Micrometeoroid/Orbital Debris (MMOD) Impact Detection and Location ...

Available online at www.sciencedirect.com

ScienceDirect Procedia Engineering 188 (2017) 233 – 240

6th Asia Pacific Workshop on Structural Health Monitoring, 6th APWSHM

Micrometeoroid/Orbital Debris (MMOD) Impact Detection and Location Using Fiber Optic Bragg Grating Sensing Technology Steven L. Rickmana*, W. Lance Richardsa, Ph. D., Eric L. Christiansenb, Ph. D., Anthony Piazzac, Francisco Penac, Allen R. Parkerc a

NASA Langley Research Center, Hampton, Virginia 23681, USA, bNASA Johnson Space Center, Houston, Texas 77058, USA c NASA Armstrong Flight Research Center, Edwards, California 93523, USA

Abstract Spacecraft in low earth orbit (LEO) experience a variety of hazards including exposure to micrometeoroids and orbital debris (MMOD). Average impact speeds for orbital debris on spacecraft in LEO are 9 to 10 km/s, and 20 km/s for micrometeoroids. Due to their high speeds, MMOD can cause considerable impact damage to sensitive spacecraft surfaces such as windows, structural elements, electronic boxes, solar arrays, radiators, thermal protection system (TPS) materials covering crew/cargo return vehicles, as well as crew modules. Prolonged exposure to the on-orbit MMOD environment can potentially compromise the TPS covering return vehicles such as the future crewed vehicles expected to visit and remain for a half-year or longer at the International Space Station (ISS). However, determination of MMOD impact damage to the TPS on crew/cargo return vehicles, or damage to other orbiting spacecraft currently requires visual inspection. For human-rated spacecraft such as the ISS and, previously, the Space Shuttle Orbiter, this has required crew time as well as vehicle assets to identify damage due to MMOD strikes. For unmanned spacecraft, there are no human assets present to conduct detailed surveys to identify potential damage. While the practice of visual inspection may successfully indicate the location of a debris strike, it does not currently allow precise determination of exactly when the debris strike occurred. It is only possible to determine that a debris strike has occurred between two successive inspection events unless damage to other components can allow inference that a debris strike occurred at a specified time. Initial development testing of a novel MMOD sensing concept amenable for potential implementation in a structural health monitoring system is discussed. Using fiber Bragg grating (FBG) sensing technology, a test article representative of a spacecraft MMOD shield layer or spacecraft structure was subjected to hypervelocity impact testing. The FBG array successfully detected that an impact had occurred, when it occurred and where it occurred on the structure. Detection of both the impact wave and the residual strain in the structure can be used to infer the location of the impact. The potential application of this technology to spacecraft is discussed. © 2016 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). © 2016 The Authors. Published by Elsevier Ltd. Peer-review under responsibility of the organizing committee of the 6th APWSHM

* Steven L. Rickman. Tel.: +1-281-483-8867; fax: +1-281-483-3861. E-mail address: [email protected]

1877-7058 © 2016 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of the 6th APWSHM

doi:10.1016/j.proeng.2017.04.479

234

Steven L. Rickman et al. / Procedia Engineering 188 (2017) 233 – 240 Keywords: Micrometeoroid, orbital debris, MMOD, fiber optic strain sensing, FBG, damage, impact detection, location

1. Introduction Spacecraft in Earth orbit are subjected to many hazardous environments including micrometeoroids and orbital debris (MMOD). Prolonged exposure to the on-orbit MMOD environment increases risk to vehicles including commercial crew and cargo vehicles expected to visit and remain for considerable periods of time at ISS. These vehicles are covered by thermal protection system (TPS) materials to protect the vehicle from reentry heat during the return of crew and cargo to Earth. MMOD can compromise the integrity of the TPS and result in loss of the vehicle during return unless the damaged TPS is detected prior to return. However, determination of MMOD impact on orbiting spacecraft currently requires visual inspection. For human-rated spacecraft (such as the ISS and, previously, the Space Shuttle Orbiter), this has required crew time as well as vehicle assets to identify damage due to MMOD strikes. For unmanned spacecraft, there are no human assets present to conduct detailed surveys to ascertain potential damage. Orbital debris fragments are generated by on-orbit explosions, collisions, breakups and degradation and is a growing threat as suggested by recent orbital debris models [1]. Two recent events, the Chinese ASAT test in 2007 and Cosmos/Iridium collision in 2009 significantly increased orbital debris for satellites in sun-synchronous orbit [2]. The MMOD threat is real and poses considerable risk to crewed and un-crewed orbiting spacecraft. While the current practice of visual inspection may successfully indicate the location of a debris strike, it does not allow for precise determination of exactly when the debris strike occurred. It is only possible to determine that a debris strike has occurred between two successive inspection events unless damage to other components can allow inference that a debris strike occurred at a specified time. However, development of a method and apparatus using fiber Bragg grating (FBG) sensing technology has been employed to demonstrate the ability to determine that an MMOD impact has occurred on a spacecraft MMOD shield or structure, when it occurred, and importantly, where it occurred. The extent of damage can be inferred from the FBG response to the impact, which serves as an important step towards conducting structural health monitoring (SHM). This paper describes the FBG system and results from testing the system to detect the location of hypervelocity particle impacts. The FBG system can be applied in a variety of space and ground applications where structural health monitoring of impact damage is required. 2. Background A considerable amount of research of has been conducted over recent years focused on using FBG sensing technology to detect impact damage in structural components [3-16]. Most approaches employ or adapt nondestructive evaluation (NDE) methodologies to demonstrate impact and damage detection capability. The common approach in the literature is to monitor the FBG signals at very high frequencies in order to characterize ultrasonic and/or acoustic waves that emanate from the damaged region at the impact location. These approaches infer damage from wave propagation and time of flight calculations between the impact location and the FBG sensor response. Dynamic signal processing techniques are used to ascertain if damage has occurred, its location and the severity of the damage. While these approaches have proven effective at detecting impact in theoretical and small scale laboratory studies, they are often challenging to implement on large-scale, realistic structures, such as the ISS, and under realistic highenergy impact conditions that can occur during ISS operation. A vast majority of the papers are based on analytical studies or on simple experiments in which impact damage was simulated by unrealistic means. Aside from a few exceptions, few examples were found where proposed methods have been experimentally demonstrated and quantified for real-world applications and amenable to in-situ deployment.

Steven L. Rickman et al. / Procedia Engineering 188 (2017) 233 – 240

3. Test Objectives and Approach The objectives of this project were to accurately and reliably ascertain the following from hypervelocity MMOD impacts: 1) that an actual damage event has occurred to a representative structure, 2) the time at which damage event occurred, 3) its location, and 4) to provide a practical method to quantify the severity of the damage event in which inspection and remediation plans can be developed. To achieve these objectives, a highly-multiplexed distributed array of fiber Bragg gratings (FBG) were bonded to the surface of a representative test article and the in-plane strain fields were measured. The test article was then impacted with a realistically-simulated MMOD projectile traveling at hypervelocity speeds. Following the impact, the post-impact residual strains caused by the damage were analyzed to detect that damage occurred and the time at which damage occurred. Numerical processing was then used in conjunction with the residual strains measured by the distributed array of FBGs to determine the location of the damage event. The long-term goal of the effort is to employ calibration techniques and/or learning algorithms to quantify the severity of the damage. 4. Test Description The proof-of-concept system was tested by NASA’s Hypervelocity Impact Technology (HVIT) group [17] using a two-stage light-gas launcher at the White Sands Test Facility (WSTF). A 1.6 mm thick aluminum 2024-T3 Al-clad plate measuring 38 cm × 38 cm with a test section measuring 30.5 cm × 30.5 cm, representative of a MMOD shield outer layer or spacecraft structure, was mounted in a frame and instrumented with an optical fiber with 36 strain sensors. The test article configuration is depicted in Fig.1. The article was tested in the 4.3 mm caliber range with spherical aluminum 2017-T4 projectiles. Table 1 summarizes test conditions for six hypervelocity shots performed on the FBG target, and Fig. 2 depicts the approximate impact location on the target for the tests. The test setup in the impact lab is shown in Fig. 3 including sensor monitoring equipment and data display that indicated impact location in near real-time. Strain sensor signals were recorded by the data acquisition system at 5 kHz sampling rate.

Fig. 1. Aluminum plate target with attached fiber optic strain sensors viewed from the back of the plate.

Fig. 2. Impact locations for the hypervelocity tests. Arrows with locations 4 and 5 indicate flight direction for oblique impacts.

235

236


Fig. 3. Test setup

4.1. Sensor Layout A sensor array consisting of FBG strain sensors was designed to maximize sensitivity to a strain gradient that radiates from the center of the test article out to the boundary test article boundary. A Bragg grating is only sensitive to strain along its primary axis, so alignment towards the center of the test article was given high priority. Given the size of the test article target area (30.5 cm × 30.5 cm), and a sensor density of 1 Bragg grating per 0.3937 cm (i.e., 1 inch) of sensing fiber, it was determined that the best coverage over the test article area was achieved with 36 Bragg gratings, with the primary axis of the Bragg gratings aligned towards the center of the test article. Fig. 4 shows the sensor layout over the test article. The physical location of each fiber Bragg grating was recorded and was used to develop a sensor map.

Fig. 4. Fiber Bragg grating instrumentation layout

4.2. FBG Data Acquisition System The FBG data acquisition system used for this application (Fig. 3) was designed to interrogate and record data from a single fiber with 36 wavelength division multiplexed (WDM) sensors installed on the test article utilizing just a few components. Figure 5 shows a block diagram of this system, which consisted of a broadband light source, an optical fiber circulator and spectrometer, standard computer and, of course, the sensing fiber.

237


Fig. 5. FOS System Block Diagram

All components were commercially available with exception to the sensing fiber, which was custom ordered for this application. The broadband light source was a 10 mW wide-band laser nominally operating at 1550 nm over an 80 nm range. The optical fiber circulator was a 3-port device hosting a flattened broadband wavelength range of operation at 1550 nm allowing light to travel in only one direction. The three-way port device allowed light from the broadband source to enter port 1 and exit port 2 with minimal loss, while concurrently allowing the reflected light from the sensors to enter port 2 and exit port 3 with minimal loss. This configuration allows for optimum power utilization of the broadband source. The reflected light was then sampled by an optical spectrum analyzer (OSA). The OSA operates over an 80 nm optical range centered at 1550 nm. The device was attached via a Universal Serial Bus (USB) interface to a standard laptop (Interrogation computer) for data processing, display and recording. The system as described was designed specifically for the purposes of this project and was capable of sensing up to 36 FBGs simultaneously at a rate of 5 KHz. Each sensor is limited to a +/-1 nm range which translates to approximately +/-800 με. Sensor data from the 36 FBGs were continuously recorded during each test and post processed.

4.3. Test Matrix The test article was installed in a test chamber at the Remote Hypervelocity Test Lab (RHTL) and data from six hypervelocity shots were acquired using the 4.33 mm caliber range. Table 1 summarizes the test matrix used in the test program, which included a variety of projectile sizes and incidence angles for various target locations on the test article. Table 1. Hypervelocity test conditions Test Sequence

Impact Location

Projectile Diameter (mm)

Projectile Mass (g)

Projectile Velocity (km/s)

Impact Angle (deg)

Description of target damage and damage diameter

1

6

0.30

0.00004

6.85

0

Crater, 1.5 mm

2

1

1.0

0.00143

7.10

0

Hole, 3.4 mm

3

2

0.5

0.00017

6.98

0

Hole, 0.5 mm

4

4

1.0

0.00148

7.00

45

Hole, 3.1×2.9 mm

5

5

0.5

0.00019

6.98

45

Crater, 2.4×2.3 mm

6

6

1.0

0.00147

6.98

0

Hole, 3.4 mm

238


4.4. Impact Detection Method Prior to deployment of the test article, the steady-state spectral distribution/strain distribution was captured and used as a reference state of the structure. The test article was then deployed into operation and continuously monitored during testing. A short time history for each Bragg grating was buffered in temporary memory and the time rate of change of each strain measurement was tracked. An event triggering system was activated by monitoring threshold on high strain rates in order to find the time of an impact. 5. Results As stated previously, the first two test objectives were to detect if an impact event had occurred, and if so at what time did the event occur. To detect an impact event, the time rate of change of each strain measurement was tracked. To trigger the event indicator, both the time rate of change of the strain and measured max strain value needed to exceed a threshold. Prior to test, the noise of the system was estimated to produce a measurement uncertainty of approximately -2 to +2 με peak-to-peak. In all six of the test sequences, the max strain detected by the FBG was well above 10 με and therefore statistically outside of the noise band. Utilizing both the event trigger and the time stamp associated with each measurement, it was then possible to detect that an event occurred and the approximate time the event occurred. An example of the time history of an individual (i.e., 1 of 36) fiber Bragg grating measurement during Test Sequence 3 is shown in Fig. 6. The max strains measured by the FBG for all load sequences are listed in Table 2. It is also noted that the peak strain experienced by the test article is anticipated to have been at a much higher magnitude than the FBG was able to detect; this is due to the limited number of sensors and the limited sample rate of the FBG used during the tests.

Fig. 6. Example time history of an individual fiber Bragg grating, Test Sequence 3

After detecting that an impact event has occurred, the next objective is to utilize the strain measurements to estimate the location of the impact, and the severity of the damage. The post impact residual strain field was used to estimate the damage caused by the simulated MMOD the impact. The post impact residual strain field is found by using the system event trigger to save a recording of the strain distribution time history starting several seconds before the impact event, and lasting until the post impact oscillations dissipate. The difference between the two states is used as the post impact residual strain. The presence of residual strain field indicates that the structure experienced plastic yielding and has accumulated some level of damage. The presence of a residual strain field also enabled the development of an automated strain based impact location estimation algorithm. The impact location was found by utilizing the post impact residual strain distribution to identify at least three (minimum) sensing stations with the highest strain magnitude, which were then stored in a subgroup. The strain magnitudes and physical locations (i.e., x,y coordinates) of the sensor subgroup can then be used with a series of algebraic equations to determine the impact location (x,y). The results from the automated impact location estimation algorithm are presented in Fig. 7 (a) to (f).

239


(a)

(b)

(c)

(d)

(e)

(f)

Fig. 7. Residual strain field, actual impact location (black circle), estimated location (purple circle). (a) Test Sequence 1, (b) Test Sequence 2, (c) Test Sequence 3, (d) Test Sequence 4, (e) Test Sequence 5, (f) Test Sequence 6

The results of the impact location estimation algorithm are presented in Table 2. The impact location estimation for Test Sequence 1 indicated the largest difference between the actual location and the estimated location, which was due to absence of a residual strain field above the noise band. Test Sequence 1 utilized a projectile with the smallest mass (0.00004 g) and the smallest projectile diameter (0.3 mm) of all the test sequences. Test Sequence 2 through Test Sequence 6 each had a max post-impact strain above the 2 με noise floor, and therefore resulted in a suitable signal for the strain based impact location estimation algorithm. The impact location estimation algorithm was within 1.5 cm for the remaining Test Sequences 2 through 6, and approximated the impact location as close as 0.1 cm. Further modifications to the automated impact location estimation algorithm may yield improved results. Table 2. Test Measurement Summary Test Sequence

Max strain detected (με)

Max post impact strain (με)

1

18

1

2

48

3

34

4

Actual Impact Location

Strain Based Estimated Location

Difference

x (cm)

y (cm)

x (cm)

y (cm)

Total distance (cm)

0.3

0.2

-3.3

8.1

8.7

10

5.4

11.9

5.0

12.2

0.5

12

-9.3

3.6

-10.1

3.9

0.9

58

12

-5.0

-10.8

-4.8

-10.7

0.3

5

46

14

10.8

-1.8

10.2

-3.1

1.4

6

58

7

0.5

0.6

0.6

0.6

0.1

6. Conclusions The MMOD threat to spacecraft, both crewed and un-crewed is real. The proof-of-concept testing and analysis described in this paper has resulted in a method and apparatus for detecting and locating MMOD impacts on spacecraft. The initial test series demonstrated the feasibility of using fiber Bragg grating sensor technology for orbital debris impact sensing as the data obtained from the test showed that an impact had occurred, when it occurred and, importantly, where it occurred. Implementation may be beneficial to all spacecraft that fly in this environment as it

240


can assist in assuring the safety of crewed spacecraft prior to entry and also aid in monitoring the health status of unmanned spacecraft. 7. Acknowledgements The authors wish to thank the staff of the Remote Hypervelocity Test Laboratory (RHTL) at NASA’s White Sands Test Facility and Mr. Jon Read of the Hypervelocity Impact Technology (HVIT) group at the NASA Johnson Space Center for their testing support during this study. 8. References [1] E. G. Stansbery, M. J. Matney, P. H. Krisko, P. D. Anz-Meador, M. F. Horstman, J. N. Opiela, E. Hillary, N. M. Hill, R. L. Kelley, A. B. Vavrin, D. R. Jarkey, NASA Orbital Debris Engineering Model ORDEM 3.0 - User’s Guide, NASA/TP-2014-217370, April 2014. [2] NASA Johnson Space Center, Satellite Collision Leaves Significant Debris Clouds, http://www.orbitaldebris.jsc.nasa.gov/newsletter/pdfs/ODQNv13i2.pdf, April 2009. [3] O. Shapira, S. Kedem, B. Glam, N.Y Shemesh, A. Dvorjetski, N. Mashiach, J. Balter, R. Shklovsky, I. Sovran, N. Gorbatov, I. Kressel, M. Tur, Implementation of a fiber-optic sensing technology for global structural integrity monitoring of UAVs, Conference Paper, February 2014. [4] L. Pei, T. Jianwei, Study on fiber Bragg grating strain sensing array detecting multi-crack damage of cantilever beam. Proc. of SPIE Vol. 9656, 2015. [5] K. Choi, F-K.Chang, “Identification of impact force and location using distributed sensors”, AIAA Journal Vol. 34, No 1, (1996) 136-142. [6] R. Di Sante, Fibre optic sensors for structural health monitoring of aircraft composite structures: recent advances and applications, J. Sensors (2015) 18666-18713. [7] M. P. Connolly, The detection of impact damage in composite pressure vessels using source location acoustic monitoring, Proceedings of the 31st Joint Propulsion Conference and Exhibit, San Diego, CA, July 1995. [8] K. Kahl and J. S. Sirkis. Damage detection in beam structures using subspace rotation algorithm with strain data", AIAA Journal, Vol. 34, No. 12 (1996) 2609-2614. [9] J. K. Shaw, J. S. Sirkis, E. J. Friebele, R. T. Jones, and A. D. Kersey. Model of transverse plate impact dynamics for design of impact detection methodologies, AIAA Journal, Vol. 33, No. 7 (1995) 1327-1334. [10] J. S. Sirkis, K. Shaw, T. A. Berkoff, A. D. Kersey, E. J. Friebele, and R. T. Jones, Development of an impact detection technique using optical fiber sensors, Proc. of SPIE Vol. 2191, 1994, pp. 158-165. [11] G. R. Kirikera, O. Balogun and S. Krishnaswamy, Adaptive fiber Bragg grating sensor network for structural health monitoring: applications to impact monitoring, J. Structural Health Monitoring, Vol. 10 No. 1 (2011) 5-16. [12] Y. Sai, M. Jiang, Q. Sui., L. Jia, and S. Lu, Low velocity impact localization system using FBG array and MVDR beamforming algorithm, J. Photonic Sensors Vol. 5, No. 4 (2015) 357–364. [13] H. Guo, G. Xiao, N. Mrad, J. Yao, Fiber optic sensors for structural health monitoring of air platforms, J. Sensors (2011) 3687-3705. [14] S. Takeda, S. Minakuchi, Y. Okabe, N. Takeda, Delamination monitoring of laminated composites subjected to low-velocity impact using small-diameter FBG sensors, J. Composites Part A: Applied Science and Manufacturing, Vol. 36 (2005), Issue 7 903–908. [15] P. Shrestha, J-H. Kim, Y. Park, C-G. Kim, Impact localization on composite wing using 1D array FBG sensor and RMS/correlation based reference database algorithm, J. Composite Structures, Vol. 125 (2015) 159–169. [16] Z. M. Hafizi, J. Epaarachchi, K.T. Lau, Impact location determination on thin laminated composite plates using an NIR-FBG sensor system Measurement, Vol. 61 (2015) 51–57. [17] NASA Johnson Space Center, Hypervelocity Impact Technology Group, http://ares.jsc.nasa.gov/HVIT/index.cfm.