N251-006 TITLE: Diagnostics, Prognostics, and Health Management for Non-Steady State, Rapid Acceleration/Deceleration, High-Load Bearings

OUSD (R&E) CRITICAL TECHNOLOGY AREA(S): Advanced Materials;Integrated Sensing and Cyber;Sustainment

OBJECTIVE: Develop and demonstrate an empirically developed system for failure risk prediction, diagnostics, prognostics, and health management of rapid acceleration/deceleration, high-load bearings to increase operational availability (Ao) of carrier-based recovery systems.

DESCRIPTION: Arresting gear on aircraft carriers quickly decelerate aircraft over a short distance (hundreds of feet or meters) and time (2–3 s). Nimitz-class carriers utilize the MK-7 arresting engine, while the Ford-class utilizes the MK-15, Advanced Arresting Gear (AAG). In both instances, an aircraft tailhook engages the cross-deck pendant (CDP). The CDP attaches to the purchase cable (PC), which transfers energy to the arresting engines. Below deck, sheaves (i.e., pulleys) are used to change PC direction and wrap the cable around shock/energy absorbers. Due to the nature of the application, the sheave bearings on the MK-7 and the spherical roller thrust bearings on the MK-15 AAG experience high-loading, and rapid acceleration followed by rapid deceleration. This cycle repeats as aircraft are continuously arrested during flight operations. Arrestments are followed by retracts, which have a lower load and acceleration, where the cable is retracted back to its pre-arrestment position.

Bearing specifications differ between MK-7 and MK-15 (AAG); they also differ within MK-7 depending on location and use. The max RPM of the MK-7 bearings is in the range of 1000–1400 RPM; the max-RPM range of the MK-15 is between 600–900 RPM, with a few follower-bearings going up to 1,400 RPM. The bearings accelerate to and decelerate from this speed over a matter of 2–3 seconds. Load ratings vary depending on where the bearing is used. On the MK-7, the maximum load ratings range from approximately 335,000 lb–360,000 lb (151.95–163.29 MT) static (with one outlier rated for 560,000 lb [254.01 MT] static.) and 190,000 lb–268,800 lb (86.18–121.93 MT) dynamic. In operation, the max operating loads are 105,000 lb (47.63 MT) cable tension, 210,000 lb (95.25 MT) resultant load. However, there can be overload scenarios that load sheaves to an approximate theoretical load of 350,000 lb (158.76 MT). Thrust loads are minimal (a few hundred pounds) for the majority of the sheave bearings. One sheave bearing type is rated at 2,100 lb (.95 MT) of thrust. Examples of the MK-7 inner diameters, outer diameters, and widths include: 7.9 in (20.07 cm) ID x 12.6 in (32.004 cm) OD x 3.5 in (8.89 cm) W, 12.6 in (32.004 cm) ID x 19.7 in (50.034 cm) OD x 2.5 in (6.35 cm) W, 18.7 in (47.5 cm) ID x 24.3 in (61.72 cm) OD x 2.7 in (6.86 cm) W. The MK-15 uses 300 mm (Timken / SKF bearing model number 29360) and 360 mm (Timken / SKF bearing model number 29372) inner diameter spherical roller thrust bearings under high-dynamic loads, as well as a few follower bearings (both pillow-block and thrust bearings) that go up to 1,400 RPM; the long-term effects of thousands of arrestments on the risk of bearing failure has yet to be determined.

When selecting a roller bearing, a common practice is to estimate the bearing’s "L10" life, defined as the number of revolutions (sometimes listed as a time at a constant speed) before there is a 10 % chance of bearing failure. A bearing can have one of many different types of failure mechanisms, including a lubricant failure causing the bearings to seize; as well as a failure in fatigue, causing rollers to crack and potentially jam the bearing. These empirical equations are based on prior tests (mostly around the 1940s) of roller bearings being continually spun until failure. These tests predominantly used continuous speed tests, and these empirical equations assume a continuously spun bearing. When a roller bearing’s speed is varied, the only available approach to estimating the bearing’s L10 life is to extrapolate from a summation of continuous speed calculations; this is at best an educated guess, and one does not truly know the probability of a bearing failure for bearing applications with significant variability in speed. It is clear, however, that a rapid change in speed will impart more fatigue and alter the lubricant properties, and thus the true long-term risk of failure from hundreds of thousands of arrestments has yet to be truly determined.

In addition, maintenance practices, failure risk, and the life of steady state bearings are better defined, and diagnostics/prognostics technologies are more mature. The rapid acceleration and deceleration of bearings in arresting gear applications is atypical; there is significant variability when quantifying the risk of a bearing failure, and this unknown risk from this unique use case leads to potentially conservative maintenance practices. This includes high-frequency greasing of the sheaves on MK-7 (every 20 arrestments) and routine teardown for inspection. The inspections include taking apart the sheaves, wiping off grease, and visually inspecting the bearing. Repeated disassembly and reassembly of the sheaves increase the maintenance, and the frequent disassembly for inspection inherently increases the risk of damage. A reduction in maintenance requirements can reduce Operations and Support (O&S) costs by (a) decreasing hours spent on inspection, and (b) preventing excessive teardown from increasing the failure rate of the sheaves. Sheave inspections take anywhere from 6–14 hours of work (per sheave) depending on the type of sheave. A method of reducing the inspections and maintenance of the bearings that preserves safety and reliability would increase Operational Availability.

Mobil Mobilith SHC 460 grease is used on the MK-7 sheave bearings, and none of the bearings are sealed. Phenolic and steel spacers act as grease retainers, but grease still escapes from small gaps between the spacers and the housings. For the majority of sheaves, the grease ports are stationary and grease is fed through grooves in the spacers. One sheave type is greased from the inner diameter of the sheave shafts. On the MK-15 AAG, Mobil Mobilith SHC 629 and 634 bearing oil is used in the spherical roller thrust bearings; seals hold the liquid oil within the bearing cavity.

The Navy is seeking an innovative solution to setting up an apparatus to subject roller bearings under a high load (relative to the bearings’ dynamic load limit). The apparatus will cyclically ramp up the bearings from stationary to a high speed (relative to the bearings’ rated speed), and then immediately and rapidly decelerate the bearings to stationary; each cyclic event should last less than 5 seconds. This cyclic acceleration and deceleration would need to continue indefinitely until the bearings have failed. Undoubtedly, this process can take a long time, but it would be essential for such an apparatus to be scaled and/or replicated in such a way such that a trend of estimated failure rates versus the number of cycles can be determined with a reasonable statistical confidence.

From this experimental apparatus, the Navy is seeking an empirically derived solution to predict the risk of a bearing failure and to track the bearing health over time. This will involve the development of diagnostics/prognostics algorithms. Insight into appropriate greasing, inspection, and maintenance intervals is required to decrease maintenance hours, extend bearing life, and alert Sailors to required maintenance prior to bearing failure. It is expected that a fixed, scheduled greasing interval will remain. Research and data are required to determine if the current 20-arrestment interval is reasonable or too conservative. In regard to diagnostics and prognostics of the bearings, real-time health monitoring is preferred, but periodic, automated inspections are also acceptable, so long as they do not increase the maintenance burden on the Fleet and enable a move towards a condition-based maintenance (CBM) approach.

Approaches may include, but are not limited to, a combination of modeling and simulation (M&S), instrumentation, sensor fusion, prognostics and health management, and/or other methodologies for data collection and data analytics, based on the empirical data. From a software perspective, advances in artificial intelligence and machine learning, or other related innovations associated with prognostics and health management may be leveraged to achieve the goals as outlined. Prior research and literature surrounding reliability, availability, and maintainability, including associated failure distributions (e.g., normal, Weibull, etc.) and other probabilistic/statistical methods are also relevant. From a hardware perspective, existing sensors, such as accelerometers, temperature sensors, thermal imaging, torque sensors, nondestructive inspection equipment, and so forth, may be appropriate; however, proposers are in no way limited to these technologies or methodologies and may offer alternative means to monitor health. Designs must be minimally intrusive, and capacity/space for additional, bulky sensing equipment is limited.

PHASE I: Demonstrate feasibility of high-load, non-steady state bearing predictions and health monitoring. Design and develop a solution that utilizes hardware to collect data at representative bearings and utilizes software to accept data for evaluation via M&S, data analytics, AI/ML algorithms, or other methods. Awardee may develop a physical, subscale bearing test bed during Phase I; however, it is not a requirement if the awardee can achieve similar results by generating realistic datasets using computer resources. The Phase I effort will include prototype plans to be developed under Phase II.

PHASE II: Design and build a full-scale prototype based on Phase I work. Demonstrate the technology in a lab environment using a bearing test bed that models the loading and acceleration profiles of the Navy’s bearings. Consideration should also be given to cyclic operations (i.e., high-sortie rate testing), as this is expected to lead to the highest temperatures, wear, and potentially bearing failure.

Validate and verify that the approach meets needs and requirements of the application by showing that diagnostics/prognostics algorithm(s) can identify (a) a proper greasing interval, (b) current bearing health state, (c) when to visually inspect and/or perform maintenance, and (d) remaining useful life or mean time between failures (MTBF) as applicable.

Assuming iterative design is used, and a more capable solution is developed gradually throughout this phase, consideration will be given to packaging to meet military specifications, data storage/processing, the health monitoring user interface, and integration with existing equipment and infrastructure.

PHASE III DUAL USE APPLICATIONS: Use any algorithms, sensor systems, bearing monitoring systems, and life-cycle prediction tools developed during Phase II to both accurately predict the expected number of arrestments a set of bearings can handle prior to an expectation of failure, as well as predict when anomalies in the bearing performance (e.g., vibrations, increase in torque) is indicative of a developing problem. Transition the monitoring systems to the ships to alert the crew when anomalies are detected, or maintenance is needed.

There are countless examples of commercial applications, which use bearings that accelerate and decelerate rapidly, and that can benefit from this technology. Some examples likely include bearings with large braking requirements, such as landing gear on aircrafts, and brakes on trains.

REFERENCES:

1. "MIL-DTL-901E. Detail specification: shock tests, H. I. (high-impact) shipboard machinery, equipment, and systems, requirements for." Department of Defense, 2017, June 20. http://everyspec.com/MIL-SPECS/MIL-SPECS-MIL-DTL/MIL-DTL-901E_55988/

2. "MIL-STD-167/1A. Department of Defense test method standard: mechanical vibrations of shipboard equipment (Type I-environmental and Type II-internally excited)." Department of Defense, 2005, November 02. http://everyspec.com/MIL-STD/MIL-STD-0100-0299/MIL-STD-167-1A_22418/

3. "MIL-STD-461G. Department of Defense interface standard: requirements for the control of electromagnetic interference characteristics of subsystems and equipment." Department of Defense, 2015, December 11. http://everyspec.com/MILSTD/MIL-STD-0300-0499/MIL-STD-461G_53571/

4. "MIL-STD-810H. Department of Defense test method standard: environmental engineering considerations and laboratory tests." Department of Defense, 2019, January 31. http://everyspec.com/MIL-STD/MIL-STD-0800-0899/MIL-STD-810H_55998/

5. Marko, M. D. "The impact of lubricant film thickness and ball bearings failures." Lubricants, 7(6), 48, 2019. https://doi.org/10.3390/lubricants7060048

KEYWORDS: Bearings; Tribology; Lubricants; Failures; Monitoring; L10

TPOC 1: Matthew Marko
(732) 323-5228
Email: [email protected]

TPOC 2: Adam Robertson
(732) 323-496
Email: [email protected]

---