Fouling Corrosion in Aluminum Heat Exchangers

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector Chinese Journal of Aeronautics, (2015), 28(3): 954–960

Chinese Society of Aeronautics and Astronautics & Beihang University Chinese Journal of Aeronautics

REVIEW ARTICLE Fouling corrosion in aluminum heat exchangers

Su Jingxin a,*, Ma Minyu a, Wang Tianjing b, Guo Xiaomei b, Hou Liguo b, Wang Zhiping a

a Tianjin Key Laboratory for Civil Aircraft Airworthiness and Maintenance, Civil Aviation University of China, Tianjin 300300, China b Aircraft Maintenance and Engineering Corporation, Beijing Capital International Airport, Beijing 100621, China

Received 15 September 2014; revised 30 October 2014; accepted 15 December 2014 Available online 20 March 2015

KEYWORDS Abstract Fouling deposits on aluminum heat exchanger reduce the heat transfer efficiency and Aluminum; cause corrosion to the apparatus. This study focuses on the corrosive behavior of aluminum cou- EIS; pons covered with a layer of artificial fouling in a humid atmosphere by their weight loss, Tafel Heat exchanger; plots, electrochemical impedance spectroscopy (EIS), and scanning electron microscope (SEM) Pitting corrosion; observations. The results reveal that chloride is one of the major elements found in the fouling Tafel plots which damages the passive film and initiates corrosion. The galvanic corrosion between the metal and the adjacent carbon particles accelerates the corrosive process. Furthermore, the black carbon favors the moisture uptake, and gives the dissolved oxygen greater chance to migrate through the fouling layer and form a continuous diffusive path. The corrosion rate decreasing over time is con- formed to electrochemistry measurements and can be verified by Faraday’s law. The EIS results indicate that the mechanism of corrosion can be interpreted by the pitting corrosion evolution mechanism, and that pitting was observed on the coupons by SEM after corrosive exposure. ª 2015 Production and hosting by Elsevier Ltd. on behalf of CSAA & BUAA. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction minimum of scales, causing major issues to the design and operation of the heat exchangers. New Zealand scientist, When heat transfer surfaces become fouled, the coefficiency of Steinhagen, investigated more than 3000 heat exchangers from heat transfer will be dramatically reduced with even a thousands of different companies and found that over 90% of them had varying degrees of scaling.1 The economic loss caused by this fouling accounted for about 0.2% of the GDP * Corresponding author. Tel.: +86 22 24092074. per year in developed countries.2 Thackery estimated that E-mail addresses: [email protected] (J. Su), erzascrlet@163. the loss caused by fouling was nearly 500 million pounds each com (M. Ma), [email protected] (T. Wang), guoxiaomei@ year.3 The inefficiency of heat exchangers in civil aircraft’ air ameco.com.cn (X. Guo), [email protected] (L. Hou), [email protected] (Z. Wang). conditioning (AC) systems induced by fouling may shut down Peer review under responsibility of Editorial Committee of CJA. the system, and jeopardize the pressure and temperature con- trols in the passengers’ cabin. Studies of the fouling problems beginning in the 1980s have been developing along three directions: theoretical analysis of Production and hosting by Elsevier the formation of the fouling in order to supply universal and http://dx.doi.org/10.1016/j.cja.2015.02.015 1000-9361 ª 2015 Production and hosting by Elsevier Ltd. on behalf of CSAA & BUAA. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Fouling corrosion in aluminum heat exchangers 955 accurate predictive models;4–6 fouling development monitoring Table 1 Results of elements’ contents in fouling samples. devices;7–9 and the technology to control and remove the fouling.10 There is little actual published literature on the corrosive Element Content (wt%) fouling found on heat exchangers. The heat exchangers on the EDS EDX aircrafts’ AC modules are made of an aluminum alloy with O 19.93 inferior corrosion resistance than the other two types of com- Na 2.24 7.283 monly applied materials: the copper alloy and the stainless Mg 2.98 2.288 steel. The specific active ingredients in the fouling, chloride Al 42.61 11.96 for example, can damage the passivated surface of the alu- Si 3.87 27.17 minum alloy, reducing the service life of the heat exchangers. P 2.31 3.87 The serious corrosion may even cause a leak in the pressurized S 1.22 air, which may de-pressurize the passengers’ cabin. Some other K 1.28 4.385 studies on the atmospheric corrosion of aluminum alloy have C 24.8 certain reference value for this study.11–13 Ti 1.36 Cl 0.665 In this study, the fouling was sampled from actual aircrafts’ Ni 0.403 heat exchangers and analyzed for its composition. Then, the Co 0.0226 aluminum coupons covered with a coating of homogeneous Ca 1.37 2.863 artificial fouling were put into an environment chamber, which Fe 20.19 12.2 was set as close as possible to the actual service environment of Cu 1.82 0.122 the heat exchangers. The corrosive behavior were characterized by measuring the corrosion rate, comparing the effect of the corrosion on different components, and studying the metal, promoting the onset of micro-cell in the voids of the cov- mechanism of corrosion at different stages. ering layers, and increases the corrosion rate due to the galvanizing effect. The other fouling components are not as 2. Experiment important under the electrochemical corrosion mechanism of aluminum alloys. So the contents of Cl and carbon must be taken into account as the key variables in the study. 2.1. Fouling analysis and experiment design Based on the above analysis, this paper presents a test for corrosion caused by the artificial fouling, which is very similar Samples of the fouling were collected from the heat exchangers to the widely applied ASTM corrodokote test,16 in order to by scratching the heat transferring surface or by centrifugally simulate the actual corrosive factors. separating a sample from the cleaning solution used to remove the fouling. Each sample was then tested with an energy dis- 2.2. Preparation of fouling and samples persive spectrometer (EDS) and an energy dispersive X-ray spectroscopy (EDX) respectively. The results are shown in Table 1. The results of these two tests deviate from each other NaCl, Fe2(SO4)3, CuSO4,Na3PO4,Na2CO3,Na2SiO3, on elements contents of O, S, and C because the fundamental Ca(NO3)2,Na3PO4,Na2CO3, and carbon black were added characteristics of these two analysis methods and the to 100 mL distilled water as per the amounts listed in corresponding samples’ nature, with one being deposit sedi- Table 2, then continuously stirred for 24 h. After that, 35 g ment while the other, the liquid extract of the sediment, are dif- of kaolin clay was added and the mixture was stirred for ferent. The artificial fouling design therefore cannot be merely another 30 min. The mixture was left standing for about 1 h determined by these tests, the operation conditions of the heat until a paste formed. The paste had a gray-green to dark-green exchangers must be taken into consideration additionally. appearance. The fouling on the gas-side of aircraft heat exchangers 2024 aluminum alloy (nominal composition by weight per- comes from atmospheric aerosols and the engine exhaust. centage: 0.50 Si, 0.50 Fe, 3.8–4.9 Cu, 0.30–1.0 Mn, 1.2–1.8 Mg, Atmospheric aerosols are complicated dispersions of liquid 0.10 Cr, 0.25 Zn, the balance is Al) sheets were machined into and solid particles into the atmosphere, which are mainly com- coupons of 50 mm · 5mm· 2 mm with a 1.5 mm diameter posed of soluble components, organic carbon, element carbon, hoisting hole near one of the short edges. Each coupon was carbon black, and inorganic alumino-silicate, among sanded up to 1200 # abrasive paper. The fouling paste was 14,15 2 + applied on one side of the coupon using a KW-4A spinning others. Water-soluble ions, such as SO4 ,NO3 ,Cl ,K , Na+,Ca2+, etc., account for a large proportion of the particles coater. The thickness of the fouling layer was approximately in the aerosols. The organic carbon and the black carbon 0.2 mm after drying out. The typical appearances of the testing mainly come from the combustion of fossil fuels in aircrafts’ coupons are shown in Fig. 1. engines. At working temperature (100–200 C), the fouling on the heat exchangers transforms into porous scales, which 2.3. Corrosion rate measurements absorb humidity from the air to form a water-soaked coating over the aluminum surface in which water-soluble components The fouling covered samples were hung in a temperature-hu- travel through the solution phase to form an electrolytic cell. midity test chamber, which was set at 38 C and 80% RH. The Cl has been well accepted as one of the dominant factors After every 24 h during the experiment, 10 samples from each of aluminum alloys’ corrosion for it damages the passivation of group were withdrawn from the chamber for measurements. these alloys. The black carbon favors the moisture uptake The bare aluminum coupons were weighed (m1) before the through porous covering layers down to the surface of the base fouling preparation. After being exposed to the corrosive 956 J. Su et al. environment, the fouling layer and the corrosion product were chemically removed by rinsing the coupons the solution, which was compounded of 20 g CrO3,50mLH3PO4 and 950 mL deionized water, at 90 C for 5 min,17 and then the coupons were weighed for the second time (m2). The Tafel plot was also measured, and the corrosion current density (icorr) was calculated for the same batch of samples as a means of validating the weight loss data. The measurements were carried out on a PARSTAT 4000 potentiostat (Princeton Applied Research), which connected to a computer with one copy of VersaStudio software installed. One three- electrode system was used in the experiments: a saturated calo- mel electrode (SCE) as the reference electrode, a carbon rod as the counter electrode, and a corroded coupon as the working electrode. The specimens were mounted to a coating evalua- Fig. 1 Typical appearances of the testing coupons with artificial tion cell after being taken out of the test chamber (see fouling layer. Fig. 2(a)), so as to leave a 1.77 cm2 window (Fig. 2(b)) in the test solution. The test solution was 3.5wt% NaCl solution, except for samples from group A where 3.5wt% Na2SO4 solution was used in order to exclude the effect of the chloride. Tafel curves were obtained by using the linear scanning voltage method, where the scanning range was between 300 and 300 mV versus the open circuit potential, and the scanning rate was 2 mV/s. The icorr was obtained from the Tafel curve by extrapolating the line of the region of the strong polarization.18

2.4. EIS measurements

EIS measurements were taken at the open circuit potential in a frequency range from 100 kHz to 100 mHz, with a 10 mV amplitude signal on one PARSTAT 4000 potentiostat. The same cell setup was used for EIS measurements as was used for the Tafel tests. Fig. 2 Coating evaluation cell used in electrochemical 2.5. Fouling layer porosity measurements measurements. indicate that almost no corrosion occurs for group A, and The pores in the fouling layer play a critical role in moisture the weight loss data of groups B and C are almost the same. absorption and oxygen delivery, which could inﬂuence the cor- This illustrates that very little corrosion happens when the rosion rate. The porosity property of the fouling was tested by chloride is not present, but adding more chloride does not Quantachrome Autosorb-1 (Quantachrome, US) according to increase the weight loss necessarily. In the corrosion system ISO 9277:1995 (the determination of the speciﬁc surface area used in this research, the base metal of the fouling layer is a of solids by gas adsorption using the BET method) for better type of aluminum alloy that keeps passivated by itself until it knowledge of the corrosion mechanism. comes in contact with halogen ions in neutral solutions, show- ing a very low corrosion rate. However, the corrosion rate of 3. Results and discussion the activated aluminum is determined by other factors than the halide ion concentration. 3.1. Weight loss The values shown in Fig. 4, by contrast, obviously indicate that the weight loss increases with the increasing proportion of The weight loss from the fouling sample groups, each with dif- carbon, indicating that the very existence of carbon black par- ferent amounts of NaCl, are shown in Fig. 3. The results ticles indeed accelerate the corrosion. This behavior can be

Table 2 Components of artiﬁcial fouling. Group Components in one batch of artiﬁcial fouling (g)

Fe2(SO4)3 CuSO4 Ca(NO3)2 Na3PO4 Na2SiO3 Na2CO3 Kaolin clay NaCl Carbon A5 5 1015151535010 B5 5 1015151535510 C5 5 10151515351010 D5 5 1015151535105 E5 5 1015151535100 Fouling corrosion in aluminum heat exchangers 957 partially interpreted under the galvanic corrosion effect between the metal and the adjacent carbon particles. Furthermore, the porosity characteristics of the fouling layers (Table 3) indicate that the fouling layers from Groups C through E have smaller speciﬁc surface areas and total pore volumes as the carbon quantity decreases. The cathodic depolarizer in this experiment was the dissolved oxygen in the solution in the micro pores of the fouling layer. A fouling layer with a greater porosity is favorable for the dissolved oxygen to migrate through it because the micro pores have greater chance to connect with one another to form a continuous diffusive path. The weight loss increases during an initial phase, but then the growth rate decreases over time for each group, which manifests as a decrease in the real time corrosion rate.

3.2. Electrochemical behavior Fig. 4 Accumulated weight loss of groups with different carbon proportion at different time. The icorr values obtained from the Tafel plots are presented in Fig. 5. It is clear that the icorr decreases by the time, revealing the corrosion rate slowdowns. This result further validates the where m1 m2 is the accumulated weight loss (g) due to corro- weight loss measurements. sion over exposure time t (h) until the sample was withdrawn, 2 According to the Faraday’s Law, the corrosion rate (CR, A is the coupons’ surface area (cm ), F is Faraday’s constant mA/cm2) can be calculated using the weight loss data by the (96500 C/mol), and M is the atomic weight of Al (27 g/mol). following equation: The C.R. values calculated from Eq. (1) compared with icorr values obtained from the Tafel plots of Group B are shown 3F dðm m Þ CR ¼ 1 2 ð1Þ in Fig. 6. The two sets of data match each other when taking 3:6MA dt both tendency and magnitude into consideration, which validates both measurements. The EIS diagrams of group A in a Na2SO4 (3.5wt%) solution and the other four groups in a NaCl (3.5wt%) solution are shown as a function of the exposure time in Figs. 7 and 8,in

Table 3 Porosity characteristics of the fouling layers. Group Carbon Speciﬁc Total pore Average (g)* surface area volume (cm3/ pore size (m2/g) g) (nm) A 10 48.923 7.374 · 102 6.029 B 10 63.635 1.037 · 101 6.517 C 10 41.907 6.235 · 102 5.968 D 5 26.637 5.668 · 102 8.511 E 0 3.877 2.208 · 102 3.827 Fig. 3 Accumulated weight loss of groups with different NaCl Note: * Carbon content in artiﬁcial fouling as described in Table 2. proportion at different time.

Fig. 5 The icorr data of the fouling covered samples from Group A to Group E. 958 J. Su et al.

Fig. 6 CR and icorr data of fouling covered samples of Group B.

Fig. 7 EIS diagrams for Group A in 3.5% Na2SO4 solution at different exposure time. which the points represent measured values and the solid lines are the results of Nonlinear least squares fitting using equivalent electrical circuit models. The fitting lines coincide well with electrode surface and the tip of the reference electrode, Q is the experimental data. The models applied in the fittings are the capacity of the fouling layer, Rpo represents the solution given in Fig. 9. resistance in the micro pores of the fouling layer, Cdl is the The EIS diagrams of group A at different measurements electric double layer capacity at the sites of the metal surface resemble one another and are all characterized by a capacitive which is in contact with the pore solution, and Rf is the equiva- loop at high and low frequency ranges. All of the diagrams lent resistance of the passivating film growth on the surface of coincide with the equivalent electrical circuit in Fig. 9(a), in the aluminum. In the theory of Cao and Zhang19 the faradaic which Rs reflects the solution resistance between the working impedance of the passivated metal can be reduced to Rf if

Fig. 8 EIS diagrams for four groups in 3.5% NaCl solution at different exposure time. Fouling corrosion in aluminum heat exchangers 959

Fig. 9 Equivalent circuit models proposed to carry out curve fitting for EIS data. maintained at a steadily passivated state, which occurred in the experiment system, because there was no halide present during the corrosion exposure. Moreover, Rf is usually on the order of 5 6 2 10 –10 XÆcm and tremendously larger than Rpo, so that the low frequency capacity loop has a maximal curvature radius. The EIS diagrams of the samples with halide in the fouling layer, i.e. Groups B–E, manifest an evolution quite dissimilar from that of Group A. Group E has the slowest corrosion development (Fig. 4) and therefore, gives the distinct EIS pattern of the early stage of the corrosion evolution, the ‘‘pitting initiation’’, which is characterized by two capacitive loops fol- lowed by one inductive loop after 24 h of moisture exposure. The inductive loop at low frequencies reflects the pitting corrosion initiation on the metal surface under the micro pores of Fig. 10 SEM observations for pitting corrosion holes. the fouling layer.20,21 It can be explained by the equivalent electrical circuit in Fig. 9(b), in which R , R and L are equiva- t L equivalent electrical circuit shown in Fig. 9(d), where W is lent elements in relation to the corrosion pitting sites’ repas- the diffusion resistance. It reflects that the corrosion process sivation.22 The EIS results of the other three groups in Fig. 8 is controlled by the diffusion of the dissolved oxygen through present no such traits, because of the faster corrosion the fouling layer and the micro pitting caused by corrosion on development. the metal surface.25–27 From 48 to 240 h, the inductive loop is no longer present in the diagrams for Group E, which evolve gradually into a two- loop pattern with each capacitive loop at high and low fre- 4. Conclusions quency ranges in the Nyquist plot and coincide with the equivalent electrical circuit shown in Fig. 9(c). Quite similar A research method including spin coating an artificial fouling patterns can also be identified in the diagrams of Group B layer and recreating exposure in humid air has been designed (from 24 h through 240 h of exposure), Group C (24 h and and validated to study the fouling corrosion of aluminum heat 48 h), and Group D (from 24 h through 144 h). exchangers. This pattern indicates that the pitting corrosion is under- The results show that weight loss rate data coincide with the going stable development and/or propagation at the interface corrosion current density, as the weight loss measurement is of the fouling layer and the metal.23,24 One of the samples with equivalent to the Tafel method. The accumulated weight loss such pattern was implanted with an epoxy resin, cut in half, curves and the EIS diagrams demonstrate that two kinds of and then observed under the SEM (1530VP, LEO, Germany) corrosion promoting elements, inorganic carbon and chloride, to verify the existence of pitting. The section surface image is express themselves during the experiments. shown in Fig. 10, in which numerous micro pores can be seen Pitting was observed on the section of the testing coupons at the bottom of the larger corrosion pits. Therefore, the two- after certain exposure times. The EIS data can also be inter- time-constant model in Fig. 9(c) is a satisfying simulation for preted by the pitting corrosion evolution mechanism of passi- the interfacial procedures, while Cdl and Rp are, respectively, vated metals. In the halide-free systems, the EIS plots are the solution resistance and the charge transfer resistance in composed of two capacitive loops, one of which is related to the corrosion pits. the passivated state of the aluminum surface. In the systems The EIS diagram of Group C and D at 240 h represent with chloride added to the fouling layer the EIS diagrams another characteristic with a capacitive loop at high frequen- demonstrate different traits during the stages of pitting initia- cies and radial at low frequencies, which coincide with the tion, pitting development, and diffusion control. 960 J. Su et al.

Acknowledgements 16. Murty VVVRS, Kattamis TZ, Devereux OF. The effect of solidification microstructure on the corrosion behaviour of a This study was supported by the National Natural Science grain-refined aluminium–copper alloy. J Mater Sci 1974;9(4): 688–92. Foundation of China (No. 21303261), and the Major Science 17. Li XH, Deng SD, Xie XG, Du GB. Inhibition effect of bamboo & Technology Research Project of Civil Aviation of China leaf extract on the corrosion of aluminum in HCl solution. Acta (No. MHRD20140110). Phys Chim Sin 2014;30(10):1883–94 [Chinese]. 18. Amin MA, El-Rehim SSA, El-Sherbini EEF, Bayoumi RS. The References inhibition of low carbon steel corrosion in hydrochloric acid solutions by succinic acid – Part I. Weight loss, polarization, EIS, PZC, EDX and SEM studies. Electrochim Acta 2007;52(11): 1. Steinhagen R, Mu¨ller-Steinhagen H, Maani K. Problems and costs 3588–600. due to heat exchanger fouling in New Zealand Industries. Heat 19. Cao CN, Zhang JQ. Electrochemical impedance spectroscopy Transfer Eng. 1993;14(1):19–30. application in corrosion science. Beijing: Science Press; 2002 2. Yu TL, De DQ, Yu XM. Power plant steam turbine condenser [Chinese]. fouling automatic cleaning and its heat transfer reinforcement. 20. Alonso C, Andrade C, Rodriguez J, Diez JM. Factors controlling Chem Equip Technol 2002;23(3):30–2. cracking of concrete affected by reinforcement corrosion. Mater 3. Thackery PA. The cost of fouling in heat exchanger plant. Effluent Struct 1998;31(211):435–41. Water Treat 1980;5(20):112–5. 21. Gill JS, Callow LM, Scantlebury JD. Corrosion measurements 4. Lalot S, Lecoeuche S. Onling fouling detection in electrical derived from small perturbation non-linearity––Part 1: harmonic circulation heaters using neural networks. Int J Heat Mass analysis. Corrosion 1983;39(2):61–6. Transfer 2003;46(13):2445–57. 22. Mansfeld F, Kendig MW, Tsai S. Recording and analysis of AC 5. Zubair SM, Sheikh AK, Shaik MN. A probabilistic approach to impedance data for corrosion studies. Corrosion 1982;38(11): the maintenance of heat-transfer equipment subject to fouling. 570–80. Energy 1992;17(8):769–76. 23. Cao CN, Wang J, Lin HC. Effect of Cl-ion on the impedance of 6. Javadi N, Ashtiani FZ, Fouladtajar A, Zenooz AM. Experimental passive-film-covered electrodes. J Chin Soc Corros Protect studies and statistical analysis of membrane fouling behavior and 1989;9(4):261–9 [Chinese]. performance in microfiltration of microalgae by a gas sparging 24. Chung SC, Sun SI, Hsien CC, Shih HC. Application of EIS to the assisted process. Bioresour Technol 2014;160(1):350–7. initial stages of atmospheric zinc corrosion. J Appl Electrochem 7. Teixeira R, Pereira A, Mendes J, Melo LF. Identifying the nature 2000;30(5):607–15. of fouling layers by online monitoring of the propagation of 25. Tsai CH, Mansfled F. Determination of coating deterioration with vibrations along the deposition surface. Heat Transfer Eng EIS: Part II. Development of a method for field testing of 2014;35(3):251–7. protective coationgs. Corrosion 1993;49(1):726–30. 8. Crattelet J, Ghnimi S, Debreyne P, Zaid I, Boukabache A, Esteve 26. Graedel TE. Corrosion mechanisms for aluminum exposed to the D, et al. On-line local thermal pulse analysis sensor to monitor atmosphere. J Electrochem Soc 1989;136(1):204–12. fouling and cleaning: application to dairy product pasteurisation 27. Yan M, Vetter CA, Gelling VJ. Corrosion inhibition performance with an ohmic cell jet heater. J Food Eng 2013;119(1):72–83. of polypyrrole Al flake composite coatings for Al alloys. Corros 9. Dreszer C, Flemming HC, Wexler AD, Zwijnenburg A, Kruithof Sci 2013;70:37–45. JS, Vrouwenvelder JS. Development and testing of a transparent membrane biofouling monitor. Desalination Water Treat 2014;52(12):1807–19. Su Jingxin is an associate professor at School of Science, Civil 10. Malayeri MR, Jalalirad MR. Mitigation of crystallization fouling Aviation University of China. The current research interest is local in a single heated tube using projectiles of different sizes and corrosion of aircrafts’ structures. hardness. Heat Transfer Eng 2014;35(16):16–7. Ma Minyu is a master degree candidate at School of Science, Civil 11. Katayama H, Kuroda S. Long-term atmospheric corrosion Aviation University of China, graduated from the same university with properties of thermally sprayed Zn, Al and Zn–Al coating exposed bachelor’s degree in 2008. in a coastal area. Corros Sci 2013;76:35–41. 12. Liu YJ, Wang ZY, Ke W. Study on influence of native oxide and Wang Tianjing is a assistant engineer at Aircraft Maintenance corrosion products on atmospheric corrosion of pure Al. Corros Engineering Co. Ltd., Beijing, majoring in aircraft accessories trou- Sci 2014;80:169–76. bleshooting, graduated from Beijing University of Aeronautics and 13. Thirumalaikumarasamy D, Shanmugam K, Balasubramanian V. Astronautics in 2004. Influence of pH, chloride ion concentration and spraying time on the corrosion behaviour of atmospheric plasma sprayed alumina Guo Xiaomei is a engineer at Aircraft Maintenance Engineering Co. coatings. Surf Eng 2014;30(3):183–94. Ltd., Beijing, majoring in aircraft air-conditioning system. 14. Babu SS, Moorthy KK, Satheesh SK. Aerosol black carbon over Arabian Sea during intermonsoon and summer monsoon seasons. Hou Liguo is a senior engineer at Aircraft Maintenance Engineering Geophys Res Lett 2004;31(6):5–7. Co. Ltd., Beijing, majoring in aircraft air-conditioning system. 15. Lou LP, Luo L, Wang W, Xu XH, Hou JA, Xun B, et al. Impact of black carbon originated from fly ash and soot on the toxicity of Wang Zhiping is a professor at School of Science, Civil Aviation pentachlorophenol in sediment. J Hazard Mater 2011;190(3): University of China. The current research interest is Failure Analysis 474–80. of Engineering Materials.