9th ANNUAL AEROSPACE HAZARDOUS MATERIALS MANAGEMENT CONFERENCE

DENVER, COLORADO SEPTEMBER 28-30,1994

TECHNICAL PAPER COVER SHEET

Paper Title: Alternative Cleaning Solvents and Lubricants at the Air Force PJKS Facility

Number of Pages: 12

Primarv Author: S.P. Pince Martin Marietta Astronautics Denver, CO

Presenter: Merritte W. Ireland Martin Marietta Astronautics

Co-Authors: B.L. Worthington Martin Marietta Astronautics

Merritte W. Ireland Martin Marietta Astronautics

W.E. McGuire Martin Marietta Astronautics

ALTEANATIVE CLEANING SOLVENTS AND LUBRICANTS AT THE AIR FORCE PJKS FACILITY*

S.P. Prince, B.L. Worthington, M.W. Ireland, and W.E. McGuire Martin Marietta Astronautics

Denver, Colorado

ABSTRACT

In accordance with required eliminations of chlorofluorocarbon 1 13 (CFC-113) and 1,1,1- trichloroethane (TCA), Martin Marietta has evaluated non-ozone depleting chemicals for implementation at the Air Force PJKS facility. These replacements are intended for all operations in which CFC-113 and TCA were previously used, and are targeted at mechanical cleaning operations, electrical cleaning operations, and the replacement of an anti-seize lubricant. Selection criteria for mechanical cleaners include the cleaning power of the solvent, the compatibility with materials being cleaned, the cleaning ability after loading with dissolved contaminants, the toxicity and relative safety of using the cleaner, and the waste disposal requirements. In addition, because the PJKS facility conducts tests on propulsion systems for missiles and satellites, the cleaner must be compatible with liquid rocket propellants (Aerozine-50 fuel and nitrogen tetroxide oxidizer). Electrical processes involved the cleaning of rotary contacts used in amplifier modules. Replacement of an anti-seize compound used to lubricate bolt threads was also required. This material is used as a shop aid in protecting threaded fasteners from exposure to high stress, high temperature, or corrosive atmospheres. After conducting screening, demonstration, and validation tests, replacement candidates for these materials were selected. These candidates met or exceeded all technical requirements identified for their intended applications.

INTRODUCTION

The Montreal Protocol and the 1990 Clean Air Act Amendments mandate the elimination of CFC-113, other chlorinated fluorocarbons, and TCA after December 31 , 1995. In response, the Air Force has formulated policy that prohibits purchase of these solvents for Air Force use after April 1, 1994. The Air Force also prohibits contractors from using ozone depleting chemicals (ODCs) on Government Owned Contractor Operated (GOCO) properties. As part of this initiative, the Air Force has funded Martin Marietta to investigate alternatives to ODCs used at

. the Engineering Propulsion Laboratory (EPL). This laboratory is operated by Martin Marietta, and is located on Air Force property known as Air Force Plant PJKS.

-_--_______________-____________I

Approved for public release; distribution is unlimited

'This work was performed under Contract No. FO4701-85-C-0019 Modification TO0827 NUSTO 93-01 8 with the Martin Marietta Astronautics Corporation, Denver Colorado.

MECHANICAL CLEANING ALTERNATIVES

Requiromonta

and - Program

- Current Process h - Hazardous Air

Pollution List

Objectives

Goals

Mechanical cleaning operations use TCA or CFC-113 to clean and degrease hardware prior to painting and precision cleaning. Gross cleaning is performed to achieve visibly clean surfaces and to remove contaminants such as oils, greases, and hydrocarbon fluids. Cleanliness is verified by visual inspection and may be aided by wipe tests, water break tests, ultraviolet inspection, and special lights and mirrors. Precision cleaning is used to attain a high degree of cleanliness, typically required for critical applications. Precision cleaning is performed in a clean room environment, and the degree of cleanliness is not visible to the human eye. Cleanliness is,verified by particle count, non volatile residue (NVR), and total filterable solids. A variety of parts are cleaned at PJKS including disassembled valves, tubing, flanges, and fittings. Metal hardware includes stainless steel, aluminum. carbon steel, brass, bronze and copper. Nonmetals include Teflon@, Kel-F@, nylon, rubber, and other plastics. Typical contaminants include propellant compatible lubricants, greases, hydraulic fluids, silicone and hydrocarbon oils, and propellant residues.

Product Survey

60 Candidatea

Initially, over 60 cleaners were identified as possible replacements for TCA and CFC-113 in mechanical degreasing operations. These initial candidates were identified from current literature sources and upon communication with various chemical manufacturers. The list included representatives from different chemical families: aqueous cleaners, semi-aqueous cleaners, organic solvents, alkaline cleaners, and terpene cleaners. Figure 1 shows the flow chart used to provide technical screening of these 60 candidates, and to provide the final

Plan

Dem / Veal

Modify Operational 3 Selections

selections for .implementation at the PJKS facility.

L

Pre-Screening Screening

Tests

1 0 Candidates 4 Candidates

Scenario

(If Required) 1

FiGURE 1 SELECTION PROCESS, MECHANICAL CLEANING COMPOUNDS

From the list of 60 initial candidates, approximately 28 candidates were eliminated upon review of manufacturer supplied data. Many of these cleaners had low flash points or would be classified as hazardous materials. In the case of chemically similar cleaners (such as the citrus based d-limonene cleaners) only one or two of the cleaners were selected for further evaluation. Pre-screening tests were performed on the remaining 32 cleaners to determine cleaning capability and materials compatibility. The ten best candidates evaluated in the pre-screening tests were taken through a more rigorous test program to determine cleaning power, soils loading, materials compatibility, and compatibility with liquid rocket propellants. From this test, four candidates were selected. Demonstration and validation tests were performed on these four remaining candidates to demonstrate cleaning performance using actual facility hardware and to validate applicable process specifications. The following paragraphs discuss the test methods and results of the screening tests and the demonstrationhalidation tests.

SCRFFNING TF.STS

The purpose of the screening tests was to evaluate the cleaning power, soils loading, materials compatibility, and propellant compatibility of the selected candidate cleaners. The ten cleaners considered for this evaluation were: Bioact 280m (Petroferm Inc.), Biogenic Regentm (Rochester Midland), Citrexm (Inland Technology), Daraclean 282m (W.R. Grace and Co.), EP 921m (Inland Technology), Formula 815 GDm (Brulin Corporation), JPX DegreaseP (Jayne Products), Partsprepm (Sentry Chemical Co.), TCA, and CFC-113. TCA and CFC-113 were included as baseline cleaners because of their present use at the PJKS facility.

In the cleaning power evaluation, test coupons 5xIOx0.3 cm (2~4~0 .125 inches) were fabricated from 304 stainless steel sheet stock. The coupons were initially prepared for testing by grit blasting using 220 grit, followed by cleaning using TCA and hot soapy water. The coupons were rinsed in deionized water and oven dried. Drilube 822m, hydraulic oil (MiI-H- 6083), lithium grease Mobil EP2TM, and Dyke" machining dye were applied to the test coupons. These contaminants were aged on the surface by baking at 38OC (10OOF) for seven days prior to cleaning. Coupons were placed in a glass beaker, cleaning solution added, and the cleaner was mechanically stirred for fifteen minutes. At the end of selected cleaning periods, samples were removed and examined for remaining residue. Visual examination was augmented using 1 OX magnification, blacklight inspection, wiping with a cleanroom cloth, and by conducting a waterbreak test. The degree of contaminant remaining was estimated numerically, Results of the cleaning power test are presented in Figure 2.

contaminant remaining

5 15 minutes

FIGURE 2 CLEANING POWER TEST RESULTS

The eight candidate cleaners as well as TCA and CFC-113 were evaluated for compatibility with materials typically processed at PJKS. These included aluminum, 304 stainless steel, carbon steel, brass, PlexiglasTM, titanium, copper, galvanized iron, Kel-FTM, graphite/epoxy and G-10 fiberglass. Other hardware tested included cadmium plated nuts and bolts, bronze fittings, solder samples, as well as Vitonm, ethylene propylene rubber (EPR) and Buna-N rubber gaskets and O-rings. Materials were placed into glass bottles containing 250 milliliters of the cleaners. Three exposure conditions were attained: liquid phase exposure, vapor phase exposure, and exposure at the liquid-vapor interface. A visual inspection of each sample was made at the end of 24 hours, and then weekly after that for a total of six weeks exposure.

The soils loading test consisted of pre-loading the cleaners with 30 weight motor oil, and measuring the ability of the cleaner to clean metal coupons contaminated with Drilube 822 or hydraulic oil. The results of this test were used to determine the frequency of cleaner changeout after cleaning contaminated hardware.

In the propellant compatibility test, small amounts of cleaner were added to liquid rocket propellants (Aerozine-50 and nitrogen tetroxide) in a sealed glass pressure vessel. The reaction pressure and temperature were monitored using an electronic data logger. Rapid rises in temperature and/or pressure indicated a chemical incompatibility between the cleaner and the propellant. Figure 3 shows the test apparatus used for conducting propellant compatibility tests.

DATA LOGGE

0 1 00 PSlG I I

LI I P

I I

I I I

SEPTUM PORT

ADDITION MARK

STIRRING MOTOR

FIGURE 3 PROPELLANT COMPATIBILITY TEST APPARATUS

Four of the eight cleaners were eliminated from the test program based on the results of the screening test: Citrex experienced a violent reaction when mixed with nitrogen tetroxide. Daraclean 282 did not clean as efficiently as some of the other cleaners and corroded the brass hardware as well asfhe cadmium plated bolts. JPX Degreaser was incompatible with nitrogen tetroxide and non-metallic materials were degraded by this cleaner. Formula 815 GD showed some indications of metals incompatibility after the six week exposure period. Although the EP-921 reacted with nitrogen tetroxide, this cleaner was retained because of its strong cleaning power. Applications involving this cleaner would be limited to hardware unexposed to liquid propellants, or to hardware in which all traces of residual propellant have been removed prior to cleaning. In addition to the EP-921, demonstration and validation tests were conducted on the remaining three cleaners (Biogenic Regent, Bioact 280, and Partsprep).

LlEMONSTRATlON AND VALIDATION TESTS

The purpose of the demonstration and validation tests was to demonstrate cleaning performance on facility hardware and to validate applicable process specifications. Two inch pneumatic valves and BarksdaleTM Control valves were used to demonstrate cleaning of facility hardware. These valves are assembled from a variety of metallic and non-metallic components, and were disassembled prior to cleaning (Figure 4). Preliminary inspection of the valves indicated oil residue and corrosion on stainless steel castings, minor rust and oil on the spring mechanism, Drilube lubricant on cartridge end plates and cylinders, grease and corrosion on brass threads, and silver paint, orange paint, and residual tape adhesive on the cartridge body. The parts were placed in a basket and suspended in a pump agitated bath containing the cleaner at room temperature for 45-60 minutes. The materials were inspected, removed from the bath, and transferred to an ultrasonic bath for an additional 20-60 minutes until cleaning was considered complete. The parts were rinsed with deionized water, rinsed with hot soapy water, rinsed again with deionized water, and oven dried. Parts processed with Bioact 280 or EP-921 were visually clean after removal from the bath, and exceeded results expected from TCA cleaning. Parts processed using Biogenic Regent and Partsprep were still contaminated with Drilube after cleaning, and results were generally poorer than those expected from TCA cleaning.

To validate process specifications, parts were contaminated with vacuum oil, grease, Drilube, or KrytoxTM. These parts consisted of metal blocks with blind holes, stainless steel tubes with 90° bends and attached "B" nuts, and valve parts. The contaminated parts were aged at 38OC (1OOOF) for seven days, and subsequently cleaned with the four candidate cleaners in an ultrasonic cleaning bath. The cleaning process consisted of ultrasonic cleaning, hot soapy water rinsing, rinsing with deionized water, and oven drying. Some mechanical brushing was used as necessary to aid in dislodging the contaminants. After cleaning, the parts were rinsed using methylene chloride, and the solvent was evaporated to determine the non volatile residue (NVR). The goal of this process was to attain parts conforming to Level 100 A of the Martin Marietta Manufacturing Process Specification MP50405 "Contamination Control Specification". This specification requires an NVR equal to or less than one milligram residue per 100 milliliters solvent. Results from this test are included in Figure 5. Biogenic Regent did not meet this standard, with respect to cleaning Drilube and Krytox contaminants, and EP-921 was considered marginal.

The apparent difference in the cleaning ability of the Partsprep and EP-921 cleaners is attributed to cleaning actual facility hardware (demonstration tests) versus cleaning hardware in which contamination was artificially introduced and baked (validation tests). The Bioact 280 is the preferred substitute cleaner for implementation at PJKS, but EP-921 and Partsprep will be retained for specific cleaning applications. Biogenic Regent was eliminated because it had not passed either of the requisite demonstration or validation tests.

I

FIGURE 4 COMPONENT PARTS UPON DISASSEMBLY OF FACILITY VALVES

45 T 40

- E 35

30 . 25

10

5

0

FIGURE 5 NON VOLATILE RESIDUE ANALYSIS OF CONTAMINATED HARDWARE

ELECTRICAL CLEANING ALTERNATIVES

Electrical cleaning operations include cleaning of contacts in rotary amplifier switches. Electrical contacts are currently cleaned with Kontact Restorerm (Chemtronics Inc.), an aerosol cleaner that is formulated with trichlorofluoromethane (Freon@-1 1) and a hydrocarbon lubricant. This cleaner is used to remove contamination from the contacts as well as to lubricate the contacts and prevent mechanical wear.

Two tests were performed in the evaluation of ODC-free contact cleaners: a contact cleaning test, and a contact life cycle test. In the cleaning test, the ability of the replacement cleaners to restore electrical conductivity of contaminated contacts was demonstrated. The electrical resistance specification for new contacts as supplied by the manufacturer was 0.010 ohms. The average new contact resistance measured by Martin Marietta was 0.0049 ohms. Contacts removed from amplifiers used in the field for over ten years had resistance measurements between 0.204 ohms and 0.331 ohms, clearly in excess of this specification. The existing cleaner (Kontac Restorer) demonstrated adequate cleaning performance of these contacts after five cleaning cycles. One of the replacement candidates, HF Contact CleanerM (CRC Industries Inc.), was able to clean the contacts to a final resistance measurement of 0.0067 ohms after three cleaning cycles. In the life cycle test, the ability of various cleaners and lubricants to

reduce contact wear during normal switching operations was demonstrated. In this test, the contact position was cycled at 30 revolutions per minute using a DC motor until degradation in electrical conductivity was observed. A photograph of the test apparatus and the rotary switches used in this evaluation is shown in Figure 6. After 40,000 cycles, electrical resistance measurements of unlubricated contacts degraded from about 0.016 ohms to 0.039 ohms. When the contacts were lubricated, in most cases restoration of the electrical resistance to 0.010 ohm was achieved (Table 1). This result demonstrated the need to provide lubrication as well as cleaning during scheduled maintenance of electrical contacts.

Table 1 Contact Resistance Measurements, Life Cycle Test

REPLACEMENT OF ANTI-SEIZE COMPQUND

An anti-seize lubricant (Fel-Pro C5Am) is applied to threaded fasteners to aid in the disassembly under high torque, after exposure to corrosive environments, or in high temperature applications. The present material contains isobutane and propane as propellants (each less than ten percent by weight), TCA as a carrier solvent (70 percent by weight), and the balance consisting of copper and graphite flakes suspended in a petroleum grease base. The copper and graphite provide a coating of non-ferrous particles on threaded surfaces. When placed on steel or stainless steel threads, these particles lubricate the threads by deforming under applied torque loading. At the PJKS facility, the lubricant is applied to fittings, nuts, and bolts of sizes 0.6 cm. (1/4") and above, and can be potentially used on valve stems. The majority of the usage consists of application to 5 cm. (2") diameter threaded parts, which display a high inclination to seize after exposure to high temperature.

Anti-galling properties of replacement anti-seize compounds were demonstrated by a torque-vs-clamping force test. Both the existing ODC lubricant (Fel-Pro CSA) and the ODC- free replacement candidates were evaluated. A more recent formulation of Fel-Pro C5A containing isopropyl alcohol as the carrier solvent was similarly tested. In this test, the lubricant was applied to high strength stainless steel (ASTM A193) bolt threads and load was applied at 2 . 2 2 ~ 1 0 ~ N (5,000 pound) increments until a bolt tension load of 1 .33x105 N (30,000 Ib) was achieved. At each increment, the torque was measured. A photograph of the bolt force tester is shown in Figure 7.

FIGURE 6 CONTACT LIFE CYCLE TEST APPARATUS

I

FIGURE 7 BOLT FORCE TESTER ..

Results of this test, shown in Figure 8 , indicate that the existing lubricant and the four replacement candidates displayed linear relationships between the applied bolt torque and the tensional force. A constant coefficient of friction using these lubricants was attained, and no thread seizure was observed. All candidates performed well at room temperature, but only the Dow Molykote 1 OOOW allowed disassembly of fittings after thermal cycling between 25OC

_ . (77OF) and 649OC (120OOF).

4 5 0 1 P 400

fi 350

300 J 3 2 250 0

0 !!! 1 5 0 J n

(A

L. c

+ 200

2 100 50

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 v) 0 lo 0 u) 0

Y v- 01 01 0

0

-8- Fel-Pro C5A (Existing) - Fel-Pro C5A (Replacement)

-.- Anti-Seize Special

d L o c t i t e A-S 767

-.- Dow Molykote 1000

BOLT LOAD, Ibs

FIGURE 8, TORQUE VS CLAMPING FORCE TEST, ANTI-SEIZE LUBRICANTS .. ,

In all appli ations tudied (mechanical cleaning, electrical cleaning, and anti-seize comDounds), alternatives to TCA and CFC-113 have been identified that exhibit identical or superior technical performance than the existing ozone depleting chemicals. Many of the identified alternatives are safe to use by operating personnel and are also environmentally friendly. Several replacements have been identified for each application so that these chemicals may be qualified for use in Martin Marietta's applications. The use of multiple replacements allows for minimization of risk due to reliance on one chemical system or product supplier. Implementation of the alternative cleaners and lubricants is in progress at Air Force Plant PJKS. Because many of these unique processes are specific to applications at PJKS, candidates rejected or accepted for use by Martin Marietta would require an individual assessment by the user prior to implementation of these materials.