Static Dynamic
Fig. 3. Convective heat transfer coefficients obtained in the static and dynamic tests with the manikin facing the incoming wind.
4.2. Convective heat transfer coefficients – body parts The convective heat transfer coefficients obtained in the tests without wind (va ¼0 m s 1) are listed in Table 5, together with the
results obtained by de Dear et al. (1997). The highest convective heat transfer coefficients are associated to the peripheral parts of the body and the lowest to the central body parts. In the static condition the highest value corresponds to the left hand (5.3 W m 2 1C 1)
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
and the lowest to the chest (2.3 W m 2 1C 1). In the dynamic condition the feet present the highest values (7.3 W m 2 1C 1) closely followed by the left hand (7.2 W m 2 1C 1). The lowest value (2.7 W m 2 1C 1) was also obtained in the chest. It is also important to point out the asymmetry of the hands, especially in the static tests. While the right hand is very close to the thigh, the left hand is opened and set apart from the thigh, thus probably outside the boundary layer developing from the feet. This detail is certainly the principal reason for the noticed differences between the left and right hands. In fact, the results show that the highest difference between the present results and those obtained by de Dear et al. (1997) corresponds to the left hand (1.2 W m 2 1C 1), whereas the right hands have the same value on both studies (4.1 W m 2 1C 1). The difference in the head is also significant since de Dear obtained an hconv value of 3.6 W m 2 1C 1 while in the present study the result was 4.3 W m 2 1C 1. In fact, the manikin used in de Dear's tests had a shoulder-length hair which can be expected to reduce dry heat losses from the head/neck region (de Dear et al., 1997). In contrast, the manikin used in the present tests had the head unprotected, i. e. without wig, thus more favourable to heat losses. It must be emphasized that details related to the experimental procedures can play an important role whenever comparisons are foreseen. For instance, the working section of the facility where de Dear's tests took place did not allow the thermal manikin to stand fully erect and the experiments were performed by slightly bending the manikin's legs at the knees. To enable both upper and lower leg segments to be measured in the vertical position, the experiments were repeated twice (de Dear et al., 1997). A remark is also due to the control methods of the manikins. In de Dear's tests the operation of the manikin is based on a constant
73
skin temperature control and in the present study the 16 parts of the thermal manikin are independently controlled according to the relation suggested by Fanger for dry heat losses within the PMV model (Fanger, 1972; ISO 7730:2005). Despite the mentioned differences, if the static values are depicted one by one, it is possible to find some similarities in the results. The graphics of Fig. 3 summarize all the tests, by presenting the hconv values of the 16 body parts for both static and dynamic conditions. It is important to point out that two different scales were adopted to make clear the differences that occur with lower velocities. Thus, the scale in the graphs up to 31 r.p.m. ( ffi 1 m s 1) is limited to 15 W m 2 1C 1 and the others to 45 W m 2 1C 1. This representation puts in evidence the differences between the 16 parts. For the central body parts the results are approximately equal, which means that the walking movements have a minor influence on these body parts. Otherwise, the dynamic experiments incorporate the pendulum effect into the hconv values and confirm the influence of walking movements on the limbs. The dynamic effect is enhanced for the lower air velocities and attenuated for the higher. Another interesting remark is related to the magnitude of the hconv values. In the experiments carried out up to 1 m/s, the hconv values are always lower than 15 W m 2 1C 1, but for the test at the higher air speed that limit is generally fixed at 30 W m 2 1C 1. The exceptions refer to the hands, where that limit is overcome. The static and dynamic results for the central body parts (chest, back and pelvic region) are almost the same which indicates that the walking movements at 45 steps/min have a negligible effect on hconv. Furthermore, the pelvic region, the chest and the back have the smallest values of hconv. Above 1 m/s the influence of the body
15 - Chest 50
40
40
hconv [W/m2.ºC]
hconv [W/m2.ºC]
8 - Head 50
30
hconv = 8.12va
20
0.49
10
Static
30
20
hconv = 6.10va0.52
10
Dynamic
0 0
2
4
6
8
0
2
4
7 - Pelvic region
8
10
16 - Back 50
40
40
hconv [W/m2.ºC]
hconv [W/m2.ºC]
6 va [m/s]
50
30
hconv = 5.99va0.50
10
Static
hconv = 5.43va
Dynamic
0
10
va [m/s]
20
Static
hconv = 5.90va0.55
hconv = 7.72va0.50
30
20
hconv = 5.66va0.55 10
Dynamic
Static
hconv = 5.33va0.59
0.54
0
Dynamic
0 0
2
4
6
8
va [m/s]
Fig. 4. Convective heat transfer coefficients – head and pelvic region.
10
0
2
4
6
8
va [m/s]
Fig. 5. Convective heat transfer coefficients – trunk.
10
74
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
movements on the convective heat transfer phenomena is not noticed. The peripheral parts of the body have larger convective coefficients. In fact, the points at a longer radius from the centre cover longer distances, and thus, have stronger heat losses. This is a common sense feeling which is clearly demonstrated here by measurements. An overview of the upper limbs puts in evidence the unexpected behaviour of the left hand. The observed asymmetry is due to the already referred position more favourable to heat losses. The highest values of hconv were obtained with the highest air velocity and correspond to this body part, both in the static (42.2 W m 2 1C 1) and dynamic (40.3 W m 2 1C 1) tests. The feet and legs have rather higher convective heat transfer coefficients than the thighs. If a comparison is made between the lower and upper limbs (hand–foot; arm–leg; upper arm–thigh), we conclude that the lower limbs have lower values of hconv.
Finally, despite the reported left hand asymmetry, precisely where the higher values of hconv were registered, there is a general symmetry between the left and right sides of the body. The convective heat transfer coefficients as a function of the air velocity are now presented in Figs. 4–7. For the central body parts (head, chest, back and crutch) the curves are similar which indicates that the walking movements have a small effect on hconv. This behaviour was also noticed during the natural convection tests. Furthermore, the pelvic region and back have the smallest values of hconv. In the lower and upper limbs the differences between the static and dynamic results become less evident in the upper arms and thighs. Thus, the body movements have a more pronounced effect on the parts with higher amplitude such as hands and feet. In fact, it is a well known phenomenon that the hands and feet are the preferred body parts for heat losses (de Dear et al., 1997; Oguro et al., 2002a, 2002b; Oliveira et al., 2012).
2 - Right foot 50
40
40
hconv [W/m 2.ºC]
hconv [W/m2.ºC]
1 - Left foot 50
30
hconv = 11.61va0.39 20
hconv = 9.61va0.51
10
Static
30
hconv = 11.88va0.41 20
hconv = 9.89va0.52
10
Dynamic
0
Static
0 0
2
4
6
8
10
0
2
4
va [m/s]
8
10
4 - Right Leg
50
50
40
40
hconv [W/m2.ºC]
hconv [W/m2.ºC]
6 va [m/s]
3 - Left Leg
30
hconv = 10.38va0.34 20
10
Static
hconv = 8.33va0.48
30
hconv = 10.34va0.35
20
10
Dynamic
0
hconv = 8.44va0.48
Static
Dynamic
0 0
2
4
6
8
10
0
2
4
va [m/s]
6
8
10
va [m/s]
6 - Right Thigh
5 - Left Thigh 50
50
40
40
hconv [W/m2.ºC]
hconv [W/m2.ºC]
Dynamic
30
hconv = 7.90va0.39 20
10
Static
30
hconv = 8.35va0.39 20
10
Dynamic
Static
Dynamic
hconv = 7.48va0.44
hconv = 7.23va0.43 0
0 0
2
4
6
8
10
0
2
va [m/s]
4
6 va [m/s]
Fig. 6. Convective heat transfer coefficients – lower limbs.
8
10
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
10 - Right hand
9 - Left hand 50
50
40
40
hconv = 15.50va0.43
hconv [W/m2.ºC]
hconv [W/m2.ºC]
75
30
20
hconv = 13.78va0.52
10
hconv = 13.37va0.41 30
20
10 Static
Dynamic
hconv = 11.60va0.50
0 2
4
6
8
10
0
2
4
va [m/s]
6
8
10
va [m/s]
11 - Left arm
12 - Right arm
50
50
40
40
hconv [W/m2.ºC]
hconv [W/m2.ºC]
Dynamic
0 0
30
hconv = 9.11va0.39 20
10
30
hconv = 9.89va0.40 20
10 Static
hconv = 8.01va0.46
Dynamic
Static
hconv = 8.58va0.44
0
Dynamic
0 0
2
4
6
8
10
0
2
4
va [m/s]
6
8
10
va [m/s]
14 - Right upper arm
13 - Left upper arm 50
50
40
40
hconv [W/m2.ºC]
hconv [W/m2.ºC]
Static
30
hconv = 8.80va0.49 20
10
30
hconv = 8.98va0.45 20
10 Static
hconv = 8.04va0.55
Dynamic
Static
hconv = 8.20va0.50
0
Dynamic
0 0
2
4
6
8
10
0
2
4
6
8
10
va [m/s]
va [m/s]
Fig. 7. Convective heat transfer coefficients – upper limbs.
The present study clearly demonstrates the reduction of the heat losses from the peripheral parts towards the centre of the body. In the central body parts the values of hconv that were obtained can be considered underestimated for use in humans because of the differences between manikins and human body motion. On the contrary, the results obtained in the extremities can be considered overestimated for humans since the swing effect achieved with the pneumatic mechanism is more prone to reproduce the motion of a military rather than the motion observed in typical daily routines. In fact, our natural movements are not that stiff. This warns for the careful use of equations proposed in the literature for the whole body and supports the fact that important errors can be committed if those correlations are assumed for the body parts. For instance, the present results highlight that the dynamic hconv value of the left hand doubles the one of the pelvic region and the corresponding difference for the static experiments almost reaches 2.5 times more.
5. Conclusion The present work is focused on the convective heat transfer coefficients for the whole body and for the body parts and provides an approach for further understanding of heat transfer phenomena around the human body, particularly when motion and air velocity are considered. These effects are analysed considering the standing static posture and a walking speed of 45 steps/min and free air stream velocities between 0 and about 10 m/s. Natural and forced convection were thus considered. In the case of forced convection, general equations for the static and dynamic conditions are presented for the whole body and for the body parts. The present study was carried out with the manikin operating under the thermal comfort regulation mode. This manikin control mode is not often considered, so the comparison between the present results and others from the literature is important to assess the differences provided by other control modes. From a
76
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
practical point of view, it can be argued that the whole body results obtained with the thermal comfort regulation mode are within the most common range available in the literature. This is particularly true when lower air velocities are considered. In addition, when the walking effect is considered, it is frequently achieved with a mechanism that allows both arms and legs to move but with the manikin walking on the spot. This condition differs from human walking movements where displacement in place occurs. The convection counterpart of the heat transfer is thus obviously different in the simulated and really body motion. Finally, any comparison must be addressed with caution and analysis based only on the hconv values is not advised, particularly in forced convection, where the flow conditions have a very significant role and are very difficult to reproduce in different wind tunnels. References Anttonen, H., 2000. Interlaboratory trial of thermal manikin based on thermal insulation of cold protective clothing in accordance with ENV 342. In: Nilsson H., Holmér I. (Eds.). Proceedings of the 3I3M – 3rd International Meeting on Thermal Manikin Testing, National Institute for Working Life, Stockholm, Sweden, 12–13 October 1999, pp. 25–28. Anttonen, H., Niskanen, J., Meinander, H., Bartels, V., Kuklane, K., Reinertsen, R., Varieras, S., Sołtyński, K., 2004. Thermal manikin measurements—exact or not? Int. J. Occup. Saf. Ergon. 10 (3), 291–300. ASHRAE, 2001. ASHRAE Handbook: Fundamentals. Bü ttner, K.J.K., 1934. Die Wä rmeü bertragung durch Leitung und Konvektion, Verdunstung und Strahlung in Bioklimatologie und Meteorologie, Nr. 404. Springer, Berlin. Chang, S.K.W., Gonzalez, R.R., 1989. Analysis of articulated manikin based convective heat transfer during walking, Technical Report T11-89. United States Army Research Institute of Environmetal Medicine, United States Army, Natick, Massachusetts. Colin, J., Houdas, Y., 1967. Experimental determination of coefficient of heat exchanges by convection of human body. J. Appl. Physiol. 22, 31–38. de Dear, R.J., Arens, E., Hui, Z., Oguro, M., 1997. Convective and radiative heat transfer coefficients for individual human body segments. Int. J. Biometeorol. 40, 141–156. Defraeye, T., Blocken, B., Koninckx, E., Hespel, P., Carmeliet, J.E., 2011. Computational fluid dynamics analysis of drag and convective heat transfer of individual body segments for different cyclist positions. J. Biomech. 44, 1695–1701, http://dx. doi.org/10.1016/j.jbiomech.2011.03.035. Directive 2010/31/EU of the European Parliament and of the Council of 19 May 2010 on the energy performance of buildings. Eur-Lex Official Journal L:13–35. Du Bois, D., Du Bois, E.F., 1916. Clinical calorimetry: tenth paper a formula to estimate the approximate surface area if height and weight be known. Arch. Intern. Med. XVII, 863–871. Fanger, P.O., 1972. Thermal Comfort—Analysis and Applications in Environmental Engineering. McGraw-Hill, New York. Ferreira, A.D., Oliveira, R.A., 2009. Wind erosion of sand placed inside a rectangular box. J. Wind Eng. Ind. Aerodyn. 97, 1–10, http://dx.doi.org/10.1016/j. jweia.2008.09.001. Fukazawa, T., Ando, T., Ikeda, S., Yamaguchi, A., Holmér, I., Tochiara, Y., 2009. Convective heat transfer coefficient from baby smaller than from adult. In: Proceedings of the 13th International Conference on Environmental Ergonomics. ISBN:978-1-74128-178-1. Gagge, A.P., 1937. A new physiological variable associated with sensible and insensible perspiration. Am. J. Physiol.—Leg. Content 120, 277–287. Gagge, A.P., Nishi, Y., Nevins, R.G., 1976. The role of clothing in meeting FEA energy conservation guidelines. ASHRAE Trans. 82, 234. Gagge, A.P., Stolwijk, J.A.J., Saltin, B., 1969. Comfort and thermal sensations and associated physiological responses during exercise at various ambient temperatures. Environ. Res. 2, 209–229, http://dx.doi.org/10.1016/0013-9351(69) 90037-1. Havenith, G., Nilsson, H., 2004. Correction of clothing insulation for movement and wind effects: a meta-analysis. Eur. J. Appl. Physiol. 92, 636–640, http://dx.doi. org/10.1007/s00421-004-1113-6. Ichihara, M., Saitou, M., Tanabe, S., Nishimura, M., 1995. Measurement of convective heat transfer coefficient and radiative heat transfer coefficient of standing human body by using thermal manikin. In: Proceedings of the Annual Meeting of the Arquitectural Institute of Japan, pp. 379–380. International Organization for Standardization, ISO 7726, 1998. Ergonomics of the thermal environment—Instruments for Measuring Physical Quantities, International Standard. International Organization for Standardization (ISO), Geneva.
International Organization for Standardization, ISO 7730, 2005. Ergonomics of the Thermal Environment—Analytical Determination and Interpretation of Thermal Comfort Using Calculation of the PMV and PPD Indices and Local Thermal Comfort Criteria. International Standard, International Organization for Standardization (ISO); Geneva. Kerslake, D.M., 1972. The Stress of Hot Environment. Cambridge University Press, Cambridge. Kurazumi, Y., Tsuchikawa, T., Ishii, J., Fukagawa, K., Yamato, Y., Matsubara, N., 2008. Radiative and convective heat transfer coefficients of the human body in natural convection. Build. Environ. 43, 2142–2153, http://dx.doi.org/10.1016/j. buildenv.2007.12.012. Kuwabara, K., Mochida, T., Kondo, M., Matsunaga, K., 2001. Measurement of man's convective heat transfer coefficient by using a thermal manikin in the middle wind velocity region. J. Hum. Living Environ. 8, 27–32. Luo, N., Weng, W.G., Fu, M., Yang, J., Han, Z.Y., 2014. Experimental study of the effects of human movement on the convective heat transfer coefficient. Exp. Therm. Fluid Sci. 57, 40–56, http://dx.doi.org/10.1016/j.expthermflusci. 2014.04.001. Madsen, T.L., 1976. Description of thermal manikin for measuring thermal insulation values of clothing. Thermal Insulation Report no. 48. Missenard, F.A., 1973. Coefficient d´échange de chaleur du corps humain par convection en fonction de la position, de l´activité du sujet et de l´environment. Arch. Sci. Physiol. 27, 45–50. Mitchell, D., 1974. Convective Heat Transfer in Man and Other Animals. B. Publishing, London. Mochida, T., 1977. Convective and radiative heat transfer coefficients for the human body. Bull. Fac. Eng., Hokkaido Univ. 84, 1–11. Mochida, T., Nagano, K., Shimakura, K., Kuwabara, K., Nakatani, T., Matsunaga, K., 1999. Characteristics of convective heat transfer of thermal manikin exposed in air flow from front. J. Hum. Living Environ. 6, 98–103. Nielsen, M., Pedersen, L., 1953. Studies on the heat loss by radiation and convection from the clothed human body. Acta Physiol. Scand. 27, 272–294. Nilsson, H., 2004. Comfort Climate Evaluation with Thermal Manikin Methods and Computer Simulation Models (Ph.D. thesis). University of Gävle. Nilsson, H., Anttonen, H., Holmér, I., 2000. New algorithms for prediction of wind effects on cold protective clothing. In: Proceedings of the 1st European Conference on Protective Clothing. Arbete och Halsa 8, Stockholm, pp. 17–20. Nishi, Y., Gagge, A.P., 1970. Direct evaluation of convective heat transfer coefficient by naphthalene sublimation. J. Appl. Physiol. 29, 830–838 (cited by ASHRAE, 2001). Oguro, M., Arens, E., de Dear, R., Zhang, H., Katayama, T., 2002a. Convective heat transfer coefficients and clothing insulations for parts of the clothed human body under calm conditions. J. Archit., Plan. Environ. Eng. (Trans. AIJ) 561, 31–39. Oguro, M., Arens, E., de Dear, R., Zhang, H., Katayama, T., 2002b. Convective heat transfer coefficients and clothing insulations for parts of the clothed human body under airflow conditions. J. Archit., Plan. Environ. Eng. (Trans. AIJ) 561, 21–29. Oliveira, A.V.M., Gaspar, A.R., Quintela, D.A., 2008. Measurements of clothing insulation with a thermal manikin operating under the thermal comfort regulation mode. Comparative analysis of the calculation methods. Eur. J. Appl. Physiol. 104, 679–688, http://dx.doi.org/10.1007/s00421-008-0824-5. Oliveira, A.V.M., Gaspar, A.R., Quintela, D.A., 2011. Dynamic clothing insulation. Measurements with a thermal manikin operating under the Thermal comfort regulation mode. Appl. Ergon. 42, 890–899, http://dx.doi.org/10.1016/j. apergo.2011.02.005. Oliveira, A.V.M., Gaspar, A.R., Francisco, S.C., Quintela, D.A., 2012. Convective heat transfer from a nude body under calm conditions: assessment of the effects of walking with a thermal manikin. Int. J. Biometeorol. 56, 319–332, http://dx.doi. org/10.1007/s00484-011-0436-3. Omori, T., Yang, J.H., Kato, S., Murakami, S., 2004. Coupled simulation of convection and radiation on thermal environment around an accurately shaped human body. In: Proceedings of the RoomVent 2004, 9th International Conference on Air Distribution in Rooms. Ono, T., Murakami, S., Ooka, R., Omori, T., 2008. Numerical and experimental study on convective heat transfer of the human body in the outdoor environment. J. Wind Eng. Ind. Aerodyn. 96, 1719–1732, http://dx.doi.org/10.1016/j. jweia.2008.02.007. Parsons, K.C., 2003. Human Thermal Environments: The Effects of Hot, Moderate, and Cold Environments on Human Health, Comfort, and Performance (527 pp.). Quintela, D., Gaspar, A., Borges, C., 2004. Analysis of sensible heat exchanges from a thermal manikin. Eur. J. Appl. Physiol. 92, 663–668, http://dx.doi.org/10.1007/ s00421-004-1132-3. Seppänen, O., McNall, P.E., Munson, D.M., Sprague, C.H., 1972. Thermal insulating values for typical indoor clothing ensembles. ASHRAE Trans., 120–130. Silva, M.C.G., Coelho, J.A., 2002. Convection coefficients for the human body parts determined with a thermal manikin. In: Proceedings of 8th International Conference on Air Distribution in Rooms, ROOMVENT 2002. Winslow, C.E.A., Gagge, A.P., Herrington, L.P., 1939. The influence of air movement upon heat losses from the clothed human body. Am. J. Physiol.—Leg. Content 127, 505–518 (cited by Fanger, 1970).
Contents lists available at ScienceDirect
Journal of Wind Engineering and Industrial Aerodynamics journal homepage: www.elsevier.com/locate/jweia
Analysis of natural and forced convection heat losses from a thermal manikin: Comparative assessment of the static and dynamic postures A. Virgílio M. Oliveira a,b,n, Adélio R. Gaspar a, Sara C. Francisco a, Divo A. Quintela a a
ADAI, LAETA, Department of Mechanical Engineering, University of Coimbra, Pólo II, 3030-788 Coimbra, Portugal Coimbra Institute of Engineering, Polytechnic Institute of Coimbra, Department of Mechanical Engineering, Rua Pedro Nunes, Quinta da Nora, 3030-199 Coimbra, Portugal
b
art ic l e i nf o
a b s t r a c t
Article history: Received 30 January 2014 Received in revised form 20 May 2014 Accepted 29 June 2014
The present experimental work is dedicated to the analysis of the effect of walking movements and air velocity on the convective heat transfer coefficients (hconv) of the human body. A wind tunnel and an articulated thermal manikin of the Pernille type with sixteen body segments were used. Beyond the standing posture (static condition), a step rate of 45 steps/min was selected, corresponding to a walking speed of 0.51 m/s (dynamic condition). The free stream air velocity was varied from 0 to about 10 m/s. The experimental conditions were thus varied from natural to forced convection. The convection coefficients for the different body segments and the whole body were determined for each air velocity giving details about the differences between them. In the case of the whole body, for the standing static reference posture and free convection, the mean value of hconv was equal to 3.5 W m 2 1C 1. In the dynamic condition the corresponding hconv value was 4.5 W m 2 1C 1. In forced convection, the highest values correspond to the highest wind speed and were equal to 22.4 W m 2 1C 1 for the static condition and 23.0 W m 2 1C 1 for the dynamic posture. & 2014 Elsevier Ltd. All rights reserved.
Keywords: Convective heat transfer coefficients Human body Thermal manikin
1. Introduction The Kyoto Protocol and all the foregoing initiatives have been underlining the need for the reduction of energy consumption related with global climate changes. Such goal has several effects on industry, transportation and built environment, among others. In a certain sense, the EU Directive 2010/31/EU on the energy performance of buildings is a recognition that further efforts are needed in order to balance between building quality, energy efficiency and comfort requirements. With this in mind, and whenever we deal with human occupancy, it seems acceptable that improved analysis of heat transfer phenomena around human body may be of interest for such matters. The present contribution is directed to this particular goal. According to the latest experimental and numerical developments characterized by an increasing need for detail, this study is focused on the experimental evaluation of convective heat transfer coefficients for the whole
n Corresponding author at: Coimbra Institute of Engineering, Polytechnic Institute of Coimbra, Department of Mechanical Engineering, Rua Pedro Nunes, Quinta da Nora, 3030-199 Coimbra, Portugal. Tel.: þ 351239790330; fax: þ 351239790331. E-mail address: [email protected] (A.V.M. Oliveira). URLS: http://www.isec.pt (A.V.M. Oliveira), http://www.dem.uc.pt (A.R. Gaspar), http://www.dem.uc.pt (S.C. Francisco), http://www.dem.uc.pt (D.A. Quintela).
http://dx.doi.org/10.1016/j.jweia.2014.06.019 0167-6105/& 2014 Elsevier Ltd. All rights reserved.
human body and the body parts. Discussion about the convective trends in natural, mixed and forced regimes is also foreseen. When heat transfer coefficients are considered, it must be recognized that most data available in the literature refer to the whole body and some of the first studies took place in the thirties by Büttner (1934) and Gagge (1937). Later on, the whole body natural convection heat transfer coefficients were characterized as equal to 5.1 W m 2 1C 1 by Colin and Houdas (1967), 4.0 W m 2 1C 1 by Seppänen et al. (1972) and 3.1 W m 2 1C 1 by Mitchell (1974). Afterwards, de Dear et al. (1997) obtained values of 3.4 and 3.3 W m 2 1C 1 for a standing and sitting thermal manikin, respectively, and Omori et al. (2004) proposed a value of 3.3 W m 2 1C 1 for the standing posture. With the development of thermal manikins, the heat losses from each part of the body have also been addressed and a set of data is now available. Oguro et al. (2002a) and Quintela et al. (2004) proposed expressions to calculate hconv in natural convection as a function of the difference between the skin and the air temperatures ðT sk T a Þ. While Oguro et al. (2002a) considered a manikin dressed and nude for both standing and sitting postures, in the work of Quintela et al. (2004) the experimental determination of convective and radiative heat transfer coefficients in standing, sitting and lying postures were considered. Nilsson (2004) assessed the repeatability of heat transfer coefficients by considering five independent calibrations performed from
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
1995 to 1998 and the results showed that the mean standard deviation was 8% for all zones. Kurazumi et al. (2008) have also investigated the convective and radiative heat transfer coefficients, focusing their research on the convective heat transfer area of the human body. Under natural convection conditions, the mannequin was placed in seven postures and the radiative heat transfer coefficient was determined for each posture and empirical formulas were proposed for the convective heat transfer coefficient of the entire human body, driven by the difference between the air and mean skin temperatures. The effect of body motion was first considered in the work of Chang and Gonzalez (1989) using the sublimation technique with circular naphthalene disks affixed to five different segments of one articulated manikin. The static posture and four walking gaits (20, 40, 68 and 80 steps/min) were studied. The results for the heat transfer coefficient by convection have an intrinsic local nature and should not be compared with more recent thermal manikins where such coefficients are averaged through the surface of different parts. More recently, the effect of walking on the convective heat transfer coefficients was considered by Oliveira et al. (2012). Beyond the standing static posture, three step rates were considered (20, 30 and 45 steps/min) and the convective heat transfer coefficients for the whole body and for the different body segments were determined for each step rate. In the case of the whole body, for the standing static reference posture, a correlation for natural convection was also presented. In the case of forced convection, a relevant study was developed by de Dear et al. (1997) through a wide range of wind speeds, varying from still air to 5.0 m/s. Both standing and sitting postures were investigated, as there were eight different azimuth angles. Another interesting work was carried out by Silva and Coelho (2002), considering three flow incidences (front, back and side) and two postures (seated and standing), and Oguro et al. (2002b) considered the same postures with the manikin nude and clothed facing upwind and downstream flows. Ono et al. (2008) measured the convective heat transfer coefficient for the human body in an outdoor environment by means of a thermal manikin placed in a wind tunnel complemented with a computational fluid dynamics (CFD) analysis. Different wind velocities were considered and the mean convective heat transfer coefficients for the whole body were 6.6, 10.0 and 16.3 W m 2 1C 1, for 0.5, 1.0 and 2.0 m/s, respectively. In a different perspective, Defraeye et al. (2011) investigated the convective heat transfer for cyclists. The convective heat transfer coefficients of the whole body and of the 19 body parts of a cyclist were determined. Wind speed correlations based on power-law equations were proposed and the results have shown that the power-law exponent does not differ significantly for the individual body segments for all positions tested. More recently, Luo et al. (2014) considered five moving speeds of a manikin (0.2, 0.5, 0.8, 1.1 and 1.3 m/s) carried out with a motor on a 10 m-length-rail, and four temperature differences (4, 8, 12 and 16 1C) by changing the heating power of the manikin. They concluded that to explain the details of the differences between the moving condition and the wind tunnel condition, which may be caused by the flow field, further analysis of the flow field should be foreseen. In a complementary perspective, extensive research has been done in the assessment of the thermal insulation of clothing, a field directly related to heat transfer by convection and radiation. When the analysis of human typical postures is considered and restricting the focus to experimental conditions close to the ones of the present study, i.e., considering the walking effect and high air velocity, the studies of Anttonen (2000); Nilsson et al. (2000); Havenith and Nilsson (2004); Anttonen et al. (2004) and Oliveira et al. (2008, 2011) are relevant contributions. In the present work the convective heat transfer coefficients are addressed in a complementary perspective with novel approaches.
67
Body motion is taken into account – a rather seldom feature considered in previous studies besides the work of Chang and Gonzalez (1989) – in both natural and forced convection, but the actual methodology as well as the thermal manikin differs significantly from the available in the eighties and followed by Chang and Gonzalez (1989). In addition, a wider range of air velocity is considered, up to 10 m/s. Such air velocity values are also not easy to find in other studies, and this distinct characteristic is presented in this paper. Several combinations of tests were performed and comparisons between static (i.e. standing still) and dynamic (i.e. walking at 45 steps/min) conditions are made. The main purpose is to measure convective heat transfer coefficients for all the conditions tested with the manikin operating under the thermal comfort regulation mode, a condition that represents another different approach and also not often addressed, besides the work of Oguro et al. (2002a) and Oliveira et al. (2012).
2. Methods 2.1. Heat transfer coefficients The exchange of heat between a solid and the environment can occur by convection (C), radiation (R) and conduction (K). In the case of the human body, the heat transfer by conduction is limited to the body parts in contact with solid surfaces, which are generally restricted to a few, namely to the feet, and thus usually neglected. Therefore, the heat is transferred mainly by convection and radiation. The study of convective heat transfer between the human body and the environment represents a great challenge. In fact, the heat transfer by convection depends on various factors and exact analytical solutions are actually not available. Hence, the heat transfer by convection is generally calculated making use of Newton's law of cooling (Fanger, 1972) C ¼ hconv ðT s T a Þ
ð1Þ
where hconv is the convective heat transfer coefficient [W m 2 1C 1] and T s and T a represent the mean temperatures of the body surface and the environment [1C]. For the dressed human body, the heat transfer by convection is calculated in an analogous way from (Fanger, 1972; Parsons, 2003) C ¼ f cl hconv ðT cl T a Þ
ð2Þ
where f cl is the clothing area factor and T cl is the mean clothing surface temperature [1C]. In the case of radiation, the heat transfer for the dressed human body is calculated by an expression similar to Newton's law (Fanger, 1972; Parsons, 2003) R ¼ f cl hrad ðT cl T r Þ
ð3Þ 2
1
where hrad is the radiative heat transfer coefficient [W m 1C ] and T r is the mean radiant temperature [1C]. Therefore, in studies concerning the human body heat exchanges with the environment, the overall sensible heat released by convection and radiation, Q_ s [W m 2], can be calculated from the following equation: Q_ s ¼ f cl hconv ðT cl T a Þ þ f cl hrad ðT cl T r Þ
ð4Þ
for the case of a clothed body (Fanger, 1972; Parsons, 2003), or from Q_ s ¼ hconv ðT sk T a Þ þ hrad ðT sk T r Þ
ð5Þ
for the naked human body (Fanger, 1972; Parsons, 2003), where T sk is the mean skin temperature [1C].
68
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
As a first approach, the radiative heat transfer coefficient for the whole body, hrad [W m 2 1C 1], can be calculated from the expression (ASHRAE, 2001)
equations with the following general form:
" #3 Arad T cl þ T r 273:2 þ hrad ¼ 4 σ ε ADuBois 2
Tables 2 and 3 list values and correlations proposed by several authors for natural and forced convection, respectively. The analysis of such values and correlations must be done with caution since the information about the experimental procedures in which the values and equations are based is rarely available. For instance, data about posture, environmental conditions of the tests and air velocity limits play an important role in the differences that are found and so conclusive comparisons are difficult to achieve.
ð6Þ
where σ is the Stephan–Boltzman constant (5.67 10 8 W m 2 K 4), ε is the average body surface emissivity, Arad is the effective radiation area of the human body, ADuBois is the Du Bois and Du Bois (1916) body surface area. The emissivity of the body can be assumed as 0.95, as already adopted by other manikin owners (de Dear et al., 1997; Quintela et al., 2004; Fukazawa et al., 2009), and the effective radiation area factor Arad/ADuBois for the standing posture is generally accepted as 0.73 (Fanger, 1972). The literature provides hrad values for the whole body and for the different body sections obtained with thermal manikins (de Dear et al., 1997; Ichihara et al., 1995). Nevertheless, preference was given to results obtained during tests carried out by Quintela et al. (2004) with the same manikin as used in this study (see Table 1). In the present study the evaluation of the convective heat transfer coefficient for the case of a nude human body is one of the main goals. For this reason, Eq. (5) can be expressed in terms of hconv by hconv ¼
Q_ s hrad ðT sk T r Þ
ð7Þ
ðT sk T a Þ
If attention is now given to the body parts, the convective heat transfer coefficient for each part can be computed from the following: hconv;i ¼
Q_ s;i hrad;i ðT sk;i T r Þ ðT sk;i T a Þ
ð8Þ
where the subscript i (i¼1,…,16) stands for individual reference. The analysis of both natural and forced convective heat transfer coefficients is considered in the literature by different approaches. In the case of natural convection, single values or correlations based on the difference between the clothing/skin and the air/ operative temperatures are the most common. On the other hand, for forced convection, the literature usually suggests regression
hconv ¼ C vna
ð9Þ
3. Measurement procedures 3.1. Wind tunnel The measurements were carried out in a wind tunnel at the Industrial Aerodynamics Laboratory. The wind tunnel is a closed circuit with an opened squared test chamber (2 2 m2) and 5 m long where the free stream air velocity can be varied from 0 to 18 m s 1 (Ferreira and Oliveira, 2009). Fig. 1a shows part of the test chamber and the downstream opening for flow recirculation and Fig. 1b shows the opening of the incoming air. The mean air velocities analysed in the present work ranged from 0 to 10 m s 1 (0 and approximately 0.25, 0.5, 0.75, 1.0, 2.5, 5 and 10 m s 1), thus gathering both indoors and outdoors conditions. 3.2. Thermal manikin A Pernille type thermal manikin named “Maria” (Fig. 1) was placed at the centre of the test chamber. The manikin is articulated and divided into 16 parts independently controlled by a computer according to the following relation between dry heat losses and skin temperature of the human body under thermal comfort conditions (Fanger, 1972; Madsen, 1976): T sk ¼ 36:4 0:054 Q_ s
ð13Þ
The manikin is made of a fibreglass armed polyester shell covered with a thin nickel wire wound around all the body to ensure heating and temperature measurement. It has joints at the
Table 1 Values of hrad [W m 2 1C 1] proposed by Ichihara et al. (1995), de Dear et al. (1997) and Quintela et al. (2004). Mannequin section
1 – Left foot 2 – Right foot 3 – Left leg 4 – Right leg 5 – Left thigh 6 – Right thigh 7 – Pelvic region 8 – Head 9 – Left hand 10 – Right hand 11 – Left arm 12 – Right arm 13 – Left upper arm 14 – Right upper arm 15 – Chest 16 – Back Total
Ichihara et al. (1995) (ε ¼ 0.96)
de Dear et al. (1997)
Quintela et al. (2004)
(ε ¼ 0.95)
(ε ¼ 0.95)
Standing
Sitting
Standing
Sitting
Standing
Lying
7.3
4.2 4.2 5.3 5.3 4.3 4.3 4.2 4.1 4.1 4.1 4.9 4.9 5.2 5.2 4.5 4.4 4.5
3.9 3.9 5.3 5.3 4.3 4.3 4.2 4.1 4.1 4.1 4.9 4.9 5.2 5.2 4.5 4.4 4.5
4.7 4.9 4.7 4.6 4.6 4.2 4.4 5.7 4 4.6 4 4.4 4.7 4.4 3.5 3.9 4.4
5 5 5 5.1 4.8 4.4 4.5 5.7 4.3 4.1 4.5 4.1 4.9 4.2 4.5 4.2 4.6
5.2 5.2 4.2 4.7 4.5 4 3.9 5.6 4.3 4.1 4.1 3.9 4.8 4.1 4 3.5 4.3
4.8 4.2 3.9 4.3 3.7 3.9 4.0 3.8 3.6
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
69
Table 2 Correlations and values of hconv for natural convection. hconv [W m 2 1C 1]
va [m s 1]
Posture conditions
hconv ¼ 2:38 ðt cl t a Þ0:25
va o 0.15
Sitting and standing
5.1 4.0 3.1
va o 0.15 va o 0.15 va o 0.2 65o M o 175M [W m 2] va o 0.1 va o 0.1 va o 0.1
Reclining position Standing Sitting Sedentary office work Sitting Standing Sitting
hconv ¼ 2:48 ðt sk t o Þ0:18
va o 0.1 va o 0.1 va o 0.15
Standing Standing Lying
hconv ¼ 1:94 ðt sk t o Þ0:23
va o 0.15
Sitting
hconv ¼ 2:02 ðt sk t o Þ0:24
va o 0.15
Standing
hconv ¼ 1:007 ðt sk t a Þ0:406
va o 0.2
hconv ¼ 1:16 ðM 50Þ0:39 3.3 3.4 hconv ¼ 0:78 ðt sk t o Þ
0:59
hconv ¼ 1:21 ðt sk t o Þ0:43 3.3
Reference Nielsen and Pedersen (1953) adopted by Fanger (1972) Colin and Houdas (1967) Seppänen et al. (1972) Mitchell (1974) Gagge et al. (1976) de Dear et al. (1997) Oguro et al. (2002a) Omori et al. (2004) Quintela et al. (2004)
Standing. exposed to atmosphere
Kurazumi et al. (2008)
Standing. floor contact Chair sitting. exposed to atmosphere
hconv ¼ 1:183 ðt sk t a Þ0:347 hconv ¼ 1:175 ðt sk t a Þ0:351
Chair sitting. contact with seat. Chair back and floor
hconv ¼ 1:222 ðt sk t a Þ0:299
Cross-legged sitting. floor contact
hconv ¼ 1:271 ðt sk t a Þ0:355 hconv ¼ 1:002 ðt sk t a Þ0:409
Legs-out sitting. floor contact
hconv ¼ 0:881 ðt sk t a Þ0:368
Supine. Floor contact
Table 3 Correlations of hconv for forced convection. hconv [W m 2 1C 1]
Reference
hconv [W m 2 1C 1]
Reference Mochida et al. (1999)
hconv ¼ 6:3 v0:5 a
Büttner (1934)
Seated manikin hconv ¼ 2:82 þ 6:35 v0:848 a
hconv ¼ 12:13 v0:5 a
Winslow et al. (1939)
Standing manikin hconv ¼ 10:14 v0:431 a
hconv ¼ 2:67 þ 8:72 v0:67 a
Colin and Houdas (1967)
Seated manikin hconv ¼ 10:1 v0:61 a
hconv ¼ 8:6 v0:531 a
Gagge et al. (1969)
Standing manikin hconv ¼ 10:4 v0:56 a
hconv ¼ 6:51 v0:391 a pffiffiffiffiffi hconv ¼ 8:3 va
Nishi and Gagge (1970)
Both seated and standing postures hconv ¼ 10:3 v0:6 a
Kerslake (1972)
hconv ¼ 14:8 v0:69 a hconv ¼ 3:5 þ 5:2 va with va r 1 m=s
Seppänen et al. (1972)
Nude and clothed manikin hconv ¼ 3:36 þ 6:86 v0:92 a hconv ¼ 3:9 va þ 3:6 (n) Standing manikin – Front flow
Missenard (1973)
hconv ¼ 4:0 va þ 3:6 (n) Standing manikin – Side flow
hconv ¼ 8:7 v0:6 a with va 4 1 m=s
Missenard (1973)
hconv ¼ 4:2 va þ 3:5 (n) Standing manikin – Back flow
hconv ¼ 8:3 v0:6 a
Mitchell (1974)
(n) Nude manikin, Upstream flow hconv ¼ 9:31 v0:6 a
hconv ¼ 5:7 ðM 0:85Þ0:39 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi hconv ¼ 3 270 v2a þ 23
Gagge et al. (1976)
(n) Nude manikin, Downstream flow hconv ¼ 9:41 v0:61 a
Mochida (1977)
(n) Clothed manikin, Upstream flow hconv ¼ 13:36 v0:6 a
hconv ¼ 15:4 v0:63 Standing manikin a
Ichihara et al. (1995)
hconv ¼ 12:38 v0:65 (n) Clothed manikin, Downstream flow a
de Dear et al. (1997)
Kuwabara et al. (2001) Silva and Coelho (2002)
Oguro et al. (2002b)
(n) – Standing.
shoulders, hips and knees. They are all made of a circular cut in such a way that standing as well as sitting postures are quite natural. 3.3. Walking device A mechanism with the capability to simulate realistic walking movements was developed to reproduce the effect of body motion. The supporting frame of the manikin has wheels with brakes in order to facilitate the transportation. The manikin is hung from the head and the feet are kept 0.15 m away from the floor. The other necessary fixing point is achieved with a waist belt. The power transmission lies on a pneumatic system with a unique cylinder, which applies a swinging movement to legs and arms by fixing supports to wrists and knees. This action can be
controlled in order to simulate different step rates. The only movable parts are then the arms and legs. To provide control of the walking movement the mechanism has an electro valve and four autonomous flow control valves. The cylinder is connected to a central flat bar that distributes the movement to each limb by spherical connections. The lower and upper limbs are then connected to the same central bar ensuring that they all move at the same speed. The cylinder, the electro valves and the flow control valves are all from Festo and the spherical connections are type Pilloballs from IKO Nippon Thompson. The mechanical components were kept to a minimum in order to make the apparatus as simple and reliable as possible. The design and construction of the walking device was made at the Industrial Aerodynamics Laboratory.
70
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
of the manikin. The mean value of the last ten minutes of each test was considered for reference. Finally, the environmental parameters in the Laboratory, namely the air temperature (ta) and the relative humidity (rh), were monitored with the data logger Testo 175-H2 (refª 0563 1758) positioned away from the test area. 3.5. Test conditions The mean skin temperature ðt sk Þ and the heat flux ðQ_ s Þ of the manikin were continuously monitored, enabling an easy check of the steady state situation. Data acquisition was started after achieving this condition and lasted for 20 min, from which just the last 10 min were considered for analysis. The hconv calculation is thus performed based on t sk and Q_ s averages of these last 10 min. During the acquisition period, the heat flux and the skin temperature of each body part were recorded every minute. Two days were necessary to perform the sixteen experiments that began with the static posture. For this posture, the air temperature ranged between 12.9 and 13.7 1C and the difference between the air and operative temperatures did not exceed 0.21C. For the dynamic condition ta varied between 13.0 and 13.8 1C and the highest difference between ta and to was also 0.2 1C. The va STD values were calculated for both static and dynamic conditions and the higher value of each test was equal to 0.0381 and 0.0484 m s 1, respectively. The higher vertical temperature differences usually occur between the head and ankle levels, with ta and tg differences less than 0.5 1C. The radiant asymmetry never exceeded 0.2 1C.
4. Results and discussion
Fig. 1. Thermal manikin in the wind tunnel.
Two main sections are considered next for analysis, one focusing the whole body and the other one dealing with the local results of each body part. The discussions concerning the body parts are supported by two different approaches, pointing out their own differences and showing the variation of the convective heat transfer coefficients with air velocity.
3.4. Physical parameters 4.1. Convective heat transfer coefficients – whole body The physical parameters of the environment were measured according to ISO, 7726 (1998) requirements and using equipment from Brüel & Kjær, namely the heat stress monitor, type 1219, the thermal comfort metre, type 1213, and the indoor climate analyser, type 1212, and from Testo, namely the hot sphere sensor (refª 0635 1049) and the data loggers 445 (refª 0560 4450) and 175-H2 (refª 0563 1758). The vertical temperature difference was controlled with the heat stress monitor, type 1219, by measuring the air (ta) and globe (tg) temperatures at three heights corresponding to the head, abdomen and ankle levels of the manikin. The air (ta) and globe (tg) temperatures also support the estimation of the mean radiant temperature ðt r Þ according to the expression suggested in ISO 7726 (1998) for natural convection. The radiant asymmetry was measured with a plane radiant temperature sensor (refª MM0036) connected to the thermal comfort metre, type 1213, and the operative temperature (to) was measure with a to sensor (refª MM0023) connected to the indoor climate analyser, type 1212. The air velocity (va) was measured with a frequency of 1/10 Hz and with a hot sphere sensor Testo (refª 0635 1049) connected to the data logger Testo 445 (refª 0560 4450). In order to avoid any flow disturbance that might affect the manikin, the sensor was positioned upstream, just after the entry section of the wind tunnel test chamber, approximately at the height of the abdomen
In free convection, the estimated convective heat transfer coefficient for the standing static condition was 3.5 W m 2 1C 1. This value agrees with those found in the literature and is particularly close to the one proposed by de Dear et al. (1997) (3.4 W m 2 1C 1). In the dynamic condition the calculated hconv value was 4.5 W m 2 1C 1. This test with the manikin walking on the spot does not correspond to a pure natural convection situation, even though the experiment without an air flow was analysed together with the natural convection test. It must be underlined that this result depends on the way how motion is created. As the manikin is walking on the spot there is no changing of position in space so there is no convection from displacement. Hence, convection is due to the balancing of the body caused by walking, i.e., by arms and legs movements since the torso and the head are practically still. These areas (head, chest and back) represent almost 31% of the body surface area and approximately 42% if the pelvic region is also considered, which means that the whole body result is affected by those body parts. The experimental conditions that were used are thus more close to gym activities like walking on a treadmill. Regarding the forced convection results, the highest values correspond to the highest wind speed and were equal to 22.4 and 23.0 W m 2 1C 1, respectively for the static and dynamic
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
Table 4 Convective heat transfer coefficients and standard deviations. Air velocity [m s 1]
0 0.22 0.51 0.75 0.96 2.48 4.88 9.43
Table 5 Convective heat transfer coefficients for the 16 body parts of the thermal manikin.
Static tests (standing still)
Dynamic tests (45 steps/min)
Schematic
Manikin
hconv [W m 2 1C 1]
Convective heat transfer coefficient hconv [W m 2 1C 1]
Standard deviation STD [W m 2 1C 1]
Convective heat Standard transfer deviation STD coefficient hconv [W m 2 1C 1] 2 1 [W m 1C ]
representation
Body part
Static Dynamic de Dear et al. (1997)
3.5 3.5 5.7 6.2 6.7 11.3 16.4 22.4
0.035 0.037 0.009 0.022 0.005 0.052 0.053 0.069
4.5 4.8 6.0 6.7 7.4 11.4 17.0 23.0
0.015 0.08 0.038 0.029 0.004 0.058 0.782 0.324
8
14 12 10 6
Whole body
4
25
hconv = 8.17va0.43
2
hconv [W/m .ºC]
20
71
2
1 – Left foot 2 – Right foot 3 – Left leg 4 – Right leg 15 13 5 – Left thigh 6 – Right thigh 7 – Pelvic region 7 11 8 – Head 9 – Left hand 10 – Right hand 9 11 – Left arm 12 – Right arm 5 13 – Left upper arm 14 – Right upper 3 arm 15 – Chest 1 16 – Back Whole body
4.4 4.4 3.4 3.9 3.9 3.9 2.8 4.3 5.3 4.1 3.7 3.6 3.2
7.3 7.3 6.8 6.6 4.6 4.7 3.2 4.2 7.2 6.9 4.5 5.1 3.4
5.1 5.1 4.1 4.1 4.1 4.1 3.4 3.6 4.1 4.1 3.7 3.7 2.9
3.4
4.0
2.9
2.3 2.8 3.5
2.7 3.1 4.5
3.0 2.9 3.4
15
10
hconv=
7.34va0.49
5
Static
Dynamic
0 0
2
4
6
8
10
va [m/s]
Fig. 2. Convective heat transfer coefficients – whole body.
conditions. Table 4 presents the averages of hconv calculations and the respective standard deviations (STD). The highest STD value is 0.069 W m 2 1C 1 for the static tests and 0.782 W m 2 1C 1 for the dynamic experiments. Table 4 shows that, generally, the STD values are higher for the dynamic condition and for the highest wind speeds, fact that can be explained by a more difficult achievement of stable conditions in the thermal manikin. Fig. 2 refers to the present results and shows that the dynamic values are always higher than the static ones. The differences between the two conditions are higher for the lower air velocities, and decrease with the air velocity. For the higher air velocities, the results even become almost equal. The data from both trials also put in evidence the air velocity dependence of hconv. The static and dynamic power regression equations presented in Fig. 2 should not be used for air velocities higher than 6 m/s due to the relative position exchange of the corresponding curves above that limit. If we compare the present results based on the equation proposed for the static condition with others of the literature, one can conclude that the present results are both over and underestimated. In fact, this was expected and also sustains the validity of the present results. Such differences have to be acknowledged when simple comparisons are performed. For instance, if the values of hconv are calculated making use of the equations proposed by Büttner (1934); Gagge et al. (1969); Nishi and Gagge (1970); Kerslake (1972); Mitchell (1974); Ichihara et al. (1995) and de Dear et al. (1997), for air velocities up to 5 m s 1 and then compared, we can conclude that the present results are higher than those proposed by Büttner (1934) and Nishi and Gagge (1970), and lower than the others, despite being close to the results of Gagge et al. (1969) and Kerslake (1972). However, if the results of de Dear et al. (1997) are considered for reference, we
conclude that de Dear's results are higher than all the others, except for the case of Ichihara et al. (1995). In this last case, for an air velocity of 5 m s 1, the value of de Dear is 25.61 W m 2 1C 1, while the one proposed by Ichihara is 42.45 W m 2 1C 1. If we look at the mean relative differences up to 1 m s 1, the conclusion is that the present results overestimate the ones obtained by Büttner (1934) and Nishi and Gagge (1970) by 14.7 and 5.8%, respectively. On the contrary, the present results underestimate the ones obtained by Gagge et al. (1969); Kerslake (1972) and Mitchell (1974) by 12.5, 11.0 and 5.5%, respectively, with higher mean relative differences obtained for Ichihara et al. (1995) (48.1%) and de Dear et al. (1997) (26.4%). Differences of the same order can be found if the comparison is now applied between these authors. For such differences, the range of variation obtained in repeatability tests of heat transfer coefficients must be considered, as clearly demonstrated by Nilsson (2004). In reproducibility trials higher values are expected, so the mean relative differences above are not as significant as a simple analysis of the results might suggest. Furthermore, all of these results are considered valid as far as the conditions on which they were derived are fulfilled. Many reasons contribute to these discrepancies, namely the ones related with the methodology and with the features of the experimental devices. Some may be difficult to assess, like the characteristics of the environment created by climate chambers and wind tunnels, while others are easier to clarify. For instance, a thermal manikin with a higher body surface area will, under the same experimental conditions, naturally lead to lower heat transfer coefficient when compared to one with a smaller area. Despite all the other different experimental procedures this is the case, for example, of the manikin used in the present study (ADuBois ¼ 1.642 m2) when compared to the one used in de Dear's tests (ADuBois ¼1.471 m2). For the present results, the differences between the static and dynamic values decrease with the air velocity and a mean relative difference of about 10% was obtained between the static and dynamic results. In addition, the value of the exponent in both equations is close to 0.5, namely for the static condition, which is the classic value suggested in the literature to represent the convective phenomena. Table 3 lists some representative correlations proposed in the literature for the case of humanoid shapes and shows that the air velocity exponent varies between 0.391 and 1.0, with the majority of the values more close to 0.6.
72
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
hconv [ W/m 2.ºC ]
Fan speed 0 r.p.m. va=0 m/s
8 head 14 right upper arm 15 12 right arm
13 left upper arm 11 left arm
10
10 right hand
Fan speed 10 r.p.m. va=0.22 m/s
9 left hand
5
16 back
4 right leg
8 head 14 right upper…15
12 right arm
16 back
5 left thigh
6 right thigh
Fan speed 17 r.p.m. va=0.51 m/s
hconv [ W/m 2.ºC ]
12 right arm 10 right hand
11 left arm 9 left hand
5
16 back
0
6 right thigh 4 right leg
12 right arm
12 right arm
16 back
5 left thigh
6 right thigh
10 right hand 16 back
11 left arm 9 left hand
6 right thigh 4 right leg
Fan speed 151 r.p.m. va=4.88 m/s
10 right hand 16 back
5 left thigh
6 right thigh
16 back
0
6 right thigh 4 right leg
11 left arm 9 left hand
9 left hand
15
15 chest 5 left thigh 3 left leg
2 right foot 1 left foot 7 pelvic region Static Dynamic
hconv [ W/m 2.ºC ] 8 head 14 right upper…45
12 right arm 10 right hand 16 back
5 left thigh
6 right thigh
3 left leg
11 left arm
30
0
15 chest
2 right foot 1 left foot 7 pelvic region Static Dynamic
13 left upper arm
4 right leg
Fan speed 301 r.p.m. va=9.43 m/s
13 left upper arm
Dynamic
8 head 14 right upper…45
Dynamic
15
3 left leg
hconv [ W/m 2.ºC ]
15 chest
hconv [ W/m 2.ºC ]
10 right hand
5 left thigh
12 right arm
3 left leg
8 head 14 right upper 45 arm 12 right arm 30
15 chest
0
Static
2 right foot 1 left foot 7 pelvic region Static
9 left hand
5
2 right foot 1 left foot 7 pelvic region
Fan speed 77 r.p.m. va=2.48 m/s
0
11 left arm
4 right leg
13 left upper arm
5
13 left upper arm
10
10 right hand
Dynamic
10
Dynamic
8 head 14 right upper arm 15
3 left leg
8 head 14 right upper arm 15
3 left leg
hconv [ W/m 2.ºC ]
15 chest
hconv [ W/m 2.ºC ]
Fan speed 31 r.p.m. va=0.96 m/s
5 left thigh
Static
2 right foot 1 left foot 7 pelvic region Static
15 chest
0
Fan speed 25 r.p.m. va=0.75 m/s
13 left upper arm
10
9 left hand
5
2 right foot 1 left foot 7 pelvic region
Dynamic
8 head 14 right upper…15
11 left arm
4 right leg
3 left leg
2 right foot 1 left foot 7 pelvic region Static
13 left upper arm
10
10 right hand
15 chest
0
6 right thigh
hconv [ W/m 2.ºC ]
30 15 0
4 right leg
13 left upper arm 11 left arm 9 left hand 15 chest 5 left thigh 3 left leg
2 right foot 1 left foot 7 pelvic region Static Dynamic
Fig. 3. Convective heat transfer coefficients obtained in the static and dynamic tests with the manikin facing the incoming wind.
4.2. Convective heat transfer coefficients – body parts The convective heat transfer coefficients obtained in the tests without wind (va ¼0 m s 1) are listed in Table 5, together with the
results obtained by de Dear et al. (1997). The highest convective heat transfer coefficients are associated to the peripheral parts of the body and the lowest to the central body parts. In the static condition the highest value corresponds to the left hand (5.3 W m 2 1C 1)
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
and the lowest to the chest (2.3 W m 2 1C 1). In the dynamic condition the feet present the highest values (7.3 W m 2 1C 1) closely followed by the left hand (7.2 W m 2 1C 1). The lowest value (2.7 W m 2 1C 1) was also obtained in the chest. It is also important to point out the asymmetry of the hands, especially in the static tests. While the right hand is very close to the thigh, the left hand is opened and set apart from the thigh, thus probably outside the boundary layer developing from the feet. This detail is certainly the principal reason for the noticed differences between the left and right hands. In fact, the results show that the highest difference between the present results and those obtained by de Dear et al. (1997) corresponds to the left hand (1.2 W m 2 1C 1), whereas the right hands have the same value on both studies (4.1 W m 2 1C 1). The difference in the head is also significant since de Dear obtained an hconv value of 3.6 W m 2 1C 1 while in the present study the result was 4.3 W m 2 1C 1. In fact, the manikin used in de Dear's tests had a shoulder-length hair which can be expected to reduce dry heat losses from the head/neck region (de Dear et al., 1997). In contrast, the manikin used in the present tests had the head unprotected, i. e. without wig, thus more favourable to heat losses. It must be emphasized that details related to the experimental procedures can play an important role whenever comparisons are foreseen. For instance, the working section of the facility where de Dear's tests took place did not allow the thermal manikin to stand fully erect and the experiments were performed by slightly bending the manikin's legs at the knees. To enable both upper and lower leg segments to be measured in the vertical position, the experiments were repeated twice (de Dear et al., 1997). A remark is also due to the control methods of the manikins. In de Dear's tests the operation of the manikin is based on a constant
73
skin temperature control and in the present study the 16 parts of the thermal manikin are independently controlled according to the relation suggested by Fanger for dry heat losses within the PMV model (Fanger, 1972; ISO 7730:2005). Despite the mentioned differences, if the static values are depicted one by one, it is possible to find some similarities in the results. The graphics of Fig. 3 summarize all the tests, by presenting the hconv values of the 16 body parts for both static and dynamic conditions. It is important to point out that two different scales were adopted to make clear the differences that occur with lower velocities. Thus, the scale in the graphs up to 31 r.p.m. ( ffi 1 m s 1) is limited to 15 W m 2 1C 1 and the others to 45 W m 2 1C 1. This representation puts in evidence the differences between the 16 parts. For the central body parts the results are approximately equal, which means that the walking movements have a minor influence on these body parts. Otherwise, the dynamic experiments incorporate the pendulum effect into the hconv values and confirm the influence of walking movements on the limbs. The dynamic effect is enhanced for the lower air velocities and attenuated for the higher. Another interesting remark is related to the magnitude of the hconv values. In the experiments carried out up to 1 m/s, the hconv values are always lower than 15 W m 2 1C 1, but for the test at the higher air speed that limit is generally fixed at 30 W m 2 1C 1. The exceptions refer to the hands, where that limit is overcome. The static and dynamic results for the central body parts (chest, back and pelvic region) are almost the same which indicates that the walking movements at 45 steps/min have a negligible effect on hconv. Furthermore, the pelvic region, the chest and the back have the smallest values of hconv. Above 1 m/s the influence of the body
15 - Chest 50
40
40
hconv [W/m2.ºC]
hconv [W/m2.ºC]
8 - Head 50
30
hconv = 8.12va
20
0.49
10
Static
30
20
hconv = 6.10va0.52
10
Dynamic
0 0
2
4
6
8
0
2
4
7 - Pelvic region
8
10
16 - Back 50
40
40
hconv [W/m2.ºC]
hconv [W/m2.ºC]
6 va [m/s]
50
30
hconv = 5.99va0.50
10
Static
hconv = 5.43va
Dynamic
0
10
va [m/s]
20
Static
hconv = 5.90va0.55
hconv = 7.72va0.50
30
20
hconv = 5.66va0.55 10
Dynamic
Static
hconv = 5.33va0.59
0.54
0
Dynamic
0 0
2
4
6
8
va [m/s]
Fig. 4. Convective heat transfer coefficients – head and pelvic region.
10
0
2
4
6
8
va [m/s]
Fig. 5. Convective heat transfer coefficients – trunk.
10
74
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
movements on the convective heat transfer phenomena is not noticed. The peripheral parts of the body have larger convective coefficients. In fact, the points at a longer radius from the centre cover longer distances, and thus, have stronger heat losses. This is a common sense feeling which is clearly demonstrated here by measurements. An overview of the upper limbs puts in evidence the unexpected behaviour of the left hand. The observed asymmetry is due to the already referred position more favourable to heat losses. The highest values of hconv were obtained with the highest air velocity and correspond to this body part, both in the static (42.2 W m 2 1C 1) and dynamic (40.3 W m 2 1C 1) tests. The feet and legs have rather higher convective heat transfer coefficients than the thighs. If a comparison is made between the lower and upper limbs (hand–foot; arm–leg; upper arm–thigh), we conclude that the lower limbs have lower values of hconv.
Finally, despite the reported left hand asymmetry, precisely where the higher values of hconv were registered, there is a general symmetry between the left and right sides of the body. The convective heat transfer coefficients as a function of the air velocity are now presented in Figs. 4–7. For the central body parts (head, chest, back and crutch) the curves are similar which indicates that the walking movements have a small effect on hconv. This behaviour was also noticed during the natural convection tests. Furthermore, the pelvic region and back have the smallest values of hconv. In the lower and upper limbs the differences between the static and dynamic results become less evident in the upper arms and thighs. Thus, the body movements have a more pronounced effect on the parts with higher amplitude such as hands and feet. In fact, it is a well known phenomenon that the hands and feet are the preferred body parts for heat losses (de Dear et al., 1997; Oguro et al., 2002a, 2002b; Oliveira et al., 2012).
2 - Right foot 50
40
40
hconv [W/m 2.ºC]
hconv [W/m2.ºC]
1 - Left foot 50
30
hconv = 11.61va0.39 20
hconv = 9.61va0.51
10
Static
30
hconv = 11.88va0.41 20
hconv = 9.89va0.52
10
Dynamic
0
Static
0 0
2
4
6
8
10
0
2
4
va [m/s]
8
10
4 - Right Leg
50
50
40
40
hconv [W/m2.ºC]
hconv [W/m2.ºC]
6 va [m/s]
3 - Left Leg
30
hconv = 10.38va0.34 20
10
Static
hconv = 8.33va0.48
30
hconv = 10.34va0.35
20
10
Dynamic
0
hconv = 8.44va0.48
Static
Dynamic
0 0
2
4
6
8
10
0
2
4
va [m/s]
6
8
10
va [m/s]
6 - Right Thigh
5 - Left Thigh 50
50
40
40
hconv [W/m2.ºC]
hconv [W/m2.ºC]
Dynamic
30
hconv = 7.90va0.39 20
10
Static
30
hconv = 8.35va0.39 20
10
Dynamic
Static
Dynamic
hconv = 7.48va0.44
hconv = 7.23va0.43 0
0 0
2
4
6
8
10
0
2
va [m/s]
4
6 va [m/s]
Fig. 6. Convective heat transfer coefficients – lower limbs.
8
10
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
10 - Right hand
9 - Left hand 50
50
40
40
hconv = 15.50va0.43
hconv [W/m2.ºC]
hconv [W/m2.ºC]
75
30
20
hconv = 13.78va0.52
10
hconv = 13.37va0.41 30
20
10 Static
Dynamic
hconv = 11.60va0.50
0 2
4
6
8
10
0
2
4
va [m/s]
6
8
10
va [m/s]
11 - Left arm
12 - Right arm
50
50
40
40
hconv [W/m2.ºC]
hconv [W/m2.ºC]
Dynamic
0 0
30
hconv = 9.11va0.39 20
10
30
hconv = 9.89va0.40 20
10 Static
hconv = 8.01va0.46
Dynamic
Static
hconv = 8.58va0.44
0
Dynamic
0 0
2
4
6
8
10
0
2
4
va [m/s]
6
8
10
va [m/s]
14 - Right upper arm
13 - Left upper arm 50
50
40
40
hconv [W/m2.ºC]
hconv [W/m2.ºC]
Static
30
hconv = 8.80va0.49 20
10
30
hconv = 8.98va0.45 20
10 Static
hconv = 8.04va0.55
Dynamic
Static
hconv = 8.20va0.50
0
Dynamic
0 0
2
4
6
8
10
0
2
4
6
8
10
va [m/s]
va [m/s]
Fig. 7. Convective heat transfer coefficients – upper limbs.
The present study clearly demonstrates the reduction of the heat losses from the peripheral parts towards the centre of the body. In the central body parts the values of hconv that were obtained can be considered underestimated for use in humans because of the differences between manikins and human body motion. On the contrary, the results obtained in the extremities can be considered overestimated for humans since the swing effect achieved with the pneumatic mechanism is more prone to reproduce the motion of a military rather than the motion observed in typical daily routines. In fact, our natural movements are not that stiff. This warns for the careful use of equations proposed in the literature for the whole body and supports the fact that important errors can be committed if those correlations are assumed for the body parts. For instance, the present results highlight that the dynamic hconv value of the left hand doubles the one of the pelvic region and the corresponding difference for the static experiments almost reaches 2.5 times more.
5. Conclusion The present work is focused on the convective heat transfer coefficients for the whole body and for the body parts and provides an approach for further understanding of heat transfer phenomena around the human body, particularly when motion and air velocity are considered. These effects are analysed considering the standing static posture and a walking speed of 45 steps/min and free air stream velocities between 0 and about 10 m/s. Natural and forced convection were thus considered. In the case of forced convection, general equations for the static and dynamic conditions are presented for the whole body and for the body parts. The present study was carried out with the manikin operating under the thermal comfort regulation mode. This manikin control mode is not often considered, so the comparison between the present results and others from the literature is important to assess the differences provided by other control modes. From a
76
A.V.M. Oliveira et al. / J. Wind Eng. Ind. Aerodyn. 132 (2014) 66–76
practical point of view, it can be argued that the whole body results obtained with the thermal comfort regulation mode are within the most common range available in the literature. This is particularly true when lower air velocities are considered. In addition, when the walking effect is considered, it is frequently achieved with a mechanism that allows both arms and legs to move but with the manikin walking on the spot. This condition differs from human walking movements where displacement in place occurs. The convection counterpart of the heat transfer is thus obviously different in the simulated and really body motion. Finally, any comparison must be addressed with caution and analysis based only on the hconv values is not advised, particularly in forced convection, where the flow conditions have a very significant role and are very difficult to reproduce in different wind tunnels. References Anttonen, H., 2000. Interlaboratory trial of thermal manikin based on thermal insulation of cold protective clothing in accordance with ENV 342. In: Nilsson H., Holmér I. (Eds.). Proceedings of the 3I3M – 3rd International Meeting on Thermal Manikin Testing, National Institute for Working Life, Stockholm, Sweden, 12–13 October 1999, pp. 25–28. Anttonen, H., Niskanen, J., Meinander, H., Bartels, V., Kuklane, K., Reinertsen, R., Varieras, S., Sołtyński, K., 2004. Thermal manikin measurements—exact or not? Int. J. Occup. Saf. Ergon. 10 (3), 291–300. ASHRAE, 2001. ASHRAE Handbook: Fundamentals. Bü ttner, K.J.K., 1934. Die Wä rmeü bertragung durch Leitung und Konvektion, Verdunstung und Strahlung in Bioklimatologie und Meteorologie, Nr. 404. Springer, Berlin. Chang, S.K.W., Gonzalez, R.R., 1989. Analysis of articulated manikin based convective heat transfer during walking, Technical Report T11-89. United States Army Research Institute of Environmetal Medicine, United States Army, Natick, Massachusetts. Colin, J., Houdas, Y., 1967. Experimental determination of coefficient of heat exchanges by convection of human body. J. Appl. Physiol. 22, 31–38. de Dear, R.J., Arens, E., Hui, Z., Oguro, M., 1997. Convective and radiative heat transfer coefficients for individual human body segments. Int. J. Biometeorol. 40, 141–156. Defraeye, T., Blocken, B., Koninckx, E., Hespel, P., Carmeliet, J.E., 2011. Computational fluid dynamics analysis of drag and convective heat transfer of individual body segments for different cyclist positions. J. Biomech. 44, 1695–1701, http://dx. doi.org/10.1016/j.jbiomech.2011.03.035. Directive 2010/31/EU of the European Parliament and of the Council of 19 May 2010 on the energy performance of buildings. Eur-Lex Official Journal L:13–35. Du Bois, D., Du Bois, E.F., 1916. Clinical calorimetry: tenth paper a formula to estimate the approximate surface area if height and weight be known. Arch. Intern. Med. XVII, 863–871. Fanger, P.O., 1972. Thermal Comfort—Analysis and Applications in Environmental Engineering. McGraw-Hill, New York. Ferreira, A.D., Oliveira, R.A., 2009. Wind erosion of sand placed inside a rectangular box. J. Wind Eng. Ind. Aerodyn. 97, 1–10, http://dx.doi.org/10.1016/j. jweia.2008.09.001. Fukazawa, T., Ando, T., Ikeda, S., Yamaguchi, A., Holmér, I., Tochiara, Y., 2009. Convective heat transfer coefficient from baby smaller than from adult. In: Proceedings of the 13th International Conference on Environmental Ergonomics. ISBN:978-1-74128-178-1. Gagge, A.P., 1937. A new physiological variable associated with sensible and insensible perspiration. Am. J. Physiol.—Leg. Content 120, 277–287. Gagge, A.P., Nishi, Y., Nevins, R.G., 1976. The role of clothing in meeting FEA energy conservation guidelines. ASHRAE Trans. 82, 234. Gagge, A.P., Stolwijk, J.A.J., Saltin, B., 1969. Comfort and thermal sensations and associated physiological responses during exercise at various ambient temperatures. Environ. Res. 2, 209–229, http://dx.doi.org/10.1016/0013-9351(69) 90037-1. Havenith, G., Nilsson, H., 2004. Correction of clothing insulation for movement and wind effects: a meta-analysis. Eur. J. Appl. Physiol. 92, 636–640, http://dx.doi. org/10.1007/s00421-004-1113-6. Ichihara, M., Saitou, M., Tanabe, S., Nishimura, M., 1995. Measurement of convective heat transfer coefficient and radiative heat transfer coefficient of standing human body by using thermal manikin. In: Proceedings of the Annual Meeting of the Arquitectural Institute of Japan, pp. 379–380. International Organization for Standardization, ISO 7726, 1998. Ergonomics of the thermal environment—Instruments for Measuring Physical Quantities, International Standard. International Organization for Standardization (ISO), Geneva.
International Organization for Standardization, ISO 7730, 2005. Ergonomics of the Thermal Environment—Analytical Determination and Interpretation of Thermal Comfort Using Calculation of the PMV and PPD Indices and Local Thermal Comfort Criteria. International Standard, International Organization for Standardization (ISO); Geneva. Kerslake, D.M., 1972. The Stress of Hot Environment. Cambridge University Press, Cambridge. Kurazumi, Y., Tsuchikawa, T., Ishii, J., Fukagawa, K., Yamato, Y., Matsubara, N., 2008. Radiative and convective heat transfer coefficients of the human body in natural convection. Build. Environ. 43, 2142–2153, http://dx.doi.org/10.1016/j. buildenv.2007.12.012. Kuwabara, K., Mochida, T., Kondo, M., Matsunaga, K., 2001. Measurement of man's convective heat transfer coefficient by using a thermal manikin in the middle wind velocity region. J. Hum. Living Environ. 8, 27–32. Luo, N., Weng, W.G., Fu, M., Yang, J., Han, Z.Y., 2014. Experimental study of the effects of human movement on the convective heat transfer coefficient. Exp. Therm. Fluid Sci. 57, 40–56, http://dx.doi.org/10.1016/j.expthermflusci. 2014.04.001. Madsen, T.L., 1976. Description of thermal manikin for measuring thermal insulation values of clothing. Thermal Insulation Report no. 48. Missenard, F.A., 1973. Coefficient d´échange de chaleur du corps humain par convection en fonction de la position, de l´activité du sujet et de l´environment. Arch. Sci. Physiol. 27, 45–50. Mitchell, D., 1974. Convective Heat Transfer in Man and Other Animals. B. Publishing, London. Mochida, T., 1977. Convective and radiative heat transfer coefficients for the human body. Bull. Fac. Eng., Hokkaido Univ. 84, 1–11. Mochida, T., Nagano, K., Shimakura, K., Kuwabara, K., Nakatani, T., Matsunaga, K., 1999. Characteristics of convective heat transfer of thermal manikin exposed in air flow from front. J. Hum. Living Environ. 6, 98–103. Nielsen, M., Pedersen, L., 1953. Studies on the heat loss by radiation and convection from the clothed human body. Acta Physiol. Scand. 27, 272–294. Nilsson, H., 2004. Comfort Climate Evaluation with Thermal Manikin Methods and Computer Simulation Models (Ph.D. thesis). University of Gävle. Nilsson, H., Anttonen, H., Holmér, I., 2000. New algorithms for prediction of wind effects on cold protective clothing. In: Proceedings of the 1st European Conference on Protective Clothing. Arbete och Halsa 8, Stockholm, pp. 17–20. Nishi, Y., Gagge, A.P., 1970. Direct evaluation of convective heat transfer coefficient by naphthalene sublimation. J. Appl. Physiol. 29, 830–838 (cited by ASHRAE, 2001). Oguro, M., Arens, E., de Dear, R., Zhang, H., Katayama, T., 2002a. Convective heat transfer coefficients and clothing insulations for parts of the clothed human body under calm conditions. J. Archit., Plan. Environ. Eng. (Trans. AIJ) 561, 31–39. Oguro, M., Arens, E., de Dear, R., Zhang, H., Katayama, T., 2002b. Convective heat transfer coefficients and clothing insulations for parts of the clothed human body under airflow conditions. J. Archit., Plan. Environ. Eng. (Trans. AIJ) 561, 21–29. Oliveira, A.V.M., Gaspar, A.R., Quintela, D.A., 2008. Measurements of clothing insulation with a thermal manikin operating under the thermal comfort regulation mode. Comparative analysis of the calculation methods. Eur. J. Appl. Physiol. 104, 679–688, http://dx.doi.org/10.1007/s00421-008-0824-5. Oliveira, A.V.M., Gaspar, A.R., Quintela, D.A., 2011. Dynamic clothing insulation. Measurements with a thermal manikin operating under the Thermal comfort regulation mode. Appl. Ergon. 42, 890–899, http://dx.doi.org/10.1016/j. apergo.2011.02.005. Oliveira, A.V.M., Gaspar, A.R., Francisco, S.C., Quintela, D.A., 2012. Convective heat transfer from a nude body under calm conditions: assessment of the effects of walking with a thermal manikin. Int. J. Biometeorol. 56, 319–332, http://dx.doi. org/10.1007/s00484-011-0436-3. Omori, T., Yang, J.H., Kato, S., Murakami, S., 2004. Coupled simulation of convection and radiation on thermal environment around an accurately shaped human body. In: Proceedings of the RoomVent 2004, 9th International Conference on Air Distribution in Rooms. Ono, T., Murakami, S., Ooka, R., Omori, T., 2008. Numerical and experimental study on convective heat transfer of the human body in the outdoor environment. J. Wind Eng. Ind. Aerodyn. 96, 1719–1732, http://dx.doi.org/10.1016/j. jweia.2008.02.007. Parsons, K.C., 2003. Human Thermal Environments: The Effects of Hot, Moderate, and Cold Environments on Human Health, Comfort, and Performance (527 pp.). Quintela, D., Gaspar, A., Borges, C., 2004. Analysis of sensible heat exchanges from a thermal manikin. Eur. J. Appl. Physiol. 92, 663–668, http://dx.doi.org/10.1007/ s00421-004-1132-3. Seppänen, O., McNall, P.E., Munson, D.M., Sprague, C.H., 1972. Thermal insulating values for typical indoor clothing ensembles. ASHRAE Trans., 120–130. Silva, M.C.G., Coelho, J.A., 2002. Convection coefficients for the human body parts determined with a thermal manikin. In: Proceedings of 8th International Conference on Air Distribution in Rooms, ROOMVENT 2002. Winslow, C.E.A., Gagge, A.P., Herrington, L.P., 1939. The influence of air movement upon heat losses from the clothed human body. Am. J. Physiol.—Leg. Content 127, 505–518 (cited by Fanger, 1970).