Title Author(s) Citation Issue Date Effect of Natural Convection on Equivalent Heat Conductivity of a Vertical Layer Packed with Glass Wool Seki, Nobuhiro; Fukusako, Shoichiro; Inaba, Hideo 北海道大學工學部研究報告 = Bulletin of the Faculty of Engineering, Hokkaido University, 84: 59-65 1977-07-11 DOI Doc URL http://hdl.handle.net/2115/41422 Right Type bulletin (article) Additional Information File Information 84_59-66.pdf Instructions for use Hokkaido University Collection of Scholarly and Academic Papers : HUSCAP 北海道大学工学部研究報告 Bulletin of the Faculty of Engineering, 第84・.暦 G1召禾日52名三) Kokkaido University, No. 84 (1977) Effect ef Natural Convection on Equivalent Heat Conductivity of a Vertical Layer Packed with Glass Wool Nobuhiro SEKI”“, Shoichiro FuKusAKoS’:’ an(至Hideo至NABA聾 (Received December 28, 1976) Abstract This report deals with the effect of natural convectioR on heat transfer in a vertical layer packed with glass wool of heat insulating material. Experiments are carried out with air or water as a saturating f}uid in the range of the specific weight of the glass wool, r, from 3 to 50 (kg/m3) and the aspect−ratio HIW from 5 to 47.5. The effect of the dimensions of cavities, the specific weight of the glass wool r and the surface temperature of the cold wall T, on the apparent heat conductivity Z,ff is investigated. The results of the experiments show a strong influence of natural convection on 2,ff in the range of T=3tv20 (kg/m3). It is ciarified that the results for water show a very complicated behavior as compared with those for air. 1. lntroductioR The apparent heat conductivity R,ff for glass wool is generally used as O.036−v・ O.057 (kcal/mhOC) in a range of r from 10 to 96 (kg/m3) by the Standard Method of Test of the JIS. However, these values can be adopted only j,n the case of heat conductive conditions, for example, where the heat insulating materials are used horizontally and the heating surface is on the upper side and the condensation of water or ice formation does not exist on the cooling surface. In case of other arrangements, especiaily vertical arrangement, the influence of natural convection on heat transfer between two vertical walls should be larger. This study investigates experimentally the effect of natural convection which is induced by the temperature difference between the hot and the cold wall surfaces on the increase of R.ff according to each r. Experimental conditions are selected with consideration to practical usage, that is, the present experiments are carried out in case of the yertical layer of glass wooi saturated with air or water. Moreover, the effect of the deyice dimensions on 2,ff is examined. ’ve 1)epartment of Mechanical Engineering, II, Hokkaiclo University, Sapporo 060, Japan 60 2 Nobuhiro SEKI, Shoichiro FUKUSAKO and Hideo INABA 2. Nomenelature Heat且ux, Inp厩(kca1)/Heating surface(恥2) q Height of cavity, (m) Widtli of cavity, (m) 研/躍 Aspect ratio Th Surface temperature of the hot wali, (OC) Surface temperature of the cold wall, (OC) Tc Apparent heat conductivity, //一丁//,=V−V一一丁・;・〉…, (Kcal/mhOC) Reff Specific we,ight of glass wool heat lnsulating material, (kg/m3) r 3. Experimental devices and Test procedure Experimentai device is depicted iR Fig. 1. The maiR parts of experimental device consist of the test section, heating and cooling parts. The test section is shown in the center part on the left side in Fig. 1. The samples of glass wool with prescribed r are packed in this section. In order to maintain each wall−surface tempature uniformly, the hot wall and the cold wall are constructed by copper plates (5 mm in thickness). Heating of the hot wall is performed by using the main heaters. The guard heaters are mounted on, the main heaters in order to minimize the heat from the main heaters to the enviroRment. Surface temperature of the cold wall is uniformly maintained by using devided cooling chambers. In order to investigate the infiuence of tke dimensions of the cavities on heat transfer, the divices are selected as shown in table 1. Surface temperature of the hot wall is uniformly kept at 20 (OC) in all of the runs, and surface temperature of the cold wall is changed from 10 (OC) to −15 (OC). Test samples of the glass wool which passed the Standard of the JIS 9505 are used. The fibrous direction of glass wool is arranged to be perpendicular to the heat fiow. Heat transfered is measured after a steady state iR the test section is reached thermally and hydrodyRarnicaily. ln each of the runs, it requires about 2・hvS hrs to obtain the steady state. 、認 騰,, 蝋,, PPI@ OOouHG P]PE IHLET 2 ^田 4 B蹉, 2 600 H/W 15 20 950 AMABATIC MNH T (塑.、主 300 P 1 曝h W (m無). 30 47.5 TEST 5ECTrOH 製…, 300 伽騰、、R !0 30 50 6 c灘:,R r ?..2 e鵬, 明 Q ・田E㎜㏄u旺5 ①MAIN限正R @ouAnb HEATER @ COPPEn maATt ④臨EUTεPしAIE ⑤IRAVER51瞬G ⑥L川P THERMoc。曜 Fig. 1 Schematic diagram of the experimental devices 571 59 116 26 9.7 5 Table 1 The dimensions of the cav三t三es used Effect of Natural Convection on Equivalent Heat Conductivity of ic Vertical Layer Pacl〈ecl with Glass Woo1 3 61 4. Experimental results and Discussion 4.1 Tke effect of the dimensions of the eavity on leff It is impossible technically to enlage the height of cavity and the width of glass wool infinitively. Therefore, the dimensions of the cavities are limited in the present study. Fig. 2 shows the relation between va and R.ff for if=571 (mm), Th=20 (OC) and T,== 一5 (Oc). When ai.r as a saturating舳id is used as shown in Fig.2(a), the vaiue of R,ff increases in a curve fashion with increasing W when 1711 {Ell 20 (mm). But when Vll>20 (mm), the value of R,ff increases lineariy with the increasing 17V. On the other hand, when water is used and glass wool is packed as shown in Fig. 2 (b), the value of 2,ff increases monotonously over the entire range of Vl and the rate of increase of R,ff is very small. Fig.3shows the re正ation between H andλe〃for躍・= 20(mm), T, =: 20(QC) and T,==一5 (OC). The value of R,ff without glass wool decreases sharpiy with the increasing H when U$1 (m). But when U>1 (m), the vaiue of R,ff decreases at a much slower pace. From Fig, 3, it is understood that these tendencies have the same 15 1.5 FLuld: Water Ftuid : Air H=O.571(m) H=O.571{m} Th=20’C,Tc=一5.C Th= 20・C , Tc =一5’C 3 e e as = o {Kgtm3) Z1,0 e v = o{Kgtm3) .U一” P0 . 10 董 e 30 妻 ξ e 20 石 受 篭 0 20 0 30 /o 一一/ 4多ノ夢ク多ン ゆ rくq5 . 10 V?”t5 o/ o / / o / e −o/ ●F瞬一_o一一一’一一一一 げ二.:d 毎島_一−i一一一一一一一〇〇一8 o o / O,05 O.10 w(m) o 0.10 WCm) Fig. 2 (a) Fig. 2 (b) Fig. 2 Effect of W on 2,が,(:rh==20℃,:r,=一5。C) O,15 O.20 ’ n O.05 . e o O.15 O.20 (a) H==571mm, Air, (b) H= 57!mm, Water Ftuid: Air O.1 5 N zo W=20(mm) ・こ\.剛C口圃C 了h=20●C,丁c=一59C しし亀 渥ミミ三三≧=:一…(Kg…3) a ・t o,io E 5 モ 乏Q。5 Xe δ k 1.5 \.一×:こ19 受 Ftuid:Water W=20(mrn) 30 × ξ oXo 石 Xo 望 Xo 蚤 ゆ ・く1,0 ’ XV;O{Kgtm3) ’x一 ×. . ’XL一 ”x. ’ 黶Fこ:一==:一;∼;∼ ≧§ ttt Q5 o o 0.5 1,0 H(m) Fig. 3 (a) 1.5 2・O O O、5 1.0 Fig. 3 Ef[ect of ff on 2βプ〆,(Th=20。C, Tc=一50C) ほニ ラ Fig。3 (b) (a)W x 9.0 mrn, Air,(b)W ・2e mm, Water i.5 zo 62 4 Nobuhiro SEKI, Shoichiro FUKUSAKO and Hideo INABA characteristics as in the experimeRtal results with glass wooi. As mentioned above, it is clear that the increase of H results in the decrease of λ、が which represents the apparent mean heat conductivity of the packed layer. This result is due to the fact the natural conyective heat transfer caused by the movement of a saturating fiuid has a significant role in the }owest part of the hot wall and in the highest part of the cold wall. 4.2 [E]he effeet of the specific weight r on 2eff Fig. 4 illustrates the effect of the specific weight r on R.ff. It may be understood that R.ff decreases sharply as r increases in the range of rx=O−v20 (kg/rn3). This result means that the infiuence of natural convection on heat transfer is predominant in this range. lf r>20 (kg/m3), R,ff decreases at as正ow rate due to the increasing resistance of the movement of the saturating fiuid. It can be seen in Fig 4 (a) and Fig. 4 (b) that a decreasing rate of R,ff for water having a large viscosity is larger than that for air. Fluid :Wqter 8 F[uidiAir Th・20。C,T、・一50C Th=200C , Tc=一sec o Wニ B W= l wW. 1, ig26 {rT]:M) ] H=s7i c rn rn ) 05 :;圃 l H=571(rnnり) ニW;1笥6 e × 。 W= り 20 H堅300(mm) oW= 50 x. ε o W= 10(mm) 6 i ttliiE‘nm’] ..,,,.., ‘0.3 s 54 e. e ε e こ o 『 ∼o 呈 9 5 ヨ K. 2 蓉 ● 、 b 卜。\\一._く. 2 賦こ・_._ O,1 \. .×. se.一e 一㌶転㌃= o θ 一_ρ_ O 10 20 1{Kg301.3) 40 50 …gO ’O 10 20 f(Kg3iO.3) 40 50 60 Fig.4 (a) Fig.4 (b) Fig. 4 Effect of r on 2eff, (Th=20 eC, Tc= 一50C) (a) Air, (b) Water 4.3 The effect of the cold wall surface−temperature Tc on leff In order to investigate the intensity of natural convection caused by the temperature dfference between the surface temperatures of the hot and the cold walls, experiments are carried out under the condition that the hot wall surface temperature Th is maintained at 20 (OC) and the cold wail surface temperature T, is changed variously in a range of T,=一15−vlO (OC). Experimental results are shown in Fig. 5. It can be understood in Fig. 5 (a)一v5 (c) for air that value of R,ff decreases linearly as T. increases. On the other hand, experimenta} results for water as shown in Fig. 5 (d)一v・5 (f) show a very complicated behavior as compared with those for air, that is, in the range of T,〈O(OC), it is clear that the heat transfer Effect of Natural Convection on Equivalent Heat Conductivity of a 5 63 Vertical Layer Packed with Glass Woo1 O.1 5 H=300(mm),W=20(rnm) H=300(mrn),W=20(rn rn} Th=20ec Th=20’C P 1.0 A9 。 ・“一一一6一一一一一。一 一SL一一SL一一一一一一一一cr一一’一“一’” ‘0.10 ∈ t’ 葦 2一§≡幕≡ 一.i:1一一v一 こ σ り む ’・ノ 壱 2 董 × 乏α5 紅 w\。叉。__ .,・ハ・\ ’ ●∼o∼●_ 一’e。∼。一〇一一一。_ e pt =10(Kglm3) Φ rくO.05 o . 20 . e 30 Ftuid:Air o 10 0 40 o o T,(oc> 一10 o Fiuid;Water 10 V =10 〈 Kgt m3) 20 30 o Tc(oc) 一20 一10 Fig.5 (d) F隻9.5 (a) 1,5 H= 571(m rn ), W= 59(m m) Th=20ec O.20 //ノ H;571(mm),W=59〈mm> 8 Th=20ec 1e 1.0 AYals = 套 0− a_o一一一一ず一一一 so一ず一一一一 _●一●一 /Q騨__●一・’● ,・/’一●、● 量 ヲ ノ 銭ム==・;ニコ=ご二二;= E U 2 /’ _気一一._._ 孕二ここiiこゴ≒_ モ こ 乏α5 o 1 =lo(Kg/m3) モ 詑q1・ 1 =10 (Kg t rn3) . 20 20 30 40 Fluid:Air e 30 0 40 FEuidl Water o 10 o 一IO 丁c(℃) Fig. 5 (e) O.05 10 o Tc〈 ec > 一10 一20 H=571(ram),W=116(mm) Fig. 5 (b) Th=200c fOXe一.一..一一一.一〇一一9一一一 1,5 O,35 H=571(rnm),W=116(mrn) / Th= 20ec 〇 o/ o/O oXU’e ’t’=r’一・’一’/’一一 8 , Q30 E こ g 二つ;二・一一一一 E ti v¢ べQ25 o Ftuid:Air 1 =10{Kgirn3) . 20 e 30 o 40 . 1.0 = / o.一e−O o−e/o 溝 Q20 e こ ”o ES AU . . 一sses e一 . イ六_・ Uり s 一”ro””=e一一。一 e rく0・5 п[一_。 。_ o Ftuid:Water i = tO(Kglm3) . 20 e 30 e 40 o 10 O 一10 一20 10 0 Tc (oc) Tc ( oc ) Fig.5 (c) Fig.5 (f) Fig. 5 Effect of Tc on z:eff, (Tn=20eC) (a) Pli==2e mm, H=300 mm, Air, (d) ViF’=20 mm, El=:300 mm, Water (b) W=59mm, H= 571 mm, Air, (e) W=59 mm, H’ :571 mm, Water (c) rv=116 mm, H=571 mm, Air, (f) Mf =116 mm, H’= 571 mm, Water 一10 64 6 Nobuhiro SEK工, Shoichiro FUK:USAKO and Hideo玉NABA between two vertical walls is affected by the ice formation on the cold wall surface and the density iRversion effect of water which has its maxim density at 4 (OC). When T, becomes larger than 4 (OC), tlte effect of density inversion disappears, whi’ie 2,ff reaches its maximum vaiue at about 5 (OC). However, the maximum value of R,ff becomes relatively smali due to the resistance of the movement of saturating fluid with the increasing r. 4.4 The effect of the relative humidity of moist air on PLeff Fig. 6 shows the relation between the relative humidity of moist air on R.ff. When moist air exists in packed layers, it is generally said that the circulation of vapor diffusion, that is, diffusion in the vicinity of the hot wall−condensation or ice formation on the cold wali−the movement of water to the hot wall by capillary action 一 evaporation in the vicinity of the hot wall, is formed in a testing layer. This movement of moisture promotes the effect of natural convection on heat transfer. It shoud be understood from the reason mentioned above that the values of λ,がfor 100%relative humidity are larger than those for 50%relative humidity. Retative Humidity of MoistAir O.1 5 Retative Humidity of Moist Air − 100。ん 。 − 1000/o o O.1 5 一一一 50『ノ。 o “一一 as=10(Kg’[あ ノー o−o P O.10 軽 1=10 づ:灘;;巽ま嶋≡ 1・10(㎏’m3) む o一 1=10 9 £O.10 ε !’ 壱 こ き 芝 H=300(mm),W=20(rnrn) Th=20ec __e一__一._一q__ e一一一 一θ一r一一♪曹一’謄國’ .4多。 u=30 r一一ree”e o_d2 0一 の ロ ロ コロ ロ ロ の ロむロ リ ロ ロ コ /8tS 五=30 / と 駕 rくQO 5 o 10 む 〃“ む 等・・5 @50 efe e H”600(rnm>,W=20(mm) Th=200c o T,(ec)O 一10 一一20 10 T,〈oc)O 一10 一20 Fig.6 (a) Fig.6 (b) Fig. 6 Effect of the relative humidity of moist air on 2eff (a) W =: 20 mm, H== 300 mm, (b) W==20 mm, H= 600 mm 5. Summary and Conclusions It can be clarified that the heat conductivity of 2,ff=O.036ivO.057 (kcal/mhOC) generally used is not always applicabie when the natural convection exists in the layer of glass wool: (1) The infiuence of natura} convection on 7,,ff is remarkable in the range of r =3−v20 (kg/m3). (2) For air, the relation between T, and 7,,ff is liRear. On the other haRd, expeyimental results for water show a very complicated behavior having a maximum vaiue of ?,,ff at about T iz 5 (OC). (3) For moist air as a saturating fiuid, the movement of water promotes the effect of natural convection on the apparent heat conductivity between two vertical walls. 7 ’E’ffect of ’Natural Convection on Equivaient 1,leat Conductivity of a Vertical Layer 1?ackecl with Glass Woo1 References 1) Lorentzen. G. and Brendeng. E.: Proceeding of the 10th Refrigeration, 2 (1959), p. 294. 2) Achtziger. J.: Kalte Technic, 12 (1960), p. 372. 3) ZeheRder. H.: Kalte Technic, 11 (1964), p. 308. Internationa1 Congress of 60r