Martempering to Improve Wear Properties of Aircraft Brake Steel Rotors by Dr

Volume 23, Number 2 - April 2007 through June 2007

Martempering to Improve Wear Properties of Aircraft Brake Steel Rotors By Dr. Sudershan Jetley

Peer-Refereed Article

Keyword Search

Aviation Technology Materials & Processes Materials Testing Metals Quality Control Research

The Official Electronic Publication of the National Association of Industrial Technology • www.nait.org © 2007 Journal of Industrial Technology • Volume 23, Number 2 • April 2007 through June 2007 • www.nait.org Martempering to Improve Wear Properties of Aircraft Brake Steel Rotors By Dr. Sudershan Jetley

Abstract This article presents a study to improve

Dr. Sudershan Jetley is an associate professor the wear properties of aviation brake ro- at College of Technology in Bowling Green State tors. Currently these rotors are being aus- Similarly Torque (τ), although not speci- University. He received his Ph.D. from University fied by the regulating agencies can be of Birmingham. He has taught and supervised tempered using a hot salt bath. This has graduate students in the areas of statics, materi- caused safety concerns in the plant, so derived as: als, automation, quality, research methods and manufacturing processes. His research interests an alternative martempering process was include Rapid Prototyping, Machining, Neural investigated using oil and water based Network applications and Machine vision. quenchants at lower temperature. The test samples were evaluated for hardness, distortion and wear under accelerated Where: simulated tests. The results show that al- K.E. = Kinetic Energy per wheel in though both hardness and wear resistance ft.lbs was lower compared to the austempering, W = Weight in lbs it met the design intent. Also the wear V = Speed in Knots rate of martempered samples was more R = Tire Radius in inches consistent which may provide advantages N = Number of wheels for maintenance purposes. τ = Torque per wheel in inch.lbs Introduction From these relationships, it can be The wheel has evolved from a simple seen that for a given aircraft, speed and wooden disk of antiquity to the sophis- weight are the governing factors in brake ticated composite structure of today design. This is again seen in graphical used in land and air vehicles. This form in figure1. evolution has been the result of meeting the challenge of ever increasing higher Due to the V2 term in the energy equa- speeds and loads. A vital component of tion, energy effects take on the overrid- the composite wheel is its brakes. The ing importance in brake functioning. primary requirement of modern brake Consider the scenario of a fully loaded systems is to be able to apply a braking aircraft aborting a takeoff just before the torque while resisting failure due to high actual takeoff. For a jet fighter, the speed friction temperatures developed. This at that moment may be close to 200 is reflected in national and international miles per hour. There is obviously little regulations for certification. reverse thrust to be had from the engines and for a jet fighter there is little aerody- For aircraft, Federal Aviation Regula- namic drag due to its control surfaces at tion, Code of Federal Regulation Part 23 that moment. The runway is also quickly and the United Kingdom Civil Authority running out, so the aircraft needs to be Specification No. 17 specify the required stopped quickly. Therefore, all the stop- energy capacity rating for Brakes. The ping action falls on the wheel brakes. A specifications provide the following commercial jetliner can require in the formula for deriving the Kinetic Energy order of 5,000 braking horsepower per (K.E.) absorption requirement (United brake (Garret, D.F., 1991). Kingdom Civil Aviation Authority, 2006):

Journal of Industrial Technology • Volume 23, Number 2 • April 2007 through June 2007 • www.nait.org

Figure1. Effects of Speed and Weight of aircraft on Landing Energy Per wheel The two essential properties needed are (Johnson, 2006) heat dissipation by

1. Carrier assembly 2. Piston cup (outer) 3. Piston cup (inner) 4. Piston (outer) 5. Piston (inner) 6. Piston end (outer) 7. Piston end (inner) 8. Pressure plate For a jet fighter at peak speed, the sion of the fluid and drum leading to loss 9. Stator drive sleeve temperatures generated in the brakes due of brake action. After the World War II, 10. Auxiliary stator and lining to friction can exceed 2000o F (Hydro aircraft started to use disc brakes which assembly Aire Inc., 2005). The situation can be are used in modern aircraft. There are 11. Rotor segment further exacerbated by cross winds that two types, namely single and multiple 12. Rotor link would place additional and uneven loads disc brakes. 13. Stator plate on the different wheels due to steering 14. Backing plate compensation needed. At different alti- Pre World War II, the landing gear of 15. Torque pin tudes, K.E. generated also increases due aircraft was not retracted and stowed. 16. Adjuster pin to increased takeoff or landing speeds However, with increased speed require- 17. Adjuster clamp required to compensate for increased ments it became necessary to retract and 18. Adjuster screw density altitudes (Johnson, 2006). The stow the landing gear to reduce drag. To 19. Adjuster washer heat generated can result in rapid wear meet these requirements, wheels became 20. Adjuster return spring of the brake surfaces leading to brake smaller to reduce stow space, leading to 21. Adjuster sleeve failure, or at the least, reduction in brake the development of multiple disc brakes 22. Adjuster nut life. (Aircraft Technology Engineering and 23. Clamp hold down assembly Maintenance, 2000). This design has be- 24. Shim Early airplanes had no brakes and relied come standard and all large planes today 25. Bleeder screw upon the friction generated from their use the multiple disc brakes. An example 26. Drive sleeve bolt tails skidding on the grassy ground. diagram of these brake assemblies is 27. Dust cover (inner) When hard runways came into use, shown in figure 2 (Integrated publishing, 28. Dust cover (outer) aircraft brakes were generally adopted 2006). from automobile brakes. These early Figure 2. Segmented rotor brake–cut- brakes consisted of drum and shoes The essential part of the brake assem- away view. systems similar to the rear brakes found blies is a rotor rotating in contact against in today’s automobiles (Garret, D.F., a stationary stator during braking. The 1991). The major problem with these industry primarily uses sintered, copper types of designs is the thermal expan- based alloys as stators and steel rotators.

Journal of Industrial Technology • Volume 23, Number 2 • April 2007 through June 2007 • www.nait.org the stator and reduced wear rate of both ing safer quenchants. Two quenchants, This is shown in figure 3. The Hardness the stator and the rotor. Although the one oil based and other water based machine can be used with several differ- basic design has remained the same over were identified. This article reports this ent scales including Rockwell A-G, H, the years, there have been many devel- study. (Note: the identity and some of and K. In this study Rockwell scales B opments in materials and processing of the details of the process are not being and C were used. these materials (Aircraft Technology provided due to confidentiality require- Engineering and Maintenance, 2000.) ments)

A significant development has been the Martempering is a commonly used use of carbon to replace steel since the heat treatment technique. The purpose 1980s. Indeed in some larger aircraft a of this process is not only to obtain the fully loaded landing and takeoff may appropriate microstructures and me- not be possible without the utilization of chanical properties but also to minimize carbon (Graphite) brakes. distortions. In martempering, the part or specimen is heated well into the aus- However, carbon brakes needs to be tenitic region. Once the entire part has thicker than steel brakes due to various reached the specified austenitic tempera- reasons such as to provide sufficient ture it undergoes an interrupted quench volume to act as a heat sink and provide at a specified temperature below the Ms structural strength. Therefore, they re- temperature of the Time- Temperature quire more space due to carbon’s lower – Transformation (TTT) diagram. The density, and are also about 1/3 more part is then slowly cooled through the costly than steel. Therefore, steel rotors, martensitic transformation. The amount also known in the industry as pucks, of time the part is kept in the martensitic remain competitive and are used in most transformation determines the properties cases (Aircraft Technology Engineering of the part. Once the temperature of the and Maintenance, 2000). part falls below the Mf line, bainite and very fine pearlite is formed. The last step The Project in the martempering process is to temper At an aircraft brake manufacturing the part at the specified temperature for Figure 3. Rockwell Hardness Tester facility, the steel rotors (pucks) are heat the specified amount of time. This re- treated using an austempering process. lieves stresses produced due to quench- Austempering is the heat treatment pro- ing and converts any retained austenite 2. Small Electric Furnace cess of heating steel into the austenitic to pearlite, thus reducing the possibility A small electric oven was used to temper phase, and then quenching to a tem- of spontaneous transformations that can the pucks after they had been quenched. perature above the martensitic transfor- lead to cracking (e-funda engineering The oven is made by Hevi Duty Electric mation temperature, which causes the fundamentals, 2006. ) Co. model number MU-56-S. The oven austenite to isothermally transform to has a voltage range of 110-200, a maxi- bainite. In the process, the quench cycle In summary, the purpose of this study is mum wattage is 3400, and a temperature is interrupted at about 700o F and held to assess the feasibility of martemper- range of 0o to1200o F. for a period at that temperature. ing heat treatment using two different quenchants. This is to replace the current 3. Large Electric Furnace The company uses salt baths to hold Austempering heat treatment being used An electric oven was used to austenize the pucks at the constant temperature. on Aircraft Brake Steel Rotors. The the pucks before they were placed into The austempered steel met the design results of martempering are assessed the quenchant. Figure 4 is a photo of the expectations of the company, but the using hardness, distortion and friction furnace. The large furnace is made by process raised concerns regarding safety testing. The main impetus for this study Therolyne, and model number F-A1730. of workers due to the salt baths used comes from the need to avoid the use of The furnace has a temperature range of for quenching. Salt bath heat treating salt baths. 0o to 2500o F. is very hazardous, especially during cleaning and in maintaining the baths at Methodology elevated temperatures (e-funda engineering fundamentals, 2006). Therefore, Apparatus a similar but different heat treatment 1. Hardness Testing Machine process, martempering, was investigated Wilson Rockwell Hardness Tester to provide similar properties but us- machine was used to measure hardness.

Journal of Industrial Technology • Volume 23, Number 2 • April 2007 through June 2007 • www.nait.org

Materials

1. Steel Parts As mentioned above, the part under investigation was the steel puck which forms the rotor part of a brake assembly. It can be seen in figures 5 and 6. The part is approximately 2.5 inches in diameter and 0.33 inch thick. It was received in annealed and machined condition. It was made with the standard 17 – 22 – A Timken steel. The composition of the steel, in percentages, is given in Table 1 (see page 6), as provided by the manufacturer. Figure 4. Large Electric Furnace 2. Quenchants The aqueous quenchant used was the 4. Height Gage Aqua Quench 3699 from Houghton North American Precision Inc. height International Inc. (http://www.hough- gage was used to measure the thickness- tonintl.com/index.html). It was used at es at eight different points on the puck. a concentration of 25% water solution, Three measurements at each position based on manufacturer’s recommenda- were taken to obtain the thickness. Fig- tion. It is a polymer based liquid and ure 5 shows a close-up view of the gage. has some of the properties as shown in table 2 (see page 6) as provided by the Figure 6. Beaker/Hanger set up and a manufacturer. The oil based quenchant, close-up view also supplied by Houghton International Inc., was mart-temp oil 2525. It is based on paraffinic mineral oils with additives. 6. Quenchant Pot This was used at 100% concentration. An electric pot with a temperature Some of its properties are shown in setting control was used to quench the table 3 (see page 6) as provided by the pucks. The electric pot is a domestic manufacturer. kitchen pot and has a half gallon hold- ing capacity. The temperature range on Procedures the pot is 0o to 400o F. The temperature 1. Puck Identification control was used to set the quenchant In order to identify the different pucks, temperature and a ‘Suregrip” thermo- these were stamped with an identifying couple temperature probe was used to letter (A to T). In order to be consistent Figure 5. Height Gage / Thickness actually monitor the quenchant tempera- in the various measurements, these were Measuring ture. This probe has a temperature range labeled so that the measurement was of -40oF to 1994o F. Figure 7 is a photo done at approximately the same place of the electric pot. and orientation. The labels consisted 5. Cooling Rack of stamped letters N, S, E, W, at two A beaker stand was used as the hold- places, in the four directions so their ing device for the pucks after treatment. orientation could be quickly found as The pucks were hung using wire coat shown figure 8. hangers bent into a shape to hold them flat while entering the hot quenchant. This presented the maximum surface area during initial contact with the fluid, and so should cause maximum distortion. The coat hanger surface touch- ing the pucks was minimized to avoid uneven cooling. Figure 6 shows this set-up. Figure 7. Electric Pot

Journal of Industrial Technology • Volume 23, Number 2 • April 2007 through June 2007 • www.nait.org

Table 1. Chemical composition of “17-22-A” S Timken Steel (AMS spec; 6302 B) C Mn P Si S Cr Mo V .28-.33 .45-.65 .040 .55-.75 .04 1.0-1.5 .40-.60 .2-.3

(http://www.timken.com/timken_ols/mail/contact.asp)

Table 2. Properties of Aqua-quench 3699 Appearance Hazy Light Amber Fluid Viscosity at 100o F 687 ± 25 cSt Specific Gravity at 60o F 1.025 Figure 8: Labeled positions Density at 60o F 8.5 lb/gallon on a steel puck pH 9.3 ± 0.5 Corrosion Protection Excellent 2. Dimensional measurements Water solubility Normal In order to assess any distortion that may occur during heat treatment, ideally profiles of the parts should be taken by Table 3. Properties of mart temp oil 2525 such a devise as a coordinate measuring o machine. However due to lack of ap- Viscosity @ 100 F 567 c St. propriate apparatus, thickness (or height) @ 210o F 33 c St. and roundness of the parts were measured. Height was measured by placing Flash point 550o F the disc on a surface plate and using the Fire Point 620o F digital height gage at the four different labeled points, taking two readings at Specific Gravity @ 60o F 0.892 each point, resulting in eight readings per part. treatment had been conducted. Since, distortion if any. In actual production, the pucks received were in an annealed the pucks should be inserted in a vertical The heights were recorded before and state, the hardness was measured using orientation to minimize the distortion, after the pucks were heat treated to see the HRb scale. which then would be less than observed if there was any distortion in the axial in this study. As stated earlier under the direction after treatment. The pucks The second and third time the hardness quenching pot description section, the were thoroughly cleaned using sandpa- was taken after quenching and temper- rise in quenchant temperature was also per to remove any scale for the after heat ing respectively. This was done with a monitored during quenching using the treatment measurements. The set-up is HRc scale. In both cases, pucks were temperature probe. shown in figure 5. cleaned thoroughly to assure accurate hardness data. In all cases, the average After quenching, the parts were tem- To assess any distortion in the radial di- of the two harnesses was used and hard- pered at 1200o F for one hour. This was rection, each puck was traced on a sheet ness measurements were taken left of done in the oven and then air cooling on of graph paper, and then retraced after the stamped letter N and at right of the the bench. it had been heat treated and tempered. stamped letter S on the pucks. Transparencies of the after heat treat- 5. Friction testing ment traces were overlaid on the traces 4. Heat treatment After the parts were heat treated they of the before heat treatment traces to The austenizing temperature used in the were tested by accelerated sub scale evaluate any differences. Parts labeled oven was 1650o F in air. For quenching, friction testing. This test simulates the K, P, E and R were used for oil quench- the oil based quenchant was used at 400o braking loads experienced by an aircraft. ing and those labeled Q, M and T for F and the water based at 150o F. These Multiple cycles of braking were simulat- aqua quenching. temperatures were recommended by the ed. Briefly, a rotor (The puck) is pressed manufacturer. The pucks were quenched against a stator, which is a Cu-Fe-C-Si 3. Hardness measurements for 30 seconds. As shown in fig.6, the powder metal flat plate. The tests allow Hardness was measured three different orientation of the pucks when inserted the control of rotational speed, pressure times during experimentation. First time into the quenchant was horizontal as this etc. These are based on the Cold Taxi hardness was measured before any heat orientation should produce maximum Stop, Service Landing and Hot Taxi

Journal of Industrial Technology • Volume 23, Number 2 • April 2007 through June 2007 • www.nait.org

Snub conditions. A test cycle consists Table 4. Height (thickness) measurements of individual pucks of all these braking conditions. The Heights (mm) Heights (mm) test yields such data as coefficient of Oil quenchant Aqua quenchant friction, temperature generated, energy used, wear loss and the like. Before After Before After hardening hardening hardening hardening Results and Discussion E M Distortion in axial direction This distortion was assessed by measur- Maximum 8.455 8.9 Maximum 8.47 8.36 ing the heights of the pucks before and Minimum 8.4 8.36 Minimum 8.42 8.32 after hardening against a surface plate at the eight positions. The total range of all Difference 0.055 0.54 Difference 0.05 0.04 measurements was 8.9 to 8.1 mm. K Q The results are shown in tables 4, 5 and Maximum 8.46 8.39 Maximum 8.46 8.35 6. The difference column in table 5 is Minimum 8.42 8.36 Minimum 8.41 8.31 the difference in average values, with corresponding standard deviation of the Difference 0.04 0.03 Difference 0.05 0.04 difference. “n” denotes the number of P T readings. Maximum 8.44 8.43 Maximum 8.5 8.37 There seem to be a reduction in thick- Minimum 8.42 8.14 Minimum 8.415 8.31 ness after hardening in both quenchants; Difference 0.02 0.29 Difference 0.085 0.06 0.032 mm for the aqueous quenchant and 0.105 mm for the oil quenchant R as seen in table 5. This is confirmed in Maximum 8.455 8.44 table 6, which shows that statistically the probability of there being zero difference Minimum 8.415 8.405 in the before and after heat treatment for Difference 0.04 0.035 all pucks is quite small, with sample size of 8, i.e. n = 8. So the reduction in thickness is due to some assignable causes. Table 5. Height (thickness) measurements overall Height Oil (E, K, Height Aqua This may be due to scale that had (mm) P & R) (mm) (M, Q, T) built-up and subsequent cleaning before Difference Difference measuring. Oxidation was found on all parts. The standard deviations of the Before After Before After means before and after hardening are Hardening Hardening Hardening Hardening almost the same. If there had been any significant distortion, the difference n 32 32 24 24 values and their standard deviations in Average 8.434 8.042 -0.032 8.44 8.34 -0.105 table 5 would have increased, rather Std. Dev. 0.015 0.014 -0.001 0.022 0.018 -0.004 than decreased. Similarly the maximum values would also have increased, due to the method of measuring, which is not Table 6. Student’s t and probability values, assuming nulls hypothesis, for height mea- observed. So we can state that there was surements of all pucks before and after heat treatment (n=8) little to no distortion due to quenching in the axial direction and the differences observed in height values could be due Oil quenchant Aqua quenchant to causes such as oxidation and cleaning. Puck E K P R M Q T This can be investigated in some future work. t 6.86 10.1 2.94 2.88 13.6 13.1 7.6 P 0.001 0.001 0.011 0.012 0.0001 0.0001 0.0001 Furthermore the parts were oriented to present maximum surface area to the P = Probability of no difference in height values before and after heat treatment hy- liquid, to cause maximum distortion. In pothesis practice this would not be the case, since

Journal of Industrial Technology • Volume 23, Number 2 • April 2007 through June 2007 • www.nait.org pucks would be orientated to reduce the Figure 9. Example of a tracing to determine radial distortion initial contact surface.

Distortion in radial direction Although radial distortion is not of much importance in the functioning of the pucks, an attempt was made to measure this. As has been stated, the ideal method for measuring radial distortion would have been to use a profilometer or a CMM. In absence of such apparatus the above described tracing method was used. In all the eight parts used, no radial distortion was detected. An estimate of the accuracy of this method can be obtained as follows: The visual error in the procedure of comparing the two tracings can be said to be half of the pencil line width, so it is Table 7. Hardness values of oil quenched parts in the order of 0.5 mm. Since the diam- Puck# R Puck# P Puck# E Puck# K Average Std. Dev. eter of the puck was about 65 mm, the measurement error translates to 1.54 %. Initial (HRb) 8 9 8 8 8.25 0.433 (1mm/65mm * 100 = 1.54). An example After Quenching 24 23 27 26 25 1.58 of the tracing is shown in figure 9. (HRc) After Tempering 22 23 26 26 24.25 1.79 Hardness (HRc) The hardness results are shown in table 7 and 8. Each value is an average of two readings. The initial low hardness Table 8. Hardness values of Aqua quenched parts was due to the annealed condition of the pucks. The hardness value increased Puck# T Puck# M Puck# Q Average Std. Dev. in both cases due to martensitic trans- Initial (HRb) 8 8.5 9 8.5 0.5 formation. The temperature differences After Quenching 38 37 39 38 1 between the hot part out of the oven and (HRc) the pre heated quenchants were 1250o F After Tempering (HRc) 31 30 33 31.33 1.53 (1650 – 400 = 1250) and 1500o F (1650 – 150 = 1500) for the oil and water based quenchants respectively. Water Table 9. Results of the friction tests. generally has 2 to 3 times greater ther- Rotor Wear Stator Wear mal conductivity as well as heat capacity Part Number Rate Rate Brake Wear Rate Hardness than oils (Wikipedia, 2006) (Huksflux, ( Aqueous) Inches/cycle Inches/cycle Inches/cycle HRb 2006). T 0.0325 0.0359 0.0684 31 Hence, these two factors should have Q 0.0283 0.0391 0.0674 33 cooled the aqua quenched parts faster M 0.0282 0.0403 0.0685 30 than the oil quenched parts. This was Average 0.029667 0.038433 0.0681 31.33333 also seen by the fact that the oil based Standard o quenchant temperature rose by 20 F Deviation 0.002454 0.002274 0.000608 and the Aqua Quench 3699 water based (Oil) polymer quenchant by 35o F. The small decrease in hardness after tempering P 0.0274 0.0351 0.0598 23 was as expected, which is mainly due to R 0.03 0.0303 0.0603 22 stress relief. K 0.0213 0.0392 0.0605 26 Average 0.02685 0.0336 0.059775 23.66667 This decrease is greater with water Standard based quenchant, indicating that more Deviation 0.004466 0.004455 0.000361 quenching stresses were produced and

Journal of Industrial Technology • Volume 23, Number 2 • April 2007 through June 2007 • www.nait.org there may have been more retained Table 10. Student’s t and probability values, assuming nulls hypothesis, for wear rates austenite, which is again to be expected difference between stators and rotors used in aqueous and oil quenchants (n= 3) due to greater cooling rate. This can be confirmed further through metallo- t P graphic analysis which would be focus of further study. Aqueous 4.54 0.011 Oil based 2.37 .077 Friction tests The results show that the temperatures P = Probability of no difference in wear rates between rotors and stators for the two generated, energy used and the wear loss quenchants. (n= 3) were in the order of magnitude higher for service landing portion of the cycles than other portions. Due to the propriety Table 11. Student’s t and probability values, assuming null’s hypothesis, for wear nature of the results, a brief summary of rates difference between aqueous and oil quenchants for stators and rotors (n= 3) the results is provided in table 9. These results show that stators wear at a greater rate than rotors in general. t P Stators 1.24 0.28 The probability of this not being true is about 1 and 7.7 percent for the water and Rotors 1.17 0.31 oil based quenched rotors respectively as shown in table 10. Similarly the wear P = Probability of no difference in wear rates between rotors and stators for the two rate of rotors is less than that of stators quenchants. (n= 3) when aqueous quenched rotors are used compared to the oil quenched rotors. Based on the limited experimental 4. The wear rate in the friction test The probability of this not being the case tests, we can speculate that when there was higher for the stators that used is greater as indicated in table 11. is a greater difference in the hardness austempered rotors than the martem- between the stator and the rotor, the pered rotors Based on the results, the company abrasion wear mechanism predomi- 5. The variation in wear rate for the reported that, when compared with mar- nates. When the hardness difference is brake assembly in the friction test tempering, the current austempering pro- reduced, the adhesion mechanism starts was higher for the tests that used cess yields much greater hardness values to predominate. This can be tested in the austempered rotors than the martem- and a lower rotor wear rate. However the future by metallographic analysis. pered rotors. Therefore, the martem- stator wear rate is significantly greater, pering process will be more benefi- i.e. there is a larger difference in the Regarding the total brake wear, the cial from maintenance viewpoint. wear rates between stator and rotor with results show that there was less variation 6. Due to the lower quenchant tem- current austempering process compared in wear rate with the martempering pro- perature and less hazardous nature to martempering process of this study. cesses when compared to the austemper- of the Aqua Quench 3699 quenchant ing process as reported by the company. compared to salt baths, martemper- The element with the lowest melt- This is useful from brake maintenance ing process based on this quenchant ing point in the whole brake system is and stator and rotor utilization view is likely to be much safer and less copper with a melting point of 1981º points, since it may lead to more effec- costly than the current salt bath F, which is the main component of the tive maintenance schedules. based austempering process. stators’ composition. Although this tem- 7. Additional friction tests and further perature may be significantly reduced Conclusions and comments micro structural and wear analysis due to alloying. The temperatures gener- 1. The martempering did not produce work should be conducted to better ated during the friction test ranged from any significant distortion, axial or understand the wear behavior during about 100 to 500º F for time periods longitudinal. the friction tests. ranging from 1.5 to 25 seconds. These 2. The mart-temp oil 2525 yielded low- conditions do not appear to be conducive er hardness than the Aqua Quench References to cause diffusion wear. Therefore, it is 3699 quenchant. Aircraft Technology Engineering and most likely that the wear is governed by 3. The wear rate in the friction test Maintenance, (2000). Stopping at all adhesion and abrasive wear mechanisms was lower for the rotors that were costs, Aircraft Technology Engineer- causing the harder rotor to wear less austempered in ing and Maintenance, Feb/March than the softer stator. the current process than the martem- 2000, www.craneaerospace.com/So- pered rotors lutions/Landing/downloads/braking.

Journal of Industrial Technology • Volume 23, Number 2 • April 2007 through June 2007 • www.nait.org

pdf, Available: 7/15/06 hydroaire.com/downloads/Brake- Wikipedia (2006). Specific Heat Capac- e-funda engineering fundamentals, Savings.pdf , Available: 10/4/05 ity, http://en.wikipedia.org/wiki/ (2006 ). Tempering, http://www. Integrated publishing, (2006). http:// Specific_heat_capacity , Available: efunda.com/processes/heat_treat/ www.tpub.com/air/12-26.htm, Avail- 7/15/06 softening/tempering.cfm , Available: able: 7/15/06 7/15/06 Johnson, P., (2006)., Kinda Kozy MKIV Acknowledgements Garret, D.F., Aircraft Systems and com- RG SVX ( C – GPLE), http://www. The author thanks Messrs. Jason Grubb ponents, Jeppesen Sanderson, Inc., canardzone.com/members/Phil- and Brad Cymbola for most of the Englewood, CO , Dec. 1991. lipJohnson/General/brakes.htm experimental work done; the brake Hukseflux Thermal Sensors, (2006). Available: 7/15/06 manufacturers for providing the test http://www.hukseflux.com/ United Kingdom Civil Aviation Author- specimens and conducting wear tests; thermal%20conductivity/thermal. ity, Specifications and tests, Speci- Houghton International for providing the htm, Available: 7/15/06 fication No. 17, 1, ISBN 1 904862 quenchants; and the College of Technol- Hydro Aire Inc, (2005). How Brake 78 0, Sept. 1986, http://www.caa. ogy, BGSU for providing the equipment control Can Prolong the Life of co.uk/docs/33/CASPEC17.PDF, and facilities to conduct the work. Carbon Brakes, pp 1-4, http://www. Available: 7/15/06