optimization of flow direction to minimize …m. s. abd-elhady et al. 897 figure 2 the different...

Heat Transfer Engineering, 30(10–11):895–902, 2009Copyright C©© Taylor and Francis Group, LLCISSN: 0145-7632 print / 1521-0537 onlineDOI: 10.1080/01457630902754142

Optimization of Flow Directionto Minimize Particulate Foulingof Heat Exchangers

M. S. ABD-ELHADY,1 C. C. M. RINDT,2 and A. A. VAN STEENHOVEN2

1Department of Mechanical Engineering, Beni-Suief University, Beni-Suief, Egypt2Department of Mechanical Engineering, Eindhoven University of Technology, Eindhoven, The Netherlands

The influence of flow direction with respect to gravity on particulate fouling of heat exchangers is investigated experimentallyto determine the optimal flow direction to minimize fouling. Four orientations of flow have been investigated: horizontal flow,upward flow, downward flow, and a flow under an angle of 45◦. It is observed that fouling starts at the point of stagnationirrespective of the flow direction, and also at the top of the heat exchanger tubes. Particulate fouling grows from these twopoints till they meet and the fouling layer covers the whole surface of the heat exchanger tube. Fouling at the upper half of thetubes is much faster than the lower half of the tubes, and the fouling rate is faster at the bottom tubes of the heat exchangersection than at the upper tubes. The best orientation for lingering particulate fouling is the downward flow, where the flowstagnation point coincides with the top point of the heat exchanger tubes and the growth of the fouling layer only starts fromone point.

INTRODUCTION

Particulate fouling is defined as the accumulation of parti-cles on a heat transfer surface that form an insulating layer,which reduces the rate of heat transfer and can lead to opera-tion failure as has been reported by many researchers, e.g., inwaste incinerators by van Beek et al. [1], in a coal-fired powerplant by Bryers [2] and in utility scale boilers by Gupta et al.[3]. Transport and capture of particles on heat transfer surfacesare functions of the size of particles, the chemical and physi-cal properties of the transported particles, and the combustionsystem conditions, such as gas flow patterns, velocity, localtemperature, etc. (Benson et al. [4]). A theoretical backgroundabout particulate fouling and especially gas side fouling canbe found in literature of Epstein [5, 6] and Marner [7]. Manyefforts have been made by numerous researchers to understandand control fouling in heat exchangers, e.g., Taborek et al. [8],Muller-Steinhagen et al. [9], Bohnet [10], and Bott [11]. Hupa[12] has performed fouling experiments using mixtures of ricehusk and eucalyptus bark as fuels. A mixture of the two fuels

Address correspondence to Dr. M.S. Abd-Elhady, Department of Mechan-ical Engineering, Beni-Suief University, New Beni-Suief, Beni-Suief, Egypt.E-mail: [email protected]

was burned in a 9-m-tall entrained flow reactor and the fly ashproduced was captured by an air-cooled probe at the bottomof the reactor. Hupa [12] found that the rice husk ashes do notstick to the test probe surface because of their large size, i.e.,several hundreds of micrometerss, while the bark ash, which hasa particle size in the order of micrometers, very readily stickson the surface and forms a gradually growing deposit on thewindward side of the probe, i.e., upstream of the flow. However,the results of Hupa [12] differ from the work of Kaiser et al.[13] and Abd-Elhady et al. [14], who studied the deposition ofparticles on cylindrical tubes and found that the formation offoulant is mostly at the rear part of the cylindrical probes, i.e.,downstream side of the tubes. The direction of flow in case ofHupa’s experiments [12] was downward, while in the case ofKaiser et al. [13] and Abd-Elhady et al. [14] it was upward,which suggests that the flow direction with respect to gravitymight influence the position at which particulate fouling starts.

Kalisz and Pronobis [15] found that particulate fouling in theeconomizer of a pulverized coal boiler occurred at the down-stream side of the tubes, while at the end of the process a littlewedge at the upstream side was observed. In the measurementsperformed by Kalisz and Pronobis [15] the upstream side ofthe tubes was the lower surface of the tubes, i.e., the flow wasupward. Nuntaphan and Kiatsiriroat [16] investigated the effect

895

Downloaded By: [Technische Universiteit - Eindhoven] At: 13:40 8 April 2009

896 M. S. ABD-ELHADY ET AL.

of fly ash deposit on the thermal performance of a cross-flowheat exchanger having a set of spiral-finned tubes as heat trans-fer surface (Nuntaphan et al. [17, 18]). The tubes of the heatexchanger were vertically mounted in a staggered array, andthe flow was horizontal. Nuntaphan and Kiatsiriroat [16] foundthat the deposition of fly ash occurred at the frontal part of thetubes following the direction of flow, i.e., upstream of the flow.It can be concluded from the presented literature survey thatthe direction of flow influences the position at which particu-late fouling starts as well as the fouling process. However, therelation between the flow direction and the position at whichfouling starts is not well understood. In the research presentedhere the influence of the flow direction with respect to gravityon particulate fouling of heat exchangers is investigated exper-imentally to study this relation and to determine the best floworientation to minimize fouling.

EXPERIMENTAL SETUP AND THE EXPERIMENTALPROCEDURE

An experimental setup has been built to study the effect offlow direction with respect to gravity on particulate fouling ofheat exchangers. A schematic of the experimental setup is shownin Figure 1. The experimental setup can be rotated 360◦ aroundits horizontal axis, i.e., the axis of tube 5, to get different orien-tations of flows, such as vertical flow, horizontal flow, and a flowat an angle to the direction of gravity. The experimental setupconsists of an air blower connected to a wooden duct of size40 cm × 40 cm × 400 cm. The power of the air blower is 1.5kW at 1450 rpm, and delivers 0.25 m3/s of air at normal operat-ing conditions. The wooden duct contains a heat exchanger with9 tubes arranged in an inline matrix, i.e., 3 × 3 tubes, as shown inFigure 1. The tubes have an external diameter of 3.2 cm and are

made of steel. The tubes are numbered from 1 to 9, so they can beeasily distinguished during the fouling experiments (see Figure1). There is no secondary fluid in the heat exchanger tubes, i.e.,the surface of the heat exchanger tubes is isothermal. A screwparticle feeder is used to dispatch calcium carbonate particlesin the airflow. By varying the rotational speed of the screw ofthe particle feeder, the injection rate can be varied. The deposi-tion of particles on the heat exchanger tubes can be monitoredthrough two glass windows mounted on the sides of the woodenduct as shown in Figure 1. Two inspection doors are installed atthe sides of the wooden duct for maintenance and for cleaningthe setup from the fouling particles. The speed of the air is mea-sured at the exit of the wooden duct by a hot wire anemometerwith an accuracy of ±0.01 m/s. Air exhaust from the setup isfiltered through a bag filter before it goes into the atmosphere.

In the experiments, ambient air is fed by the air blower intothe duct, and particles are added by a particle screw feeder.Calcium carbonate particles of average diameter of 40 µm witha standard deviation of ±16 µm are used in the experiments as asource of fouling. In the experiments the volume flow rate of airand the injection rate of particles are kept constant at 0.2 m3/sand 20 g/m3 of air, respectively. Fouling of the tubes of the heatexchanger occurs as the air–particle mixture passes the heatexchanger. Fouling of the heat exchanger tubes is monitoredusing a digital camera through the installed glass windows, asshown in Figure 1. The digital camera is a Finepix S602 zoom,with a resolution of 2832 pixels × 2128 pixels, i.e., 6.03 millionpixels, and optical zoom of 6×. The fouling layer thickness ismeasured optically. Images of the heat exchanger tubes are takenbefore and during fouling. The clean heat exchanger tube radiusis taken as a reference. The tube radii in the camera imagesand hence the fouling layer thicknesses are determined with anaccuracy of ±1 pixel, which corresponds to ±0.04 mm.

Four experiments are performed with different flow orien-tations, i.e., horizontal flow, flow at 45◦, upward flow, and

× ×

Figure 1 The experimental setup used to study the effect of flow direction on particulate fouling of heat exchangers.

heat transfer engineering vol. 30 nos. 10–11 2009


M. S. ABD-ELHADY ET AL. 897

Figure 2 The different orientations of the experimental setup to give (a)horizontal flow, (b) flow at 45◦, (c) upward flow, and (d) downward flow; g isthe direction of gravity. Point A is at the upstream area of the heat exchanger,which is the same in all orientations.

downward flow, to investigate the influence of flow directionwith respect to gravity on particulate fouling of heat exchang-ers. In each experiment, the whole experimental setup is rotatedas shown in Figure 2 to give the required flow orientation.

EXPERIMENTAL RESULTS

Horizontal Flow

The growth of the fouling layer as a function of time forthe horizontal flow case is shown in Figure 3. The pictures arefor tubes 5 and 8 in the second column of tubes of the heatexchanger section shown in Figure 1. The flow is horizontal andpassing into the plane of the page as indicated in Figure 3b. Thetubes at the beginning of the fouling experiment, i.e., time = 0,is shown in Figure 3a. The gray color corresponds to the cleansteel surface of the tubes. Points A and B indicated in Figure3a correspond to the stagnation point (Achenbach [19] and Lamet al. [20]) and the top of the heat exchanger tube 8, and the lineof stagnation is indicated in Figure 3b. The tubes after 10 and30 min of operation are shown in Figures 3b and c, the whitecolor corresponds to the fouling layer deposited on the tubes.The fouled part of the tubes shown in Figure 3 is the frontal partof the tubes in the upstream. It can be concluded from Figure 3,b and c, that fouling starts at two points, the point of stagnation,i.e., point A, and the top of the heat exchanger tube, i.e., pointB, and the growth of the fouling layer continues from thesetwo points, as can be seen from the preceding pictures. Thisobservation is in agreement with the findings of Nuntaphan andKiatsiriroat [16]. Fouling at the upper half of the tubes is muchfaster than at the lower half of the tubes, because particles thatstick at the upper half of the tubes can not fall off by gravitydue to the obstruction of the tube surface; however, particlesthat stick to the lower half can fall off due to gravity. It isalso found that the fouling rate is faster at the bottom tubesof the heat exchanger section, i.e., tubes 7, 8, and 9, than atthe upper tubes, i.e., tubes 1, 2, and 3. This is attributed to theincreased concentration of particles in the lower portion of theheat exchanger caused by the falling off particles from the uppertubes.

Figure 3 Fouling of the heat exchanger tubes 5 and 8 as a function of time.The position of tubes 5 and 8 in the heat exchanger section is indicated inFigure 1. The airflow is horizontal, and only the upper quarter of the tubes inthe upstream is shown. Points A and B indicated in Figure 3a correspond to thestagnation point and the top of heat exchanger tube 8.

Flow at 45◦

The growth of the fouling layer as a function of time in caseof a 45◦ flow is shown in Figure 4. The tube shown is tubenumber 5. The flow is coming out of the plane of the page asindicated in Figure 4b. It can be seen that fouling starts at the topof the tube, which is the area bounded by points B and C. Thegrowth of the fouling layer over the heat exchanger tube withrespect to flow and gravity is illustrated in Figure 5. Points Band C in Figure 4a correspond to points B and C in Figure 5. It isalso found that fouling starts at the stagnation area of flow, i.e.,point A presented in Figure 5. The growth of the fouling layerover the upper half of the tubes, i.e., area bounded by the curveA–B–C as indicated in Figure 5, is much faster than the foulingrate at the lower surface area of the tubes, i.e., area bounded bythe curve A–D–C. The average thickness of the fouling layerat the upper half of the tubes after 3 h of operation was 1 mm;however, the thickness of the fouling layer at the lower surfaceof the tubes was very small, i.e., less than 0.05 mm.

Vertical Upward Flow

The growth of the fouling layer as a function of time in caseof an upward flow is shown in Figure 6. The top of the heatexchanger tubes is shown in Figures 6a and b, while the bottom




Figure 4 The growth of the fouling layer as a function of time in the case of a 45◦ flow. The presented area is the rear end of tube number 5, and points B and Cin Figure 4a correspond to points B and C in Figure 5.

is shown in Figure 6c. The top and the bottom of tube 4 after 30min of operation are shown in Figures 6b and c, respectively. Itis found that fouling starts at two places, which are the top andthe bottom of the tubes. This observation agrees with the workof Kalisz and Pronobis [15], Kaiser et al. [13], and Abd-Elhadyet al. [14]. The bottom of the heat exchanger tubes in case ofupward flow is the area of stagnation. The fouling layer growsfrom the top as well as from the bottom of the tubes until thewhole surface of the tubes is fouled. The fouling layer growthat the top is faster than at the bottom, and the growth is moreor less symmetrical around the vertical axis of the tubes. Thefouling layer in the case of a horizontal flow covers a quarter ofthe heat exchanger tube after 3 h of operation, in the case of a45 degree flow covers the upper half of the tube, and in the case

Figure 5 Areas at the tube at which fouling starts in the case of a flow at 45degrees to the direction of gravity.

of an upward flow covers the whole tube. It can be concludedthat the flow direction with respect to gravity affects stronglythe growth of particulate fouling layers over the tubes of heatexchangers.

Vertical Downward Flow

The growth of the fouling layer in case of a vertical downwardflow is shown in Figure 7. The pictures were taken during 2 hof operation. It was seen that the growth of the fouling layerwas symmetrical around the vertical axes of the tubes. It canbe seen from Figure 7 that the bottom of the tube is clean anddid not foul during the experiment; however, the top of the tubehas fouled from the beginning of the experiment and continuedfouling till the end. It can be concluded that fouling starts onlyat one point, which is the top of the heat exchanger tube, andat which the fouling layer continues its circumferential growth.This conclusion is in agreement with the observations of Hupa[12]. The top of the heat exchanger tube in the case of a verticaldownward flow coincides with the stagnation area of flow.

DISCUSION OF RESULTS

The growth of the fouling layer as a function of the flowdirection with respect to gravity is summarized in Figure 8.Points A and B indicated in Figure 3a correspond to points Aand B in Figure 8a, and points B and C indicated in Figure 4acorresponds to points B and C in Figure 8b. It is found thatin all orientations of the flow, fouling starts at the stagnationarea of flow around the heat exchanger tube and at the top ofthe tube. The top of the heat exchanger tube in the case of a




Figure 6 Growth of the fouling layer over the heat exchanger tubes in caseof upward flow as a function of time; g is the direction of gravity. (a, b) Topof the heat exchanger tube; (c) bottom of the heat exchanger tube. The whitesurface is the fouling layer, while the black part is the tube surface.

vertical downward flow coincides with the stagnation area offlow, which is the main reason for having only one position atwhich fouling starts. The reasons for fouling to start at the topof the heat exchanger tubes and at the stagnation area of floware discussed in the next paragraphs.

Submicrometer particles are transported by large particles,usually dp > 5 µm, or vapor diffusion, whereas the intermediate-sized particles, i.e., dp < 5 µm, can be transported by a ther-mophoretic or an electrophoretic force (Raask [21]), and thelarge particles mainly by inertial impaction. Inertial impactiontakes place when large particles have too large an inertial mo-mentum to follow the gas stream lines around the heat transfersurface, and instead impact on the surface. Inertial impaction ac-counts for the bulk of deposit growth in the presented research.The temperature difference between the gas side and the surfaceof the heat exchanger tubes affects the transport of submicrom-eter and intermediate-sized particles but hardly influences thetransport of large particles (De Best [22]). Despite the fact that

Figure 7 Growth of the fouling layer over the heat exchanger tubes in thecase of a downward flow as a function of time; g is the direction of grav-ity. The white surface is the fouling layer while the black part is the tubesurface.

there is no secondary fluid in the heat exchanger tubes and thetemperature difference between the gas side and the tube surfaceis zero, the performed experiments can still explain the onsetof large particulates on the tubes of heat exchangers and theposition at which fouling starts. Particles stick to the heat ex-changer tube if their speed is lower than or equal to the criticalsticking velocity (Abd-Elhady et al. [23]), which is defined asthe maximum impact speed at which a particle hitting a surfacewill stick and does not rebound. The critical sticking speed Vs

for the used calcium carbonate particles hitting a steel tube is




Figure 8 Areas at the tubes at which fouling starts as a function of flow direction, i.e., (a) horizontal flow, (b) flow at 45 degrees, (c) upward flow, and(d) downward flow; g is the direction of gravity.

equal to 0.02 m/s, and it is calculated from Thornton and Ning[24] by

Vs = 1.84

[(�/R)5

ρ3E∗2

] 16

(1)

where � is the surface energy between the interacting bodies,R the radius of the particle, ρ the density of the particle, andE* the reduced Young’s modulus, which is calculated fromAbd-Elhady et al. [25] by

E∗ =(

1 − ν21

E1+ 1 − ν2

2

E2

)−1

(2)

with the subscripts 1 and 2 representing the interacting particleand the tube surface. E is the Young’s modulus of elasticity andν the Poisson’s ratio (Fenner [26]). The material properties ofthe calcium carbonate and the steel used to calculate the criti-cal sticking velocity are shown in Table 1 [27, 28]. The criticalsticking velocity for the calcium carbonate particles is very lowcompared to the main stream velocity, i.e., 1.25 m/s, and there-fore deposition of particles on the tubes surface occurs at areaswhere the speed of particles is low, such as at the area of sepa-ration of flow and at the stagnation area of flow. If the particlehas too much kinetic energy in excess, it can also bounce off theheat transfer surface and become entrained in the flue gas streamas observed by Benson et al. [4], Raask [21], and Baxter [29].

Separation of flow occurs toward the top of the heat ex-changer tube as observed by Sons and Hanratty [30] and Wu

Table 1 Physical properties of calcium carbonate and steel(Gilman [27] and Rogers and Reed [28])

Calcium carbonate Steel

Young’s modulus, E (N/m2) 0.35×1011 2.15×1011

Density, ρ (kg/m3) 2830 7800Poisson’s ratio ν 0.27 0.28Surface energy � (J/m2), 0.29

calcium carbonate–steel

et al. [31] in the case of a horizontal flow, upward flow, and aflow at 45 degrees, and the separation angle is a function of theReynolds number of the flow around the tube. Based on the flowconditions, i.e., main stream velocity of 1.25 m/s and air tem-perature 25◦C, the Reynolds number of the flow for the tubesof the heat exchanger is equal to 3.2 × 104, which correspondsto a separation angle of 80◦ from the stagnation point accordingto Sons and Hanratty [30]. Due to separation of the flow at thetop of the heat exchanger tube, both the removal of particles byshear flow decreases as found by Cabrejos and Klinzing [32]and Halow [33], and the speed of the particles decreases. Botheffects increase the deposition rate of particles at the top of theheat exchanger tube.

At the area of stagnation, where the flow velocity is zero,removal of particles by shear flow ceases, and consequentlythe particles that stick to the surface of the tube at the area ofstagnation are not removed by the flow. The impact speed of theparticles is so high at the area of stagnation due to the blockageof the flow by the tubes, such that the incident particles bounceoff the tubes. However, when the incident particles bounce off atthe area of stagnation they are decelerated by the flow such thatwhen they hit the tube again, they impact at a very low speedand they stick to the tube surface. This explanation has beenobserved by Kuerten [34] from the numerical simulations hehas performed for a particle-laden flow through a tube bundle.

CONCLUSIONS

An experimental setup has been built to investigate the in-fluence of flow direction with respect to gravity on particulatefouling of heat exchangers. Four orientations of flow have beenstudied: horizontal flow, upward flow, downward flow, and aflow at 45 degrees. It was found that in all orientations foulingstarts at the stagnation area of flow around the heat exchangertube and at the top of the tube. The area of stagnation coin-cides with the top of the heat exchanger tube in the case of a




downward flow, which is the main reason for having only oneposition at which fouling starts. Fouling at the upper half ofthe tubes is much faster than at the lower half of the tubes, andthe fouling rate is faster at the bottom tubes of the heat exchangersection than for the upper tubes. Fouling starts at two positionson the heat exchanger tubes in the case of a horizontal flow,upward flow, and a flow at 45 degrees, while in the case of adownward flow it starts only at one position, which promotes thedownward flow as the best orientation for lingering particulatefouling.

NOMENCLATURE

dp diameter of the particle, mE* reduced Young’s modulus, N/m2

E Young’s modulus, N/m2

g gravityR radius of the particle, mt timeVs critical sticking speed, m/s� surface energy, J/m2

ρ density of the particle, kg/m3

σy yield strength, N/m2

ν Poisson’s ratio, dimensionless

REFERENCES

[1] van Beek, M. C., Rindt, C. C. M., Wijers, J. G., and van Steen-hoven, A. A., Analysis of fouling in refuse waste incinerators,Heat Transfer Engineering, vol. 22(1), pp. 22–31, 2001.

[2] Bryers, R. W., Fireside slagging, fouling and high temperature cor-rosion of heat-transfer surface due to impurities in steam-raisingfuels, Progress in Energy and Combustion Science, vol. 22, pp.29–120, 1996.

[3] Gupta, R. P., Wall, T. F., and Baxter, L. L., The thermal con-ductivity of coal ash deposits: relationships for particulate andslag structures. In The Impact of Mineral Impurities in Solid FuelCombustion, eds. R. P. Gupta, T. F. Wall, and L. L. Baxter, KluwerAcademic Press, New York, pp. 65–84, 1999.

[4] Benson, S. A., Jones, M. L., and Harb, J. N., Ash forma-tion and deposition, in Fundamental of Coal Combustion—ForClean and Efficient Use, Coal Science and Technology, ed. L. D.Smoot, Elsevier Science Publishers, Amsterdam, pp. 299–373,1993.

[5] Epstein, N., Thinking about heat transfer fouling: A 5 × 5 matrix,Heat Transfer Engineering, vol. 4(1), pp. 43–56, 1983.

[6] Epstein, N., Fouling in heat exchangers, in: Proceedings of the6th International Heat Transfer Conference, pp. 235–253, 1987.

[7] Marner, W. J., Progress in gas-side fouling of heat-transfer sur-faces, Applied Mechanics Review, vol. 43, no. 1, pp. 35–66, 1990.

[8] Taborek, J., Aoki, T., Ritter, R., and Palen, J. W., Fouling: Themajor unresolved problem in heat transfer, Chemical EngineeringProgress, vol. 68, pp. 59–67, 1972.

[9] Muller-Steinhagen, H., Reif, F., Epstein, N., and Watkinson, P.,Influence of operating conditions on particulate fouling, CanadianJournal of Chemical Engineering, vol. 66, pp. 42–50, 1988.

[10] Bohnet, M., Fouling of heat-transfer surfaces, Chemical Engi-neering Technology, vol. 10, pp. 113–125, 1987.

[11] Bott, T. R., The Fouling of Heat Exchangers, Elsevier Science,New York, 1995.

[12] Hupa, M., Interaction of fuels in co-firing in FBC, Fuel, vol. 84,pp. 1312–1319, 2005.

[13] Kaiser, S., Antonijevic, D., and Tsotsas, E., Formation of foulinglayers on a heat exchanger element exposed to warm, humid andsolid loaded air streams, Experimental Thermal and Fluid Science,vol. 26, pp. 291–297, 2002.

[14] Abd-Elhady, M. S., Rindt, C. C. M., Wijers, J. G., van Steen-hoven, A. A., Bramer, E. A., and van der Meer, T. H., Minimumgas speed in heat exchangers to avoid particulate fouling, Inter-national Journal of Heat and Mass Transfer, vol. 47, no. 17–18,pp. 3943–3955, 2004.

[15] Kalisz, S., and Pronobis, M., Investigations on fouling rate inconvective bundles of coal-fired boilers in relation to optimizationof sootblower operation, Fuel, vol. 84, pp. 927–937, 2005.

[16] Nuntaphan, A., and Kiatsiriroat, T., Thermal behavior of spiralfin-and-tube heat exchanger having fly ash deposit, ExperimentalThermal and Fluid Science, vol. 3, no. 8, pp. 1103–1109, 2007.

[17] Nuntaphan, A., Kiatsiriroat, T., and Wang, C. C., Air side per-formance at low Reynolds number of cross-flow heat exchangerusing crimped spiral fins, International Communication in Heatand Mass Transfer, vol. 32, pp. 151–165, 2005.

[18] Nuntaphan, A., Kiatsiriroat, T., and Wang, C. C., Heat transfer andfriction characteristics of crimped spiral finned heat exchangerswith dehumidification, Applied Thermal Engineering, vol. 25, no.2–3, pp. 327–340, 2005.

[19] Achenbach, E., Distribution of local pressure and skin frictionaround a circular cylinder in cross-flow up to Re = 5×106, Journalof Fluid Mechanics, vol. 34, no. 4, pp. 625–639, 1968.

[20] Lam, K., Li, J. Y., Chan, K. T., and So, R. M. C., Flow patternand velocity field distribution of cross-flow around four cylindersin a square configuration at a low Reynolds number, Journal ofFluids and Structures, vol. 17, no. 5, pp. 665–679, 2003.

[21] Raask, E., Mineral Impurities in Coal Combustion, HemispherePublishing Corporation, New York, 1985.

[22] De Best, C., Novel Aerosol Condensing Heat Exchanger for Small-Scale Biomass Combustion Applications, PhD thesis, EindhovenUniversity of Technology, Eindhoven, The Netherlands, 2007.

[23] Abd-Elhady, M. S., Rindt, C. C. M., Wijers, J. G., and van Steen-hoven, A. A., Particulate fouling in waste incinerators as influ-enced by the critical sticking velocity and layer porosity, Energy,vol. 30, pp. 1469–1479, 2003.

[24] Thornton, C., and Ning, Z., A theoretical model for the stick be-haviour of adhesive, elastic-plastic spheres, Powder Technology,vol. 99, pp. 154–162, 1998.

[25] Abd-Elhady, M. S., Rindt, C. C. M., Wijers, J. G., and van Steen-hoven, A. A., Modelling the impaction of a micron particle with apowdery layer, Powder Technology, vol. 168, no. 3, pp. 111–124,2006.

[26] Fenner, T. R., Mechanics of Solids, Blackwell Scientific Publica-tions, Oxford, UK, 1989.

[27] Gilman, J.J., Direct measurement of the surface energies of crys-tals, Journal of Applied Physics, vol. 31, no. 12, pp. 2208–2218,1960.

[28] Rogers, D. E., and Reed, J., The adhesion of particles undergoingan elastic-plastic impact with a surface, Journal of Physics D:Applied Physics, vol. 17, pp. 677–689, 1984.




[29] Baxter, L. L., Ash deposition during biomass and coal combustion:A mechanistic approach, Biomass and Bioenergy, vol. 4, no. 2,pp. 85–102, 1993.

[30] Sons, J. S., and Hanratty, T. J., Velocity gradient at the wall forflow around a cylinder at Reynolds numbers from 5×103 to 105,Journal of Fluid Mechanics, vol. 35, no. 2, pp. 353–368, 1969.

[31] Wu, M., Wen, C., Yen, R., Weng, M., and Wang, A., Experimen-tal and numerical study of the separation angle for flow around acircular cylinder at low Reynolds number, Journal of Fluid Me-chanics, vol. 515, pp. 233–260, 2004.

[32] Cabrejos, F. J., and Klinzing, G. F., Incipient motion of solidparticles in horizontal pneumatic conveying, Powder Technology,vol. 72, pp. 51–61, 1992.

[33] Halow, J. S., Incipient rolling, sliding and suspension of particlesin horizontal and inclined turbulent flow, Chemical EngineeringScience, vol. 28, pp. 1–12, 1973.

[34] Kuerten, J. G. M., Numerical simulation of particle-laden flowthrough tube bundles, Report for EU Project SES6-CT-2003-502679, BIOASH, 2007.

Mohamed S. Abd-Elhady has been an assistant pro-fessor since July 2006 in the field of energy technol-ogy at the Department of Mechanical Engineering inBeni-Suief University, Egypt. He started as a researchassistant in October 1995 within the same group. Hereceived an M.Sc. degree in 1998 and Ph.D. degreein 2005 from Eindhoven University of Technology,the Netherlands. During his Ph.D. he worked on gas-side particulate fouling in biomass gasifiers and wasteincinerators. He is currently working on developing

new techniques to reduce particulate fouling of heat exchangers and his interestsare renewable especially solar energy.

Camilo Rindt has been an associate professor sinceFebruary 2001 in the field of heat transfer in the Di-vision Energy Technology at the Department of Me-chanical Engineering, Eindhoven University of Tech-nology, the Netherlands. He started as an assistantprofessor in January 1991 within the same group. Hereceived an M.Sc. degree in 1985 and Ph.D. degree in1989 from Eindhoven University of Technology. Un-til January 1991 he worked in the field of biomedicalengineering on experimental and numerical analyses

of blood flow in the carotid artery bifurcation. His main interests are in thefields of conjugate heat transfer and radiation modeling (electronics coolingand PV/T combi-panels) and fouling of heat exchangers due to particles (wasteincinerators and biomass gasifiers).

Anton van Steenhoven has been a full professor ofenergy technology since 1990 at the Department ofMechanical Engineering in Eindhoven University ofTechnology, the Netherlands. He joined the Facultyof Mechanical Engineering in 1979, first in the Pro-duction Technology division and next in the Engi-neering Mechanics division. He became an associateprofessor in 1985 in (bio-)fluid mechanics. His maininterests are heat transfer and transitional flows, witha special focus on renewable energy systems. He is

member of the “Hollandsche Maatschappij der Wetenschappen,” FOM pro-gram manager, and chairman of the KNAW jury for the annual DOW En-ergy prize. As part of his international activities he is a Dutch delegate inthe assemblies of IHTC and IUTAM; he is also a member of the IAB ofIPPT (Warsaw) and participates in the scientific councils of Eurotherm andICHMT.



optimization of flow direction to minimize …m. s. abd-elhady et al. 897 figure 2 the different...

Documents