National Academies Press: OpenBook

Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure (2008)

Chapter: Chapter Three - Effects of Pavement Deicing Products on Aircraft Components

« Previous: Chapter Two - Use of Airport Pavement Deicing Products at Airports
Page 8
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 8
Page 9
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 9
Page 10
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 10
Page 11
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 11
Page 12
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 12
Page 13
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 13
Page 14
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 14
Page 15
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 15
Page 16
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 16
Page 17
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 17
Page 18
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 18
Page 19
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 19
Page 20
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 20
Page 21
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 21
Page 22
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 22
Page 23
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 23
Page 24
Suggested Citation:"Chapter Three - Effects of Pavement Deicing Products on Aircraft Components." National Academies of Sciences, Engineering, and Medicine. 2008. Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure. Washington, DC: The National Academies Press. doi: 10.17226/13913.
×
Page 24

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

8Alkali-metal-salt-based PDPs such as KAc and KF entered the European market to a significant extent in the mid- to late-1990s (approximately 1995 for KAc and 1998 for KF). A few years later, these modern PDPs entered the U.S. mar- ket. In both cases, these salts were introduced as alternatives to urea and glycols used in traditional PDPs for freezing point depression to mitigate the environmental concerns re- lated to airfield deicing and anti-icing operations. It became apparent soon after their introduction that these new deicers presented new challenges. For instance, introduction of the alkali-metal-salts as PDPs coincided with the increased fre- quency of failures and returns of aircraft carbon–carbon (C/C) brakes, whereas aircraft C/C brakes had not changed signif- icantly since their introduction in mid-1980s. Aircraft C/C composite brakes mainly consist of a carbon matrix reinforced by carbon fibers (either woven or randomly dispersed). In 1985, Airbus Industries introduced the first Messier–Bugatti production carbon brakes on the A310-300 and A300-600 aircraft (http://www.messier-bugatti.com/article.php3?id_ article=180). Other issues concerning the aircraft durability and operations have also contributed to the debate surround- ing the use of such PDPs. The following sections of this chapter will discuss the effects of PDPs on aircraft components, including catalytic oxidation of C/C composite brakes, corrosion of aircraft alloys (with a focus on cadmium plating), and interaction with air- craft deicing and anti-icing products. CATALYTIC OXIDATION OF CARBON–CARBON COMPOSITE BRAKES Composites of a carbon matrix reinforced by carbon fiber (C/C composites) possess excellent mechanical and thermal properties. These C/C composites are much lighter than steel, maintain a near-constant friction coefficient over a broad tem- perature range, possess a much higher heat capacity and ther- mal conductivity than steel, and have tensile strength gener- ally twice that of steel at elevated temperatures. Owing to their advantages over metallic friction materials, C/C com- posites have been increasingly used on aircraft brakes (Chai and Mason 1996). Since their aviation debut aboard the Con- corde and the Vickers Super VC-10 (“Company History” 2007), there has been a general transition from metallic brakes to C/C composite brakes on modern medium and large air- craft (Chai and Mason 1996). The typical C/C composite brake assembly consists of a primary thrust stator and multiple stator disks in an alter- nating stack with multiple rotor disks (Figure 3). The float- ing stators are kept from rotating with the motion of the land- ing gear wheel by interlocking stator tenons and torque tube splines, while the similarly keyed rotors rotate with the wheel. When activated, an annular piston assembly presses the thrust stator into the stack, causing very high friction force between the rotors and stators. Such friction can cause the carbon discs to reach 400°C–600°C in normal operations and up to 1400°C in some extreme cases such as in the event of refused take-off (Wu 2002). A known drawback of C/C composites is their susceptibil- ity to thermal oxidation at brake operating temperatures. At temperatures above 400°C, the reaction of carbon with oxy- gen can easily occur and thus wears the unprotected C/C com- posites. Existing research has demonstrated that oxidation (gasification) is the predominant mechanism for weight loss (an indicator of loss in structural integrity) of the C/C com- posites under high-energy brake operation conditions (Wu 2002). These conditions led to the popular practice of pro- tecting the surface of C/C composite brakes with anti-oxidant coatings. As such, thermal oxidation usually will not occur until the anti-oxidant layer is disrupted. Dynamometer testing at 50% relative humidity and 100% normal aircraft landing energy led to significant oxidation of friction layer and fric- tion debris samples obtained from commercially available C/C composite brake material, accompanied by the release of gaseous H2O, CO2, and CO (Penszynska-Bialczyk et al. 2007). Therefore, some level of thermal oxidation is expected and effective oxidation protection is essential to the design of C/C composite brakes. As commercial and military aviation use of C/C composite brakes continues to grow, so does the scope of challenges asso- ciated with the technology. Alkali-metal-salt contaminants (e.g., sodium from the marine environment and potassium from cleaning and deicing chemicals) can reach the carbon surface and act as catalysts to facilitate oxidation of C/C composite brakes under conditions milder than those required for thermal oxidation. Concerns have been raised about the effect of modern PDPs (mainly alkali-metal-salts) on aircraft brakes. The rest of this section will synthesize the information on the validity and nature of the effect of modern PDPs (primar- CHAPTER THREE EFFECTS OF PAVEMENT DEICING PRODUCTS ON AIRCRAFT COMPONENTS

9FIGURE 3 Typical airplane wheel and brake configuration: cross-sectional view. Wheel Assembly Brake Assembly Rotor disks Stator disks Rotor tenons keyed to wheel splines Stators keyed to torque tube splines Primary thrust stator FIGURE 4 Carbon–carbon brake damage attributed to catalytic oxidation: (a) oxidized rotor drive lugs, (b) oxidized stator tenons (courtesy: Continental Airlines). (a) (b) ily alkali metal salts) on aircraft brakes, describe the related standards and test protocols, discuss ways to prevent and mit- igate such effects, and identify knowledge gaps. Validity of the Effect of Modern Pavement Deicing Products on Aircraft Brakes Thermal oxidation is the primary design specification gov- erning durability of aircraft C/C composite brakes. This is substantiated by data from Goodrich’s Problem Analysis Report System, which has monitored failures and premature removals of C/C composite brakes (McCrillis 2007). Whereas thermal oxidation can be seen uniformly at the hottest brake locations, catalytic oxidation typically corresponds to local- ized regions in cooler brake locations in contact with con- taminant residue (e.g., aft aircraft positions) (Walker 2007). Figure 4 illustrates some C/C brake damage attributed to cat- alytic oxidation. Catalytic oxidation of C/C composite brakes owing to air- field PDPs has become a growing concern to be monitored in the ever-changing operation environment (McCrillis 2007). As nontraditional chemical contaminants, modern PDPs may be responsible for the more rapid structural failure of C/C composite brakes in recent years. To avoid potential safety implications, this concern has to be mitigated through more

10 illustrate the catalysis of carbon oxidation by a KAc-based PDP, with the C/C composite samples impregnated with the PDP showing a higher oxidation rate and a higher weight loss at lower temperatures (Filip 2007). As the brake frictional characteristics are changed by the use of alkali-metal-salt-based PDPs, airline operators are con- cerned over the adverse effect of such PDPs on the braking performance and safety of aircraft (Jansen 2007). In addition to reduced brake life, such effect introduces the possibility of brake failure during aborted take-off, with the concomitant risk of fire from hydraulic fluid released during such an event. Since 2003, Dunlop has analyzed the contamination of material samples of heat packs returned for service using the induction coupled plasma spectroscopy (Hutton 2007). A plain geographic trend can be seen (as shown in Table 5), with those aircraft operated in Northern Europe experiencing the highest rates of potassium contamination and those oper- ated in Southern Europe and the Mediterranean region expe- riencing the lowest rates of contamination. According to Dunlop (Hutton 2007), this trend corresponds to the reported frequency of catalytic oxidation in the BAe 146 fleet oper- ated throughout Europe—highest in Nordic countries where KAc and KF are regularly used for snow and ice control and lowest in Southern European countries where PDPs are not needed. Although aircraft are moving objects exposed to var- ious PDPs, climatic conditions, and maintenance practices at different airports, the contaminants encountered other than PDPs often are essentially the same type of fire extinguish- ing or cleaning fluids as defined in the manufacturers’ con- sumable materials list of allowable materials. Therefore, PDPs play a key role in the trend seen in Table 5. Evidence from aircraft operators and manufacturers in North America, Europe, and Asia corroborate the role of PDPs in catalytic oxidation of C/C composite brakes. Irregular C/C composite brake wear and damage has been reported on vir- tually every aircraft platform equipped with them by inter- national and regional carriers (Duncan 2007). ACRP survey FIGURE 5 Weight loss curves of samples of new, used, and KAc-impregnated aircraft brakes obtained by TGA at 500°C (a) and 600°C (b) [adapted from Carabineiro et al. (1999)]. TABLE 4 TGA DATA AS A FUNCTION OF C/C COMPOSITE AND DEICER CONTAMINATION a New Brake 1 0,8 0,6 0,4 0,2 New Brake Used Brake Used Brake Time (h) 0 2 4 6 8 10 12 14 16 18 20 22 24 Impregnated Brake W ei gh t / In iti al w ei gh t 1 0,8 0,6 0,4 0,2 0 W ei gh t / In iti al w ei gh t Impregnated Brake b Sample Weight Loss (%) Virgin HWCCA 45 HWCCA+ PDP 71 Virgin HWCCD 53 HWCCD+ PDP Onset of Oxidation (°C) 447 421 382 325 Maximum Rate of Weight Loss (% per min) 0.617 0.902 0.619 0.720 76 Adapted from Filip (2007). Note: HWCCA, HWCCB, and HWCCD stand for different C/C composite materials, specifically for CarbenixTM C2000 2D pitch/charred resin, CarbenixTM C2000 2D-modified pitch/charred resin, and CarbenixTM 400 3D PAN/CVD, respectively. frequent proactive maintenance and inspection activities incur- ring high direct and indirect costs. A growing body of field evidence from airline operators suggests that the use of KAc and KF on airfield pavements leads to catalytic oxidation of C/C composite brake compo- nents. The thermogravimetric analysis (TGA) data in Figure 5 illustrate the catalysis of carbon oxidation by potassium species, with the new brake samples impregnated with KAc showing high-oxidation reactivity at relatively low tempera- tures (Carabineiro et al. 1999). The TGA data in Table 4 also

11 respondents also indicated that no landing gear components made from C/C composite were immune to catalytic oxida- tion. Stators and related components tend to be more prone to catalytic oxidation than rotors, although rotors will read- ily load catalytic contaminants when the aircraft is stationary (Walker 2007). Direction of contaminants to brake parts can be influenced by landing gear design; redesign of landing gear on existing aircraft, however, is not practical. Oxidation testing conducted jointly by the Center for Ad- vanced Friction Studies (CAFS) at Southern Illinois Univer- sity, Dunlop, Aircraft Braking Systems Corporation (ABSC), and Messier–Bugatti incorporated a range of materials, cata- lyst treatment methods, and heating temperatures (Table 6). Selected data regarding catalytic oxidation of unprotected C/Cs were presented at the SAE G-12 committee meeting in May 2007 (Filip 2007), and the results were predictably var- ied owing to differences in test methods employed in the lab- oratories (Figures 6 and 7). Weight loss percentages recorded by Messier–Bugatti were much higher than those recorded in the other two laboratories—perhaps as a result of differences in heating temperature and air—which merit further investi- gation. Nonetheless, there was a consistent trend in weight loss data confirming the catalysis of contaminants for C/C oxidation. Figures 6 and 7 also indicate that the different C/C composite materials (both virgin and catalyst-loaded sam- ples) experienced different levels of oxidation, which will be discussed later in this chapter. Dunlop and ABSC are currently conducting catalytic oxidation round-robin testing for the SAE A-5 committee (Hutton 2007), and preliminary data from both laboratories indicate a significant increase in the oxidation of C/C in the presence of KAc and KF. In the most recent round of tests, uniform C/C coupons (material type not specified) were pre- treated with 25% (by weight of solution) KAc, KF, or urea; heated at 550°C for 4 h; and then weighed. The coupons treated with KF and KAc showed significantly higher weight loss than the control coupons, whereas those treated with urea showed weight loss similar to that of the control coupons TABLE 5 BAe 146 AIRCRAFT BRAKE CONTAMINATION ANALYSIS Mean K Content per Sample (ppm) 547 2,183 74 749 991 563 Operator A B C D E F G Base of Operations (destinations) United Kingdom (Europe) Finland (Scandinavia, Europe) Greece (Greek Islands, S. Europe) Germany (Europe) Germany (Europe) Belgium (Europe) Belgium (Europe) No. of Aircraft 20 9 6 8 5 31 11 No. of Heat Packs Analyzed 101 44 42 51 18 93 32 991 Adapted from Hutton (2007). TABLE 6 TESTING PARAMETERS IN VARIOUS LABORATORIES Sample With a KAc-based PDP— Cryotech E36 Contamination Heating CAFS Cylinder, D = 10 mm, H = 8 mm, mass ~ 1.2 g Soaked for 60 min 700°C for 20 min, still air Dunlop Cylinder, D = 49.9 mm, H = 5.94 mm, mass ~ 20 g Soaked for 30 min and dried at 150°C for 120 min 550°C for 4 h, still air Messier–Bugatti 40 mm × 40 mm × 40 mm, mass ~ 50g Impregnated under vacuum for 60 min and dried at 105°C for 120 min 650°C for 5 h, constant O2 partial pressure CAFS = Center for Advanced Friction Studies.

12 eral mechanisms proposed to account for the catalytic effects of metals, oxides, and salts in carbon oxidation (Walker et al. 1968). At approximately 800°C the diffusion of oxygen through the surrounding gas to the carbon surface becomes the limiting step for thermal oxidation (Wu 2002). At typical peak operating temperatures for C/C aircraft brakes (near 500°C), the alkali-metal-salt-based PDPs (e.g., KAc) are known to decompose to alkali metal carbonates and oxides (e.g., K2CO3 and K2O). The active species (S) are believed to serve mainly as “more effective adsorption and dissociation agents for the gaseous reactant than carbon itself” and to transfer the adsorbed oxygen (O) to the carbon (C). They can go through an oxidation-reduction cycle, as represented here (Wu 2002): Such an oxygen-transfer mechanism has been supported by laboratory investigation pertinent to this subject. Environmen- tal scanning electron microscopy experiments demonstrated that K-oxide particles very effectively catalyzed the gasifica- tion of isotropic carbon fibers in a C/C composite. In situ X-ray diffraction experiments suggested that a K-peroxide acted as the reactive intermediate species (Carabineiro et al. 1999). Experimental observations and molecular orbital (MO) calculations supported Wu’s (2002) theory as follows: the presence of catalyst or inhibitor on carbon materials affects the oxidation behavior by influencing the concentration and sta- bility of two types of oxygen complexes on the carbon surface during the C–O2 reaction. S O C C O Sf( ) + → ( ) + ( )2 2 2 12S O S O+ → ( ) ( ) FIGURE 6 Selected results from carbon oxidation testing by Center for Advanced Friction Studies (CAFS) and Aircraft Braking Systems Corporation (ABSC)/Dunlop [adapted from Filip (2007)]. HWCCA = Carbenix™ C2000 2D pitch/charred resin, HWCCB = Carbenix™ C2000 2D-modified pitch/charred resin, and HWCCD = Carbenix™ 400 3D PAN/CVD. FIGURE 7 Selected results from carbon oxidation testing by Messier–Bugatti [adapted from Filip (2007)]. CAFS ABSC/Dunlop H W CC A pu re H W CC A ca ta lys t lo a de d H W CC B pu re H W CC B ca ta lys t lo a de d H W CC D pu re H W CC D ca ta lys t lo a de d 0% 2% 4% 6% 8% 10% 12% 14% 16% 18% W ei gh t l os s 0% 20% 40% 60% 80% 100% HWCCA HWCCB HWCCD W ei gh t L os s Pure Catalyst loaded (Figure 8). Although KAc- and KF-treated coupons experi- enced higher weight loss in the ABSC laboratory than in the Dunlop laboratory, the results demonstrated the roughly equal effectiveness of both potassium salt deicers as catalysts for carbon oxidation. Nature of the Effect of Modern Pavement Deicing Products on Aircraft Brakes Existing research in the laboratory has demonstrated the cat- alytic effects of potassium, sodium, and calcium on carbon oxidation (Lang and Pabst 1982; Krutzsch et al. 1996; Wu 2002). Oxygen transfer and electron transfer are the two gen-

13 The electron-transfer mechanism, on the other hand, sug- gests that catalysts [e.g., alkali metals (AM)] have unfilled electron shells and accept electrons from carbon matrix, as represented here (Filip 2007): This wetting ability (or lack thereof) is also relevant in the catalysis of C/C composites. The low melting temperatures AM O CO AM CO2 2 3+ → 2 5( ) 2AM CO e AM O CO2+ −+ + → +2 42 ( ) CO C CO 2e32− −+ → +2 3 3( ) of K salts and their decomposition products—all below 600K (or 327°C)—allow them to migrate easily on the carbon sur- face and form good interfacial contact with it (Wu 2002; Wu and Radovic 2005), facilitating oxygen transfer (Figure 9a). In some cases heavy catalyst loading may retard the activity of K species by “crowding” particles on the carbon surface (Figure 9b). Calcium oxide, with a melting point higher than 1500K (or 1227°C) (Wu 2002), relies on initial loading and impregnation to achieve the necessary surface contact. As oxidation proceeds, immobile calcium oxide may act as a barrier for additional catalysts, lowering the oxidation rate (Figure 9c). This retarded reaction recommends Ca-based PDPs as less detrimental to C/C aircraft brakes than K-based PDPs (Wu and Radovic 2005). FIGURE 8 Weight loss of C/C after heating at 550°C for 4 h: (a) Dunlop data; (b) ABSC data [adapted from Hutton (2007)]. 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% KAc KF Urea Control Treatment W ei gh t l os s 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% KAc KF Urea Control W ei gh t l os s Treatment (a) (b)

Several other less-studied factors also deserve attention. Temperature has been observed to play a role in governing the intensity of catalytic oxidation. Walker et al. (1997) observed a threshold of between 650°C and 700°C where the reaction rate of catalytic oxidation of C/C friction material in the presence of KAc drops off dramatically. Electrical con- ductivity may also be an important factor, as it appears to reflect catalytic potential in common pavement deicers. Con- ductivities of KAc and KF compared with that of urea appear to mirror relative catalytic abilities of these deicers (Table 7). Standards and Test Protocols SAE AMS 1431C and AMS 1435B are the accepted standards for solid and liquid pavement deicers, respectively. Neither currently contains requirements for C/C catalytic oxidation testing. The SAE G-12 Carbon Oxidation Working Group is in the process of refining a carbon compatibility test protocol with assistance from the SAE A-5A Brake Manufacturers Working Group for inclusion in the next revision of both standards. Boeing provides for testing of runway and facility deicers in its comprehensive test protocol for the Evaluation of Main- tenance Materials, Specification D6-17487. The current revi- sion of the protocol, released in 2003, specifies details for test- ing solid and liquid deicers, but does not address their roles in C/C composite catalytic oxidation. Boeing has no plans at this time to add catalytic oxidation testing to D6-17487, citing a requirement included in every Boeing brake assembly engi- neering drawing: “. . . that the brake be designed to be com- patible with different materials including runway de-icer flu- ids” [M. Arriaga, Boeing Company, personal communication, July 2007]. 14 In addition, the SAE A-5A Working Group is in the pro- cess of developing an oxidation test method for anti-oxidant (AO)-treated coupons and the details are summarized as fol- lows. C/C composite brake material will be cut from pro- duction discs and tested in the configuration of cylinders. A generic AO treatment based on mono-aluminum phosphate, phosphoric acid, and water (to be determined) will be used to simulate the application of the AO protection system. Although AO treatments are typically applied to nonfriction surfaces only, it is proposed to cover all surfaces of the test coupon to demonstrate the effect of the AO treatment. Dipping coupons in the AO treatment is proposed because it is less operator-sensitive than other application methods (e.g., brush- ing and spraying). The AO treatment will then be cured by heating the coupons at a ramp rate of 60°C/h to 300°C/h in air and at 300°C for 1 h. The AO treatment will be made cen- trally before the test coupons are distributed to test facilities. To test the catalytic oxidation, runway deicing solutions will be used at 25% w/w concentration. The weight loss of the AO-treated coupons will be tested under the same tempera- ture for the testing of bare carbon; that is, 1022°F ± 10°F or better (550°C ± 5°C) in still air (SAE A-5A . . . 2007). As discussed previously, there were large variations ob- served in the carbon oxidation testing results from different laboratories. These variations highlight the need to develop reliable, standard testing procedures to evaluate the catalysis of PDPs for carbon oxidation, which would allow better practices in preventing or mitigating such catalysis. To ensure reproducible results, protocol parameters to be defined will include sample material, density, dimensions, coupon orienta- tion, contamination concentration, temperature, duration, post- treatment, heating temperature, and ambient air velocity. Prevention and Mitigation There are potential opportunities for all stakeholder groups to collaborate in addressing the catalytic oxidation issue of C/C aircraft brakes with respect to aircraft and component design, brake testing, aircraft operations, airfield maintenance, etc. In the domain of brake technologies, chemical modification of C/C appears to offer greater potential than structural changes or defect elimination in mitigating catalytic oxidation (Filip 2007). Chemical modification generally involves the intro- duction of groups of atoms to reduce gasification of carbon, the reduction of catalyst mobility, and the formation of a FIGURE 9 Catalyst loading and activity on brake surface [adapted from Wu and Radovic (2005)]. TABLE 7 ELECTRICAL CONDUCTIVITIES OF PAVEMENT DEICERS (b) intermediate catalyst activity(a) high catalyst activity (c) low catalyst activity carbon surface catalyst particles active catalyst inactive catalyst catalyst layer Type Electrical Conductivity 1% Solution (µS/cm) Urea 18 KAc ~7000 KF ~9000

15 barrier for transport of oxygen and reaction products (Filip 2007). Defect elimination generally involves the reduction of C/C composite porosity and the elimination of active oxida- tion sites (by means of improved crystalline order of carbon and reduced defects) (Filip 2007). A firm understanding of the catalysis mechanism has nurtured development of various proprietary AO formulations tailored to material, environ- mental, and performance needs of specific brake and wheel designs. It is also recognized that catalytic oxidation of C/C composite friction material by some alkali-metal-salt-based PDPs—such as those based on KAc and KF—cannot be fully arrested in situ by current methods. Dozens of U.S. and European patents have been assigned to formulae and methods for blocking active oxidation sites on and below the C/C composite brake material surface. The oxi- dation inhibiting composition usually penetrates at least some of the pores of the C/C composite and, once heated, forms a deposit within the penetrated pores and the surface of the C/C composite. For instance, Stover and Dietz (1995) and Stover (1998, 2003) created AO formulations primarily of phosphoric acid, metal phosphates, and aluminum and zinc salts in a polyol/alcohol base. Walker and Booker (2000) demonstrated the effectiveness of P-13 (a standard phosphoric-acid-based AO) and a potassium compound (KH2PO4) for inhibiting cat- alytic oxidation by KAc by blocking active oxygen transfer sites on the surface of C/C friction material. It was proposed that the addition of KH2PO4 blocked sites on the carbon sur- face that were particularly prone to “activation” by K cataly- sis. Phosphate-based AO paints are now a standard, effective tool in reducing catalytic oxidation (T. Walker, Honeywell International, personal communication, July 2007). Wu et al. (2001) and Wu (2002) explored inhibition by phosphorus (P)- and boron (B)-deposition and B-doping in fine detail and reported several key findings. Data sug- gested the presence of two catalyst-deactivation mechanisms. Surface-deposited P and B compounds were found to block (with varying success) catalysts from contact with active oxi- dation sites on the carbon substrate. Thermally deposited P compounds were demonstrated to be effective in inhibiting the carbon oxidation catalyzed by KAc and calcium acetate (CaAc); and the characterization of P-deposited carbon sam- ples and ab initio molecular orbital calculations both suggested that the inhibition effectiveness derived from the formation of possibly C-O-PO3 groups and C-PO3 groups (Wu and Radovic 2006), which preferentially block the active carbon sites (Wu 2002). The effect on K catalysis was much smaller owing to the high wetting ability and mobility of K species. Also sug- gested was the possibility that in sufficient concentration, these deposition compounds form stable oxide glazes over the friction surface, acting as an oxygen barrier. B-doping pro- moted better graphitization of the C substrate, denying free electron sites to catalysts. A secondary benefit to B-doping was the lower curing temperature (by 400°C–500°C) needed for satisfactory graphitization. Wu (2002) suggested that a combined use of P and B might offer more effective inhibi- tion of catalytic carbon oxidation than used individually (Wu 2002). Emphasizing the impossibility of eliminating K cata- lysis of C/C composites, an inhibition system employing a combination of painted or ceramic coatings on nonfriction surfaces, B-doped C/C substrate, and deposited P and/or B2O3 was suggested. Application of the laboratory research has been met with mixed success. Industry experiments with B-doping have not shown promise (T. Walker, Honeywell International, per- sonal communication, July 2007), although elemental B has been used successfully as a barrier coating. Brush-on phos- phate AO coatings continue to be widely employed, with peri- odically improved formulations. These AO treatments typi- cally include multiple cycles of a brush-applied or sprayed phosphate- or phosphoric acid-based coating on the C/C material, followed by high-temperature curing. In addition, a ceramic-based oxygen barrier coating is applied to non- friction surfaces (Webb 2007). C/C composites composed of three-dimensional nonwoven fabric with preform chemical vapor deposition matrix have generally replaced those com- posed of chopped polyacrylonitrile (PAN) or pitch fibers, owing to the increased load-bearing and thermal properties of the former. Combinations of these advances have resulted in marked reductions in unscheduled replacements owing to cat- alytic oxidation. Materials-deicer compatibility testing conducted by the Concurrent Technologies Corporation on behalf of the Air Force Research Laboratory in 2003 and 2004 included catalytic oxidation testing of four Honeywell C/C friction materials and three pavement deicers that were then new to the market or still in development (Concurrent Technologies Corp. 2004). For each contaminant, ten specimens—five brushed with Honeywell P-13 phosphate-based AO compound— of each friction material were soaked in deicer or deionized water for 20 min and dried in still air at 110°F (43°C) for 30 min. Specimens were then heated for two 4-h periods at 1300°F (704°C) and allowed to cool to room temperature in still air after each session. Three of the C/C materials were composed of chopped pitch or PAN fiber with a phenolic char chemical vapor deposition matrix, whereas the fourth material, Carbenix™ 4000, consisted of 3D needled, non- woven PAN fabric with a chemical vapor deposition preform matrix. As shown in Figures 10 and 11, unprotected 3D PAN samples suffered weight and hardness losses similar to those of the other unprotected friction materials. In contrast, AO- treated 3D PAN samples experienced the lowest weight and hardness losses in the group. This observation suggests that even though the 3D material showed no inherent resistance to catalytic oxidation, one or more of its unique traits improved the performance of the AO surface treatment. Although the CTC report did not offer explanations for the favorable per- formance of the AO-treated 3D C/C composite material, implementation of this improved material in the field in

concert with new AO systems has demonstrated similar margins of success. New brakes with a 3D PAN preform substrate and improved AO protection fitted aboard Boeing 767s led to a 90% reduction in brake removals before end- of-service life on those aircraft, and 3D PAN substrates have become the standard substrate on new C/C compos- ite brake designs (Walker 2007). Solid pavement deicer formulations were endorsed by sev- eral ACRP survey respondents as less aggressive catalysts, although no specific justification was provided. Catalytic oxidation of C/C brakes may also be mitigated by using more carbon-friendly PDPs on airfield pavements. Anhydrous betaine (N,N,N-trimethylglycine), a naturally occurring organic byproduct of sugar beet processing, is being developed with support from the Finnish Government and Finavia as a pave- ment deicer (Hänninen 2006; Simola 2006). Not yet com- mercially available, betaine has shown favorable results in metallic corrosion and deicing performance testing, and its low electrical conductivity also compares favorably with other deicers (Table 8). Betaine’s relatively high BOD, nitro- gen content (15% by weight), and high cost may present some challenges to using it as the sole freezing point depres- sant for PDPs. 16 As early as 2002, a U.S. patent was under review for an aqueous liquid aircraft runway deicer composition featuring minimal catalytic oxidation effect on C/C composites. The composition contains 20%–25% w/w of an alkaline earth metal carboxylate, 1%–15% w/w of another alkaline earth metal carboxylate, 1%–35% w/w of an aliphatic alcohol, 0.01%–1% of an alkali metal silicate, and up to about 1% w/w of a triazole (Moles et al. 2002). Through partnership with DuPont Tate & Lyle Bio Products LLC, this evolved into a commercial product marketed by Cryotech (BX36), which now includes a bio-based active ingredient Susterra propanediol. Joint efforts by Honeywell and Cryotech led to preliminary testing of BX36, which showed less conductivity FIGURE 10 Average weight loss of (a) unprotected; (b) AO- treated C/C friction materials exposed to KF-based deicers and de-ionized water [adapted from Concurrent Technologies Corp. (2004)]. (Note: Carbenix™ is a Honeywell proprietary C/C material.) FIGURE 11 Average hardness loss of (a) unprotected; (b) AO- treated C/C friction materials exposed to KF-based pavement deicers and de-ionized water [adapted from Concurrent Technologies Corp. (2004)]. (Note: Carbenix™ is a Honeywell proprietary C/C material.) A ve ra ge w ei gh t l os s ( %) 0 10 20 30 40 50 60 70 80 90 Carbenix 1000 Carbenix 2000 Carbenix 2110 Carbenix 4000 C/C friction material De-ionized water Safeway KF Hot AVIFORM L50 METSS RDF-2 (a) (b) A ve ra ge w ei gh t l os s ( %) 0 10 20 30 40 50 60 70 Carbenix 1000 Carbenix 2000 Carbenix 2110 Carbenix 4000 C/C friction material De-ionized water Safeway KF Hot AVIFORM L50 METSS RDF-2 0 10 20 30 40 50 60 70 80 90 Carbenix 1000 Carbenix 2000 Carbenix 2110 Carbenix 4000 C/C friction material A ve ra ge h ar dn es s l os s ( %) De-ionized water Safeway KF Hot AVIFORM L50 METSS RDF-2 (a) (b) 0 10 20 30 40 50 60 70 80 90 Carbenix 1000 Carbenix 2000 Carbenix 2110 Carbenix 4000 C/C friction material A ve ra ge h ar dn es s l os s ( %) De-ionized water Safeway KF Hot AVIFORM L50 METSS RDF-2 TABLE 8 ELECTRICAL CONDUCTIVITY AND BIOLOGICAL OXYGEN DEMAND OF DEICERS Deicer BOD5 (mg O2/g) Betaine 759 KAc ~300 KF ~100 Urea Chemical Formula C5H11NO2 CH3COOK HCOOK H2NCONH2 Electrical Conductivity of 1% Solution (µS/cm) 99 ~7,000 ~9,000 18 ~2,100

17 and more than 50% less catalytic activity and passed all AMS 1435A deicing criteria, including the proposed corro- sion criteria (Boeing Method) (Walker 2007). Knowledge Gaps Although the fundamental mechanisms of catalytic oxidation by PDPs are well understood in well-controlled laboratory settings and advances in technologies for its prevention and mitigation have been made in the last decade or so, the prob- lem appears far from solved. Action in the following areas may be beneficial for further advances. There is still a need to establish a comprehensive PDP cat- alytic oxidation test protocol. To this end, a test protocol has been in development in the SAE G-12 Working Group since early 2003 and is currently being refined for inclusion to AMS 1431 and 1435. Incorporation of such a test protocol (includ- ing a conductivity test as suggested by several ACRP survey respondents) into AMS 1431C and 1435B will provide nec- essary guidance for developing the next generation of PDPs and C/C aircraft brakes. The proprietary nature of PDP and C/C aircraft brake technologies may hinder the development of such a test protocol, and the ever-changing nature of these technologies may entail continued efforts in updating the test protocol. Furthermore, more research is needed to better understand relationships between brake design, AO treatment, and PDP contamination as factors in catalytic oxidation. PDP devel- opment is still an active field, and new products will continue to be introduced to the market. AO treatments designed to mitigate catalytic oxidation by PDPs are still immature and mostly proprietary. CADMIUM CORROSION Cadmium (Cd) had been the standard for protection of steel parts on aircraft wheels and brakes even before the 1980s. Cd- plating is the most popular surface treatment technology for corrosion protection of aircraft steel parts (e.g., airframe com- ponents and fasteners), which is of great importance to flight safety and aircraft durability. This is attributable to the unique combination of its excellent corrosion protection properties in traditional service environments and its other service charac- teristics. Cd-plating serves as a highly effective barrier coat- ing, especially in the marine environments often experienced by aircraft. It also serves as a sacrificial coating to protect steel and features nonvoluminous corrosion products. Cadmium offers better corrosion resistance and a greater immunity domain than zinc (Badawy and Al-Kharafi 1998). Cadmium is also galvanically compatible with aluminum alloys (Baldwin and Smith 1996). Other attractive properties of Cd- plating include its good conductivity and surface lubricity, high ductility, solderability, and potential to be repaired in the field. The main drawback of Cd-plating is the high toxicity asso- ciated with Cd and its compounds. Cadmium can accumulate in the human body with acute or chronic exposure, and even- tually lead to softening of the bones and kidney failure in humans and many animals (“Metals as Toxins—Cadmium” 2007). In addition, the Cd-plating process often involves the use of toxic cyanide baths and the process itself can weaken steel components through hydrogen embrittlement if post- application precautions are not taken. Despite the disadvantages of Cd-plating and a large body of research on its alternatives (Smith 1992; Baldwin and Smith 1996; Thomson 1996; Zhirnov et al. 2003), the desirable qual- ities of Cd-plating have yet to be matched or exceeded in a single alternative. Field reports increasingly suggest that the contact with modern PDPs (such as potassium acetate- and formate-based products) promotes damage to aircraft components, including those that are Cd-plated. In April 2002, the Aerodrome Safety Branch under Transport Canada issued an Aerodrome Safety Circular, recommending that airport operators refrain from using deicing fluids containing KF on airside movement areas. The recommendation was based on a Boeing Service Bulletin indicating that all B737-600, -700, -700C, -800, and -900 air- plane models were prone to suffer KF-promoted corrosion of electrical connectors located in the wheel well. In August 2004, Transport Canada cancelled the Circular, based on new evidence that the problem appeared to be limited to the Boeing 737, but suggested that airport operators inform air carriers serving their airport of the PDPs used on airside move- ment areas (Transport Canada . . . 2004). In September 2005, the FAA updated an existing airworthiness directive that applies to all Boeing B737-600, -700, -700C, -800, and -900 airplane models. The existing directive required “either determining exposure to runway deicing fluids containing KF, or performing repetitive inspections of certain electrical con- nectors in the wheel well of the main landing gear for corro- sion and follow-on actions.” The amendment was prompted by anecdotal evidence showing similar corrosion effects of KAc- based PDPs and added a new inspection requirement and related corrective actions. The goal was to “prevent corrosion and subsequent moisture ingress into the electrical connectors, which could result in an electrical short and consequent incor- rect functioning of critical airplane systems essential to safe flight and landing of the airplane” (“Airworthiness Direc- tives . . .” 2005). The U.S. Air Force has found that KAc-based runway deicing fluids caused numerous problems with its air- craft components, mainly electronics (e.g., failure of switches and wire harnesses), likely owing to high conductivity of the deicers (“Runway Deicing . . .” 2007). Cadmium corrosion has been observed in Continental Air- lines (CO) and Scandinavian Airlines System 737-NG and CO EMB-145 MWW (main landing gear wheel well) electrical connectors, MWW components, and air conditioning bay packs. Aluminum corrosion has been observed in CO 737-NG

and other airline 737-Classic MWW and wing aluminum hydraulic lines. The PDPs were also suspected to cause the premature corrosion of landing gear joints, accelerate the de- gradation of electrical wire harness insulation, and promote the corrosion of aluminum belly skin (Duncan 2006). Similar effects have also been observed on ground support equipment (GSE) units. Among them, the foremost concern has been the effect of modern PDPs on Cd corrosion, although the other problems are more anecdotal and are more easily mitigated through better aircraft design or maintenance practices. As such, the rest of this section synthesizes the information on the validity and nature of the effect modern PDPs (mainly alkali-metal-salts) on Cd corrosion, describes the related stan- dards and test protocols, discusses ways to prevent and miti- gate such effect, and identifies pertinent knowledge gaps. Validity of the Effect of Modern Pavement Deicing Products on Cadmium Corrosion Until recently, the principal evidence connecting alkali-metal- salt-based PDPs with Cd-plating corrosion has been a trend of increased reports of the latter occurring simultaneously with the introduction of the former (ACRP survey; Duncan 2006). In the United States, the introduction of alkali-metal-salt-based PDPs in recent years coincided with a rise in the number of reported cases of failed or replaced aircraft components result- ing from Cd corrosion (Duncan 2006). A majority of the rele- vant reports involved mechanical and electrical connectors accessible to runway deicers through spraying or splashing, such as main landing gear wheel wells and air conditioning bays. Scandinavian and Northern European airports saw wide- spread corrosion of Cd-plated components and electrical fail- ures on 737-NG aircraft in the 2000/2001 winter season, which Boeing attributed to their exposure to KAc- and KF-based PDPs (Hunter 2005). It is interesting to note that all PDPs used at these airports passed the Cd-corrosion test required by AMS 1435 and AMS 1431. The test protocol—ASTM F1111—involves the continuous immersion of Cd-plated steel specimens in the PDP for 24 h, which was later consid- ered unable to simulate the field exposure conditions. Boeing initiated round-robin testing based on ASTM F1111 Cd-plated steel specimens and a custom-designed Cd cor- 18 rosion test protocol involving cyclic immersion in the PDP for 31 days (instead of ASTM F1111, as discussed later). With the first round of tests since 2005, participating Euro- pean and U.S. laboratories observed generally consistent mass gain patterns in Cd-plated coupons in the presence of KF (Nicholas 2007), and the experiment was redesigned and the second round of tests began in 2007. The material used was changed from F1111 Cd-plated steel to AMS QQ-P- 416B Cd-plated steel. The test environment was adjusted to 90°F (32.2°C) and 30% relative humidity, and the test dura- tion was reduced from 31 to 14 days. The number of brush strokes on the specimen was regulated at 12 stokes per side. Methanol was used to dry the specimen to ensure moisture removal. Data indicated a very reliable test within the same laboratory, but variations between laboratories. The latter was likely derived from deviations in detailed procedures (Nicholas 2007). The only anomalous weight gain (see Table 9, highlighted in bold) was attributed to an unidenti- fied deviation in procedure. The next round of testing was planned for late 2007 and was to be conducted under signif- icant changes in experiment design to further reduce vari- ability between laboratories and to incorporate a fluid to simulate KAc-based PDPs (along with a KF-based fluid and urea as control) into the testing scheme. Nature of the Effect of Modern Pavement Deicing Products on Cadmium Corrosion Because corrosion of metals is an electrochemical process, the thermodynamics of metallic corrosion is generally gov- erned by the combination of pH of the electrolyte and electro- chemical potential of the metal in the electrolyte. For the Cd- water system at 25°C, its potential-pH equilibrium diagram was established as early as the 1960s (Deltombe et al. 1966). Such theoretical predictions of corrosion and passivity were experimentally validated by testing the dissolution of Cd in solutions of pH 1–15. Consistent with the potential-pH equi- librium diagram, the immersion test results indicated signifi- cant corrosion of Cd in both acid (pH < 7) and highly alkaline (pH > 12) solutions and an intermediate region of a low cor- rosion rate of Cd near a pH of 12 (Tomlinson and Wardle 1975). It can be reasonably assumed that Cd in the first two solutions dissolved to Cd2+ and HCdO2−, respectively, whereas in the intermediate region it formed a passive layer of TABLE 9 WEIGHT CHANGE OF CD-PLATED STEEL AFTER ROUND-ROBIN TESTING IN KF, SPECIMEN SIZE 25.4 MM × 50.8 MM × 1.22 MM Specimen Type Weight Change of Specimens (g) ASTM F1111-02 AMS QQ-P-416B Lab A 0.0114 0.0109 Lab B 0.0919 0.0707 Lab C +0.057 0.0155 Lab D 0.0077 0.0132 Lab E 0.0618 0.0955 Adapted form Nicholas (2007).

19 Cd(OH)2. The formation of Cd(OH)2 or CdOH2O in neutral and alkaline solutions has been confirmed and this passive film may be unstable in highly alkaline solutions through the following reactions (Badawy and Al-Kharafi 2000): In addition to pH, both the dissolved oxygen and tempera- ture are expected to have an influence on the kinetics of metal- lic corrosion. The corrosion of Cd in water at pH 8.3–10.55 was found to proceed under cathodic control through the reduction of oxygen. The corrosion rate of Cd was reduced substantially by limiting the available oxygen in water or increasing concentrations of OH− and CO32− species (Posselt and Weber 1974). In another study, however, the presence of oxygen was found to passivate the Cd surface in neutral and alkaline solutions (Badawy et al. 1998). The electrochemical impedance spectroscopy data indicated the presence of two phase maxima in neutral solutions, signifying two consecutive charge transfer reactions with different time constants occur- ring at the Cd/electrolyte interface (Badawy et al. 1998). Sim- ilar to most chemical reactions, the rate of Cd corrosion usu- ally increases with temperature. Very little research has been conducted to investigate the mechanism of Cd corrosion or Cd-steel corrosion in the pres- ence of alkali-metal-salts (e.g., KF and KAc), partly owing to the high toxicity associated with Cd and its compounds. It is known that the corrosion properties of Cd resemble those of zinc in the range pH 8–11, except for the higher corrosion resistance of Cd (Posselt and Weber 1974). As such, a recent laboratory study conducted at the Western Transportation Cd OH OH Cd OH( ) + → ( )− − −3 42 7( ) Cd OH OH Cd OH( ) + → ( )− −2 3 6( ) Institute might shed some light on the PDP effect on Cd-steel corrosion (Fay et al. 2007, with expanded dataset). The corro- sion behavior of mild steel (ASTM A36) and galvanized steel (highway guardrail) was studied using various deicer products and analytical-grade KAc. For deicers diluted at 3% by weight or volume (for solid and liquid deicers, respectively), electro- chemical testing of their corrosion to mild steel and galvanized steel showed that the acetates and formates (except the solid analytical KF) were much less corrosive to mild steel than the chloride-based deicers. Steel is considered to be passive when its corrosion current density icorr < 0.1 µA/cm2, and active cor- rosion occurs when icorr > 1.0 µA/cm2. As such, it can be con- cluded that the acetates and formates tested (except the solid analytical KF) were noncorrosive to mild steel, whereas the chloride-based deicers were very corrosive (as shown in Table 10). Nonetheless, the galvanized steel in the acetates and formates (except the solid NaF-based product) was found to be corroding at comparably high rates, as seen in the chloride- based deicers. The corrosion potential (Ecorr) data shown in Table 10 indicate that acetates and formates (except the solid analytical KF) significantly shifted Ecorr of mild steel to the noble direction, but failed to do so with Ecorr of galvanized steel. The latter might be attributed to the presence of sacrifi- cial zinc in the galvanized steel, which likely changed the potential-pH equilibrium of the steel and moved the metal from a passive state to an active corrosion state. It can be assumed that the sacrificial Cd in the Cd-plated steel plays a similar role to the sacrificial zinc in the galvanized steel when exposed to alkali-metal-salt deicers. A few mechanisms may now be proposed that may be re- sponsible for the effect of modern PDPs on the corrosion of Cd-steel aircraft components. First, PDP residues may be highly concentrated on localized areas of aircraft, owing to TABLE 10 ELECTROCHEMICAL ANALYSIS OF DEICER EFFECT TO MILD STEEL (ASTM A36) AND GALVANIZED STEEL (Guardrail) Note: 1 MPY = 1 milli-inch per year = 0.0254 mm per year. Deicer icorr(µA.cm2) MgCl2 (liquid) 6.0 ± 2.5 Salt/Sand 5.4 ± 1.3 KAc (liquid) KAc (solid, analytical) NaAc (solid) NaF(solid) 5.5E-03 ± 2.0E-04 5.5E-03 ± 6.0E-04 6.8E-02 ± 9.3E-02 5.5E-03 ± 6.0E-04 KF (solid, analytical) 19.8 ± 23.9 MgCl2 (liquid) 3.5 ± 0.6 Salt/Sand 1.6 ± 0.1 KAc (liquid) 3.0 ± 0.9 KAc (solid, analytical) 4.4 ± 1.3 NaAc (solid) 1.8 ± 0.3 NaF(solid) 0.2 ± 9.1E-02 KF (solid, analytical) Corrosion Rate (MPY) 2.7 ± 1.1 2.5 ± 0.6 2.5E-03 ± 9.1E-05 2.5E-03 ± 3.0E-04 7.1E-03 ± 4.1E-03 2.5E-03 ± 2.1E-04 8.5 ± 11.5 1.7 ± 0.2 0.8 ± 2.0E-02 1.7 ± 0.6 1.8 ± 1.2 0.9 ± 0.2 5.2E-02 ± 5.9E-02 1.6 ± 0.9 Ecorr(mV, vs. SCE) 616.0 ± 1.8 764.3 ± 6.0 155.3 ± 30.2 132.3 ± 13.3 204.3 ± 68.6 199.5 ± 12.0 598.0 ± 316.5 1037.5 ± 5.0 1047.5 ± 5.0 1032.5 ± 5.0 1050.0 ± 0.0 1035.0 ± 19.1 1003.3 ± 8.3 1060.0 ± 0.0 4.1 ± 0.6 M ild S te el G al va ni ze d St ee l

the hydroscopic nature of these PDPs (e.g., KAc and NaF). This may lead to localized high alkalinity that can disrupt the passive film of Cd(OH)2 by means of the reactions in Eqs. 6 and 7. Second, the potential corrosion products (CdAc and CdF) are highly soluble in water, which may facilitate the cor- rosion of Cd-steel. Third, Cd serves as a sacrificial anode to protect the steel components and the corrosion of Cd may be accelerated by the high conductivity derived from its con- tamination by PDPs and the increasingly smaller area ratio of anodic sites (Cd) to cathodic sites (steel). Finally, the pres- ence of PDPs might promote the hydrogen embrittlement of Cd-plated steel. It should be cautioned no conclusions should be drawn about the corrosivity of PDPs without stating the specific deicer product and its concentration, the test protocol used (SAE, ASTM, National Association of Corrosion Engineers, or electrochemical test), and the type of metal tested. Standards and Test Protocols In May 2002, ASTM Committee F07 on Aerospace and Air- craft published an updated version of ASTM F1111-02, Standard Test Method for Corrosion of Low-Embrittling Cadmium Plate by Aircraft Maintenance Chemicals. As discussed earlier in this chapter, the testing parameters of ASTM F1111-02 have been considered insufficient for dis- criminating PDPs for their corrosivity, likely resulting from the relatively mild temperature (95°F or 35°C) and short, continuous immersion period (24 h). In response to this defi- ciency, Boeing developed its current Cd corrosion test (dis- cussed later) and integrated it into the broader Boeing test protocol. Similar to those found in ASTM F1111-02, the plating specifications contained in the Boeing Cd corrosion test are intended to be used only for evaluation purposes and differ from those in AMS QQ-P-416 B, the accepted standard for electrodeposited Cd-plating in aerospace applications. AMS QQ-P-416 B references separate ASTM and NASA standards for determining corrosion resistance to salt spray, but does not set forth parameters for general corrosion test- ing such as those specified in ASTM F1111-02 or the Boeing protocol. Boeing Document D6-17487, Evaluation of Airplane Maintenance Materials, contains a section (§20) specific to measuring the corrosivity of runway deicers to Cd-plating. The Boeing test sample specification was derived from the ASTM F1111 specification using a specimen size of 25.4 mm × 50.8 mm × 1.22 mm. The Boeing test protocol uses a 31-day cyclic immersion of Cd-plated steel in the electrolyte, instead of a 24-h continuous immersion used in the ASTM specification. Using this protocol, Cd-plated steel specimens were exposed to eight runway de-icing fluids and three con- trol fluids, with three replicates in each solution. The test pro- tocol was demonstrated to be capable of distinguishing the corrosivity of PDPs to Cd-plated steel, with a KAc-based PDP and a KF-based PDP being the most and the least corro- sive, respectively (Hunter 2005). 20 The Boeing standard limits the Cd weight loss owing to corrosion to no more than 0.03 mg/cm3, equivalent to no more than a 0.0077 g loss from a single test coupon. Because the majority of Cd weight loss occurs in the first 14 days of test- ing, Boeing reduced the test duration from 31 days to 14 days and plans to establish a new weight loss threshold level based on the revised test method. Currently, the Boeing Cd corrosion test protocol is being modified. As discussed earlier, round-robin testing data in- dicated good consistency within the same laboratory, but revealed large variations between different laboratories. These variations highlight the need to develop reliable, stan- dard testing procedures that can be used to evaluate the cor- rosivity of PDPs to Cd-plated steel, which would allow better practices for preventing or mitigating such corrosion. Details of the test protocol such as the Cd-plated steel material, dimen- sions, and configuration; specimen pretreatment and post- treatment; and testing environment (relative humidity, temper- ature, etc.) should be well-defined and controlled to ensure repeatable and reproducible results. Prevention and Mitigation There are potential opportunities for all stakeholder groups to collaborate to prevent and mitigate the effects of PDPs on air- craft components from aspects of aircraft and component design, aircraft operations, and airfield maintenance. Installa- tion of TR (engine thrust reverser) cascades with improved design, replacement of Cd-plated connectors with stainless steel and anodized aluminum connectors, and application of corrosion-inhibiting compounds (CICs) are suggested by Boeing (Duncan 2006). Frequent inspection or online moni- toring of corrosion-prone components, although costly, is another way to mitigate corrosion of Cd-plating and aluminum corrosion. In the domain of CIC technologies there is still great potential for improvement. At this stage, the research has to build on the existing knowledge base of inhibiting Cd cor- rosion in the absence of PDPs. For instance, some quino- line derivatives (quinaldic acid, oxine, 2-methyloxine, and oxine-5-sulfonic acid) were found to form stable chelate compounds with Cd and thus inhibit the general corrosion of Cd (Kato et al. 1973). Quinoline, however, is a known hazardous air pollutant. Precoating the Cd surface with CO32− or sodium metasilicate was reported to greatly reduce its corrosion rate (Posselt and Weber 1974). Triazoles are known to be effective CICs, but they have been banned in most Northern European countries owing to toxicity concerns (Duncan 2006). Methanol, ethanol, isopropanol, and n-propanol were found to inhibit Cd corrosion in aqueous solution, and the electro- chemical testing indicated their corrosion inhibition effi- ciency at 0.1 M to be 29.9%, 37.8%, 39.3%, and 98.6%,

21 respectively. The inhibition mechanism validated by surface analysis was proposed as follows. The alcohol molecules absorb on the active Cd sites by means of functional −OH groups and stabilize the passive film, with their hydrophobic alkyl tails limiting the access of electrolyte to the metal sur- face. A concentration of ≥0.075 M of n-propanol in an alka- line solution (pH = 13) achieved 97% reduction in weight loss after a 180-min exposure. It should be noted that the experi- ment employed highly polished, spectroscopically pure Cd rods rather than AMS- or ASTM-compliant Cd-plated steel (Badawy and Al-Kharafi 1998, 2000). Commercially available CICs for application to fasteners and other exposed metal are available in wipe-on, brush-on, and spray-on types (Groupe Meban 2007). Most employ an adsorption mechanism with an active ingredient(s) similar to n-propanol as described earlier, which blocks the active sites on the metal surface. As a secondary effect, the CIC base may then harden or dry on the metal surface, forming a temporary but durable physical barrier to salt spray, cleaners, and other contaminants. There is scant—if any— relevant research available to the public on the effectiveness of these CICs in mitigating Cd-plating corrosion, especially in the presence of PDPs. In lieu of a comprehensive prevention solution to Cd- plating corrosion or a satisfactory Cd-plating replacement, shop-level mitigation practices such as additional and en- hanced maintenance and inspection should help reduce the effects of PDPs on corrosion-prone Cd-plated steel aircraft components. Such best practices would also minimize the impact of PDPs on other aircraft components. In addition, the corrosion of aircraft components (e.g., Cd- plating and aluminum parts) can be mitigated by utilizing less corrosive PDPs on airfield pavements. U.S. patents were granted in 2001 (USP 6287480), 2003 (USP 6623657), and 2005 (USP 7063803) for deicing formulae based on potas- sium succinate, succinic anhydride, and succinct acid. At least one potassium lactate formula has been marketed in the United States, but only to military facilities (Shi 2007). Such deicing compositions are claimed to be suitable and effective for airport applications in which corrosion of aircraft alloys and concrete runways are of concern, but they have not been tested at commercial airports. In addition, no testing in rela- tion to C/C catalytic oxidation is known to have been con- ducted on these formulae. PDPs with low electrical conductivity have been sug- gested to potentially pose less risk for aircraft components as well as C/C composite brakes. PDPs based on betaine (a bio- based freezing point depressant and corrosion inhibitor with low electrical conductivity) have been developed in Finland and are qualified for AMS 1435A certification owing to low corrosivity. In addition to its high cost, however, betaine’s relatively high BOD and nitrogen content may present some challenges, as discussed earlier (Duncan 2006). The U.S. Air Force suggested adding the solution conduc- tivity test (ASTM D1125) to AMS 1431 and AMS 1435 as a required test for PDPs. In addition, other tests such as wet arc propagation resistance (SAE AS4373, Method 509), immer- sion volume swell (SAE AS4373, Method 601), bend (SAE AS4373, Method 712), and voltage withstand (SAE AS4373, Method 510) were suggested to be added to AMS 1431 and AMS 1435 if possible (USAF Research Laboratory 2007). Knowledge Gaps Although the fundamental mechanisms of Cd corrosion in water are relatively well studied, the link between alkali-metal- salt-based PDPs and Cd-plating corrosion has yet to be exper- imentally validated and thoroughly investigated. Action in the following areas may be beneficial for further advances in miti- gating the effects of PDPs on Cd-plating and aircraft alloys in general. First, there is still a need to establish a comprehensive metal- lic corrosion test protocol for PDPs. To this end, a Cd corro- sion test protocol has been in development in the SAE G-12 Working Group since 2003 and is currently being refined for inclusion to AMS 1431 and AMS 1435. Incorporation of such a test protocol (including a conductivity test as suggested by several ACRP survey respondents) into AMS 1431C and AMS 1435B will provide necessary guidance for developing the next generation of PDPs and aircraft components. The pro- prietary nature of PDP and aircraft components may hinder the development of such a test protocol, and the ever-changing nature of these technologies may entail continued efforts in updating the test protocol. For instance, the need for an accept- able alternative to Cd-plating has led to extensive research on this subject and several promising alternatives such as Zn- Ni-P (Veeraraghavan et al. 2003), Zn-Sn-P (Zhirnov et al. 2006), and Zn-Ni (Thomson 1996; Claverie and Chaix 2007). Second, more research is needed to better understand the interactions among the aircraft component design, the CICs used, and the contamination of PDPs in the processes of metal- lic corrosion. This is further complicated because the use of Cd-friendly PDPs is still in the burgeoning stage and new products will be continually introduced to the market. Simi- larly, CICs designed to mitigate Cd-plating corrosion by PDPs are still immature and mostly proprietary. Furthermore, field corrosion of metals may be affected by component design and exposure conditions, and various other mechanisms that are unique to the operational environment (e.g., galvanic corro- sion, pitting corrosion, crevice corrosion, stress corrosion cracking, corrosion fatigue, erosion corrosion, and microbially influence corrosion). Finally, there is still a lack of academic research data from controlled field investigations regarding the aircraft metallic corrosion by PDPs, which would help differentiate the con- tribution of PDPs to such corrosion from other possible

contaminants in the field environment. One challenge is that aircrafts are moving objects exposed to various PDPs and other contaminants, climatic conditions, and maintenance practices at different airports. For instance, Cd-plated steel was found to be affected by paint vapors (Brough 1987), by microorganisms in hydraulic fluids (Weyandt and Schweisfurth 1989), and by aircraft fuel additives (Hinton and Trathen 1992). INTERACTION WITH AIRCRAFT DEICING AND ANTI-ICING FLUIDS The SAE Aircraft Deicing/Anti-icing Fluid (ADAF) Specifi- cations provide guidelines for the holdover time of Types I, II, III, and IV fluids (“FAA-Approved . . .” 2007). To meet these criteria, Types II, III, and IV fluids currently used for aircraft anti-icing contain thickeners to keep these fluids on surfaces after application. These thickeners are gel polymer additives known to gradually precipitate out of solution and form dry residues that can remain in aerodynamically quiet areas of the aircraft for long periods. If not discovered, these residues can accumulate over time, rehydrate and expand in rain or aircraft washes, and freeze during cold weather or high altitude flight. This can negatively affect in-flight han- dling of the aircraft if deposits occur on or near control sur- faces or linkages. Initial research has shown that thickener separation is accelerated by contact between aircraft deicing fluids and runway deicing fluids (Ross 2006; Hille 2007). A typical aircraft deicing operation is shown in Figure 12. The rest of this section synthesizes the information on the validity and nature of the interaction between modern PDPs and ADAFs, describes the related standards and test proto- cols, discusses ways to prevent and mitigate such interaction, and identifies pertinent knowledge gaps. Validity of Interaction Between Modern Pavement Deicing Products and Aircraft Deicing/ Anti-Icing Fluids The first glycol-based, non-Newtonian ADAFs were intro- duced in the early 1960s. (“Non-Newtonian” describes a fluid that, when subjected to an external force, experiences increased viscosity until the external force is removed, upon 22 which it returns to its normal state. This characteristic allows most ADAFs to be applied easily and then resist being blown off the aircraft surface during flight.). Rates of in-flight in- cidents connected to thickener deposits from these deicers appeared to rise during the mid-1990s, shortly after the intro- duction of alkali-metal-salt-based PDPs (Ross 2006). A greater proportion of these reports appear to have come from Euro- pean operators than from North America or Asia (Hille 2007). The geographic imbalance of these reports is believed to be connected to the general method of aircraft deicer application favored at European airports. In Europe, aircraft are treated in a single application with a solution of Type II fluid in hot water. The two-step method favored elsewhere consists of an initial application of heated Type I (nonthickened) deicing fluid, followed by application of a heated Type IV solution (Hille 2007). User experience has shown that application of the pure or diluted Type I fluid removes thickener residue from previous deicer applications. The synergistic generation of residue when an ADAF on aircraft is splattered with modern PDPs (e.g., KAc and NaF) presents serious concerns about residue gel rehydration and refreezing in flight and has produced potentially dangerous rough residues on leading edge surfaces on aircraft. Air- craft and runway deicing fluids tend to mix in two different locations—the aircraft and the runway (Hille 2007). Aircraft deicer may run off the aircraft, mix with runway deicer, and then be splashed back onto another aircraft by landing gear spray or blown on by thrust reversers. Runway deicer may reach freshly applied aircraft deicer by the same means or from overspray during runway application. After mixing, the now less viscous aircraft deicer may remain in place or it may migrate to aerodynamically quiet areas through control surface gaps or vent holes, where thickeners can precipitate unnoticed. Nature of the Interaction Between Modern Pavement Deicing Products and Aircraft Deicing/Anti-Icing Fluids Thickeners used in aircraft deicers increase viscosity through charge–charge interaction; organic salts such as KAc and KF are known to disrupt this interaction (Ross 2006). In theory, contamination with KAc or KF should cause a measurable reduction in the viscosity of the aircraft deicer. Preliminary evidence shared with the SAE G-12 committee by Kilfrost, Ltd. appears to support this theory. Samples of Type II and IV aircraft deicer fluids contaminated with small amounts of KF- or KAc-based PDPs experienced immediate reductions in vis- cosity, followed quickly by precipitation of thickener addi- tives. These laboratory data appear to corroborate anecdotal reports of increased rates of thickener residues in environ- ments where alkali-metal-salt-based PDPs have been used. Nonetheless, this issue is being addressed by Kilfrost. Kilfrost also gathered data on the effect of runway deicers on dried thickener residue. The thickener was observed to rehydrate only slightly in a 5% KF solution when compared FIGURE 12 Typical aircraft deicing operation [adapted from Ambrose (2007)].

23 with the control. Immediately following this treatment, the same residue sample exceeded the control in weight gain when rehydrated with dematerialized water (Ross 2006). This suggests that not only do alkali-metal-salt-based PDPs accel- erate the precipitation and buildup of thickener residues, but under the right conditions they may also encourage greater moisture uptake by the thickeners. Standards and Test Protocols There are two independent standards used to designate ADAFs for military and commercial use. The military specifi- cation—MIL-A-8243D—is entitled Anti-icing and Deicing- Defrosting Fluids and approves only propylene glycol-based fluids as Type I and ethylene glycol-based fluids as Type II to be used by the U.S. Air Force. The commercial specifica- tions—SAE AMS 1424 (Type I) and AMS 1428 (Types II, III, and IV)—classify ADAFs based on their viscosity and hold- over properties, other than their chemical composition (Pro- Act Fact Sheet . . . 1998). AMS 1428F is the accepted standard for SAE Types II, III, and IV thickened, non-Newtonian aircraft deicers, and there is no provision for testing compatibility with PDPs con- tained in the July 2007 revision of this standard. Likewise, AMS 1435, the SAE standard for liquid runway deicers, con- tains no provision for testing compatibility with ADAFs. Prevention and Mitigation Although interaction between runway and aircraft deicers is inevitable, there are opportunities to control the effects of the interaction. When applying runway deicer, extra care can be taken to avoid overspray around parked airplanes. Thickener residue can be reduced to a minimum through frequent in- spection and cleaning of areas prone to buildup, such as spar areas and leading edge cavities. Dried residue can be re- hydrated with warm water spray and then flushed or wiped away. Nonetheless, challenges remain for such operational practices in commercial aviation, in light of the financial and environmental constraints. In addition, spray from PDP pools is unpredictable during aircraft take-off and landing. Inter- action with Type IV ADAFs has been seen to rapidly pro- mote rough, persistent residue on wing leading edges with unfavorable aerodynamic properties. Knowledge Gaps The contamination effects of ADAFs by runway deicing flu- ids have been well-observed, but not yet thoroughly quanti- fied. Acquisition of hard data will assist in the generation of inspection schedules (Hille 2007) and may spur development of improved thickener formulae for ADAFs. To this end, the SAE G-12 Fluid Residues Working Group is leading re- search efforts in this field. Currently available thickeners were designed to enhance the holdover properties of ADAFs and did not take the potential interaction with PDPs into account. From a residue mitigation standpoint, air carriers and airports would be well-served by a controlled comparison of the single- and double-step processes currently favored for application of ADAFs to identify the most effective method for controlling thickener residue buildup. Further research is needed to better understand the inter- actions between ADAFs and PDPs, as new ADAFs and PDPs are continually introduced to the market. For instance, envi- ronmentally benign alternatives to glycol-based ADAFs such as formulations based on glucose-lactate have been tested with promising results (“FAA-Approved . . .” 2007). In addi- tion to the freezing point depressant and additives used, the interactions between ADAFs and PDPs may be affected by the aircraft type, maintenance and inspection practices, and weather (Ross 2006). CONCLUDING REMARKS The U.S. aviation industry as a whole has enjoyed greater pavement frictional characteristics (safety) and longer oper- ating hours for aircraft (nonclosure of runways) because of the effectiveness of modern PDPs. In spite of their environmental advantages over older for- mulae such as urea and glycols, alkali-metal-salt-based PDPs present new challenges to the aircraft operating and manu- facturing industries. Table 11 summarizes the effects of mod- ern PDPs on aircraft components, including the key findings and knowledge gaps. It should be noted that the effects of modern PDPs on air- craft components lead to substantial financial consequences such as increased maintenance, inspection, and replacement costs and flight delay costs. Continental Airlines forecasted out-of-service and flight delay losses owing to catalytic oxidation of C/C aircraft brakes starting at $200,000 and $500,000 annually (Duncan 2006). Advances in anti-oxidant technology and C/C composite substrates are helping to con- trol this figure. Corrosion of Cd-plating and aluminum parts by runway deicers requires modifications and repairs to com- ponents on, in, and around landing gear and wheel wells, with an the annual cost estimated at approximately $1.3 mil- lion per national carrier for the foreseeable future (Duncan 2006). Increasing costs like these are making alternatives to Cd-plated steel such as anodized aluminum and stainless steel more attractive to manufacturers (Duncan 2006).

24 TABLE 11 SUMMARY OF EFFECTS OF MODERN PDPS ON AIRCRAFT COMPONENTS PDP Impact Information Sources What Is Known What Is Unknown 2. Existing research in the laboratory has demonstrated the catalytic effects of potassium, sodium, and calcium on carbon oxidation. 2. More research is needed to better understand relationships between brake design, AO treatment, and PDP contamination as factors in catalytic oxidation. 2. More research is needed to better understand the interactions among the aircraft component design, the CICs used, and the contamination of PDPs in the processes of metallic corrosion. 3. There is still a lack of academic research data from controlled field investigation regarding the aircraft metallic corrosion by PDPs. 2. Further research is needed to better understand the interactions between ADAFs and PDPs, as new ADAFs and PDPs are continually introduced to the market. 1. A growing body of field evidence from airline operators suggests that the use of KAc and KF on airfield pavements leads to catalytic oxidation of C/C composite brake components. 1. There is still a need to establish a comprehensive PDP catalytic oxidation test protocol. Catalytic oxidation of carbon–carbon composite brakes 1. Academic-peer-reviewed literature 2. Industry-peer-reviewed publications and reports 3. Survey of stakeholder groups 1. There is still a need to establish a comprehensive metallic corrosion test protocol for PDPs. 1. Industry-peer-reviewed publications and reports 2. Survey of stakeholder groups Corrosion of aircraft alloys (with a focus on cadmium plating) 1. Until recently, the principal evidence connecting alkali-metal-salt- based PDPs with Cd-plating corrosion has been a trend of increased reports of the latter occurring simultaneously with the introduction of the former. 2. Very little research has been conducted to investigate the mechanism of Cd corrosion or Cd- steel corrosion in the presence of alkali-metal-salts (e.g., KF and KAc), partly owing to the high toxicity associated with Cd and its compounds. Interaction with aircraft deicing and anti-icing products 1. Industry-peer-reviewed publications and reports 2. Survey of stakeholder groups 1. The contamination effects of ADAFs by runway deicing fluids have been well-observed, but not yet thoroughly quantified. 1. Recent laboratory data appear to corroborate anecdotal reports of increased rates of thickener residues in environments where alkali-metal- salt- based PDPs have been used.

Next: Chapter Four - Effects of Pavement Deicing Products on Airfield Infrastructure »
Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s Airport Cooperative Research Program (ACRP) Synthesis 6: Impact of Airport Pavement Deicing Products on Aircraft and Airfield Infrastructure explores how airports chemically treat their airport pavements to mitigate snow and ice, and the chemicals used. The report also examines the effects of pavement deicing products on aircraft and airfield infrastructure, and highlights knowledge gaps in the subject.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!