NDE 2017 Conference & Exhibition of the Indian Society for NDT (ISNT), 14-16 December 2017, Chennai,

T.N., India

A Novel Ultrasonic Inspection Methodology for Submerged Metallic Structures

using a Remotely Operated Vehicle (ROV)

Ashish ANTONY JACOB1, Manas GALIPALI1, Vineet UPADHYAY1, a), Prabhu

RAJAGOPAL1,2 and Krishnan BALASUBRAMANIAM1,2
Planys Technologies, IIT Madras Research Park, Chennai – 600113, T.N., India.
1
2
Centre for Nondestructive Evaluation and Dept. of Mechanical Engg., IIT Madras, Chennai – 600036,
T.N., India.
a)
vineet@planystech.com

Abstract
Currently, ultrasonic inspection of submerged structures is performed with Remotely Operated Vehicles (ROVs)
in conjunction with divers. In order to obtain reliable measurements, the ROV attaches itself to the specimen
under examination using a manipulator arm and places the ultrasonic transducer in contact with the test
specimen. However, these methodologies cause the inspection to be cumbersome, difficult to pilot and may
result in high wear to transducer casing. In this paper, the authors present a custom designed ROV based
inspection methodology which does not require contact with the specimen, thereby alleviating stringent
conditions required to perform subsea ultrasonic inspection. Proof of concept ultrasonic pulse-echo experimental
studies with varying standoff distances performed using an ROV in a test facility show promising results.
Advantages and limitations to practical realisation are also discussed.

Keywords: Ultrasonic inspection, ROV, underwater

1
NDE 2017 Conference & Exhibition of the Indian Society for NDT (ISNT), 14-16 December 2017, Chennai, T.N., India

1. Background and Introduction

Submerged structures operating in oil, gas, bridges and other civil industries are
susceptible to internal and external corrosion, manufacturing defects, seismic variations and
other types of problems during operation. Moreover, it is typically predicted that these
structures will operate well beyond their designed life. Periodic inspection of critical
structures is, therefore, of paramount importance in order to prevent catastrophic accidents
resulting in loss of life and property. The approach to subsea inspection is generally
categorised as Type 1 Inspection which involves visual inspection of the structures, Type 2
Inspection which involves detailed visual inspection of selected areas after cleaning of marine
growth, and Type 3 Inspection which specifies surface preparation following a detailed
Nondestructive Evaluation (NDE) of the structure [1]. Even in relatively shallow waters,
subsea inspection is challenging owing to formidable conditions resulting from tidal waves,
marine growth, visibility and accessibility. Conventionally, Type 3 Inspection is performed by
trained divers despite this method being subject to operator error, constraints on the depth of
operation, the speed at which the operation is performed and the inherent risk to human lives.
More recently, such inspection operations are being performed by Remotely Operated
Vehicles (ROV) [2,3], resulting in more efficient, reliable and safe methodologies.
Ultrasonic inspection is a widely used methodology for NDE of submerged structures
in numerous sectors including oil, gas, petrochemical and civil engineering. The
methodologies used for submerged structures closely resemble topside NDE techniques. The
modifications to the topside ultrasonic methodologies typically involve waterproofing and
adaptation of relevant instrumentation to the subsea environment [4]. Conventionally,
ultrasonic testing (UT) of underwater structures is performed by Remotely Operated
Underwater Vehicles (ROV) or divers, which requires them to carry a probe and make an
appropriate contact of the probe with the target surface in order to take the measurement.
ROVs deploy probes mounted on seven degrees of freedom (dof) electro-hydraulic
manipulator arms in order to execute the task [5]. Operation of ROV with the arm is often
prolonged thus resulting in lower productivity. Further, in rough environmental conditions,
there is a significant risk of damaging the probe and the manipulator due to the uncontrollable
additional force exerted on them. Piloting and operational skills of the crew are crucial to the
success of this task [6].
Planys Technologies Pvt. Ltd. (www.planystech.com) is an IIT Madras incubated
startup that provides immersed inspection and survey solutions using indigenously
manufactured submersible ROVs. Planys emerged from and is closely associated with the
Centre for Nondestructive Evaluation (CNDE), IIT Madras (www.cnde.in). In this paper, we
present an ROV based ultrasonic inspection system which allows the pilot to take a thickness
measurement by just performing a slow ‘fly-by’ manoeuvre while directing the ultrasonic
transducer towards the structure under examination. Further details on the components of the
ROV based inspection system are provided in later sections. Requirements of highly trained
pilots and an involved manipulator arm to perform ROV based ultrasonic NDE can be
obviated using the proposed methodology. Further, the inspection can be performed at a faster
rate with improved throughputs. The methodology has the potential of leading to significant
cost savings for ROV operators by reducing the operational complexity, setup time,
dependency on crew skills and need for complicated hardware.
This paper is organised as follows. Firstly, a brief introduction to the ultrasonic
inspection of submerged structures is provided. Then, we describe the design features of the

2
NDE 2017 Conference & Exhibition of the Indian Society for NDT (ISNT), 14-16 December 2017, Chennai, T.N., India

ROV used and the proposed ultrasonic NDE system, following which experimental results are
presented and discussed. The paper concludes with a summary and directions for future work.

3. Methodology
3.1 ROV Beluga
Planys technologies’ indigenously developed ROV Beluga, illustrated in Fig. 1, is a
compact, modular and portable ROV that can sustain 2-4 meters of wave height, moderately
rough swell, up to 1 m/s of sea current and can dive up to 200 meters of depth. Beluga is an
observation class ROV [7] and possesses a state-of-the-art dual hull heavy bottom
hydrodynamic design, thus providing it with static and dynamic stability in rough
environmental conditions. Beluga is equipped with eight thrusters providing control in 5
degrees of freedom along with forward thrust of 17.5 kg-f and an operating speed of 3-4
knots. Rolling of the ROV is inhibited by virtue of its heavy bottom design. Beluga carries an
Inertial Measurement Unit (IMU) for orientation feedback, a pressure sensor for depth
feedback and a Global Positioning System (GPS) for position feedback on the water surface.
Beluga’s modular architecture allows it to carry a variety of different payloads such as high-
definition cameras, high-intensity lights, LASERs for scaling and crack measurement,
altimeter, side-scan Sonar, an ultrasonic thickness measurement unit, cathodic potential
measurement unit, bio-fouling cleaning equipment, 2D & 3D Imaging/Scanning SONARs [8].
Detailed technical specifications of Beluga are provided in Table 1.

Fig. 1 A photograph of ROV Beluga highlighting its key features

Parameter Specifications and Features

Dimensions 730 x 820 x 600 [mm]
Depth range 200 m
Total weight in air 53 kg
Maximum Payload 10 – 15 kg
Power rating 3000 W (maximum)
Endurance Unlimited
Control System Auto-depth, Auto-Heading
Depth Sensor 0.01 % Full-Scale Accuracy; Temperature Compensated
Table. 1. Design parameters, specifications and key features of the ROV Beluga developed by
Planys Technologies

3
NDE 2017 Conference & Exhibition of the Indian Society for NDT (ISNT), 14-16 December 2017, Chennai, T.N., India

Beluga is loaded with capabilities to perform bathymetric surveys using SONARs to

study seabed topography, flooded member detection on offshore platforms by using acoustic
techniques, hydrographic surveys to characterise water properties such as conductivity,
temperature, pH, dissolved oxygen (DO), turbidity and oxidation-reduction potential. Various
other different types of sensors or payloads can be integrated through multiple plug-and-play
interfaces available on the ROV. Many new capabilities are under development such as
cleaning of marine and bio-fouling growth from marine installations, pipeline surveys and
concrete inspection. Beluga has provided frequent services in the offshore oil & gas sector,
shipping industry (hull, propeller, rudder and ballast tank inspection), ports & marine
terminals, dams, power plants, thermal plants and nuclear plants, including visual and non-
destructive techniques (NDT) based inspection of immersed underwater structures.
3.2 Ultrasonic Inspection System
At the core of the developed ultrasonic inspection is the processing unit consisting of both
hardware (pulser-receiver, data acquisition units, analog to digital convertors) and software.
As briefly explained earlier, the methodologies for performing NDE of submerged structures
using an ROV are similar to topside inspection systems. The problems and challenges which
hinder inspection to these topside systems also continue to exist. Conventionally, the effects
of ROV instability are mitigated by using a manipulator arm to clamp the test specimen or its
surrounding structures. However, in the proposed methodology, to develop a reliable and
robust NDE tool, and to overcome various challenges including ROV instability and probe
non-perpendicularity to the test specimen, custom designed software and hardware have been
developed by Planys Technologies.
The reflected signal obtained from pulser receiver unit from the target specimen is
processed over various steps in the data processing unit which includes implementation of
signal filters, smoothing algorithms, peak detection algorithms, heuristic algorithms for
thickness measurement and data selection algorithms. The software is developed as an
Application Programming Interface (API) and is customizable to perform both contact and
immersion ultrasonic inspection. The detailed data process flow and system architecture are
shown in Fig 2.

Fig. 2. Illustration of process and control flow of the software to enable real-time
measurement of thickness of the test sample under examination

4
NDE 2017 Conference & Exhibition of the Indian Society for NDT (ISNT), 14-16 December 2017, Chennai, T.N., India

The digital signal from the pulser-receiver unit is considered as the raw signal data,
shown in Fig. 3 (a), which is transferred to the thickness measurement API developed to
obtain the metal thickness. The configuration settings are controlled using graphical user
interface (GUI) or can be selected by the API calls. Based on the API configuration settings,
RSD is processed through noise filter or a set of noise filters. The filtered signal data is further
processed over a set of smoothing algorithms to obtain a smooth signal data, and then an
optimised Hilbert transformation is implemented to produce Hilbert Signal Data (H-SD). On
H-SD, the Heuristic peak detection algorithm (HPDA) is applied to detect peaks, based on the
type of measurement the output of HPDA. The heuristic thickness measurement algorithm is
used to obtain a thickness of the target metal structure, and also is used to calculate the length
of water column between the probe and the metal structure. Thickness values are collected
over a user-defined or machine learnt time period and are optimised using pseudo thickness
optimisation algorithm based on signal amplitudes of surface reflections and number of re-
occurrences of the same thickness estimate.
Step by step sample amplitude vs time (A-scan) results obtained using the process
flow control described in Fig. 2, when contact ultrasonic measurements were performed on 8
cm thick Aluminum test specimen, are illustrated in Fig. 3. The A-scan given in Fig. 2 (a)
shows the raw data with elevated noise levels obtained from the pulser-receiver. We note here
that the main source of noise is from the hardware. Subsequent steps, as described in Fig. 2
are implemented to obtain the A-scans indicating multiple reflections from the free surface of
the test specimen as shown in in Figs. 3 (b-d).

(a) (b)

(c) (d)
Fig. 3. Sample amplitude vs time plots obtained as (a) raw signal from pulser-receiver, (b) A-
scan after applying noise filters and signal smoothing algorithms, (c) A-scan upon application
of Hilbert Transform and (d) A-scan indicating peaks used for the estimation of the thickness
of the test sample.

5
NDE 2017 Conference & Exhibition of the Indian Society for NDT (ISNT), 14-16 December 2017, Chennai, T.N., India

3.3 Experimental Arrangement

Experiments are performed in controlled water conditions with a 6 mm thick stainless
steel plate as the test specimen. The ROV uses a custom designed probe assembly to perform
the ultrasonic inspection. The ultrasonic probe is enclosed inside a 100 m depth-rated Delrin
watertight enclosure with a single O-ring face seal. The structural elements making the probe
assembly are shown in Figs. 4 (a-b). The UT probe mates tightly with an impedance matched
layer which is precisely manufactured to ensure maximum ultrasonic transmission from the
ultrasonic transducer. The entire assembly is fastened using tightening discs and nut-bolts to
make sure the O-ring is compressed sufficiently to avoid water ingress. This assembly allows
the underwater usage of any commercially available ultrasonic transducer without any
significant performance loss. To prevent any mating issues between the matching plate and
the probe arising due to vibrations, a spring mount inside the enclosure has been provided.

(a) (b)
Fig. 4. The custom designed ultrasonic transducer enclosure (a) a schematic illustrating the
exploded view, and (b) a photograph of its components

The experiments are performed using a commercially available ultrasonic transducer

(Olympus, Panametrics, Japan) with a center frequency of 2.25 MHz and a nominal diameter
of 1 inch. However, these parameters may be changed as per the inspection requirements.
Using the user interface of the inspection system on the pilot's control station, numerous
parameters controlling the excitation of the ultrasonic transducer can be chosen including
recording data length, gain, input voltage and filters on the basis on the nature of the test
specimen and the standoff distance of the ROV from the test specimen. A schematic of the
experimental arrangement is shown in Fig. 5. The ROV is deployed into a test facility with a
water column 4 meters high. The ROV is directly towards the test specimen by the pilot using
various controls available in the control station [2]. Once the ROV is close enough, the pilot
engages the ultrasonic inspection system to begin recording amplitude vs time plots. These
results are presented in the following section.

Fig. 5. A schematic illustrating the experimental arrangement in a test facility

6
NDE 2017 Conference & Exhibition of the Indian Society for NDT (ISNT), 14-16 December 2017, Chennai, T.N., India

4. Results and Discussion

In this proof of concept experimental study, amplitude vs time plots were recorded
with varying standoff distances of the ROV from the test specimen. Sample A-scans obtained
during the scan trials at standoff distances 50 mm and 100 mm are shown in Fig. 6. Table 2
indicates the measured thickness using the developed system along with the error in
measurement against different standoff distances of the ROV from the specimen.

Fig. 6. Sample amplitude vs time plots obtained during the ROV based ultrasonic
measurements using the custom designed software at standoff distances (a) 50 mm, and (b)
100 mm in a test facility

Thickness of the specimen (Vernier in mm) = 5.95

Sl. Standoff Thickness measured using ultrasonic
No. distance (mm) measurements (mm)
1 50 6.05
2 100 6.1
3 150 5.85

Table 2. Sample measured thickness of the test specimen at different standoff distances of the
ROV, measured using the developed ROV based ultrasonic inspection system
We observe that the scans, in Fig. 6, obtained from the custom-designed software are
satisfactory with the SNR > 20 dB. Multiple echoes from the specimen’s free surfaces are also
seen, which is used to improve the reliability of measurement taken by the developed
software. While the developed inspection system offers significant advantages over
conventional inspection routines, the system has the following limitations. The system is
highly sensitive to disturbances induced in the ROV due to varying environmental forces
acting on it which often leads to loss of reflected echo or reduced amplitudes. Further, the
received signal amplitude is significantly lower in the case the specimen has marine growth
over it.

5. Conclusion and Future Work

This paper describes a novel ROV based ultrasonic inspection system for NDE of
submerged metallic structures. Design details and critical features of ROV Beluga are

7
NDE 2017 Conference & Exhibition of the Indian Society for NDT (ISNT), 14-16 December 2017, Chennai, T.N., India

described, followed by a description of the NDE system’s processing unit. Experimental

studies conducted in a test facility using the developed ROV based inspection methodology
provide satisfactory results.
Future work is directed towards developing compensational systems to the ultrasonic
measurements to improve its reliability using the ROV’s inertial measurement unit (IMU)
Further, compensations to the input parameters to the ultrasonic transducer are also to be done
to obtain more reliable and accurate ultrasonic measurements.

6. References

[1] G. Hayward, J. Pearson and G. Stirling, An Intelligent Ultrasonic Inspection System

for Flooded Member Detection in Offshore Structures, IEEE Trans. Ultrason.
Ferroelectr. Freq. Control. 40 (1993) 512–521. doi:10.1109/58.238103.
[2] T. Jhunjhunwala et al., MIKE: A remotely operated vehicle (ROV) for visual
inspection of underwater structures, in: Natl. Semin. Exhib. Non-Destructive Eval.
NDE 2015, Hyderabad, 2016.
[3] T. Jhunjhunwala and P. Rajagopal, Remotely operated vehicle (ROV) for inspection of
underwater pipelines using bulk and guided ultrasonic waves, in: Natl. Semin. Exhib.
Non-Destructive Eval. NDE 2015, Hyderabad, 2016.
[4] P. Rizzo, NDE/SHM of Underwater Structures: A Review, Adv. Sci. Technol. 83
(2012) 208–216. doi:10.4028/www.scientific.net/AST.83.208.
[5] S.W. Moore, H. Bohm and V. Jensen, Chapter 10. Hydraulics and Payloads, in:
Underwater Robotics: Science, Design and Fabrication, 2013: pp. 540–580.
[6] R.D. Christ and R.L. Wernli, Chpater 8. Manipulators. The ROV Manual - A User
Guide for Remotely Operated Vehicles The ROV Manual A User Guide for Remotely
Operated Vehicles, 2014: pp. 522-600.
[7] J. Snyder, Doppler Velocity Log (DVL) Navigation for Observation-Class ROVs,
IEEE Ocean. 2010 MTS/IEEE SEATTLE. 978 (2010) 1–9.
[8] V. Upadhyay, S. Gupta and P. Agarwal, Multi-functional Remotely Operated
Submersible Vehicle (ROV) System, 2016.