Professional Documents
Culture Documents
Ultrasonic Range: Neural Vehicles
Ultrasonic Range: Neural Vehicles
Related terms:
Neural Vehicles
Ben Kröse, Joris van Dam, in Neural Systems for Robotics, 1997
Collision-Free Navigation
Instead of using vision for road following, similar approaches have been presented
which use ultrasonic range measurements in order to navigate without collisions.
An example is the work carried out within the Esprit project ANNIE [32]. In this
work a simulated vehicle is equipped with sensors which measure the range to
objects which are nearest within their field of view. The corresponding range values
are assigned logarithmically to a discrete set of values, in correspondence with the
sensitivity to near obstacles. The 16 range values are coded in 4 bits (the authors do
not indicate why this input representation was chosen). There are 9 sensors, resulting
in a network input of 36 (binary) input units. The network has 8 hidden units and an
output layer of 3 units, which give a coded representation of the 8 different actions.
The network was trained supervisedly by navigating it around 10 different obstacles
separately. After teaching all obstacles, the vehicle is able to drive without collisions.
Since the wave is guided by the geometric boundaries of the medium, the geometry
has a strong influence on the behavior of the wave.18−20 In contrast to ultrasonic
waves used in conventional ultrasonic inspections that propagate with a constant
velocity, the velocity of the guided waves varies significantly with wave frequency
and geometry of the medium. In addition, at a given wave frequency, the guided
waves can propagate in different wave modes and orders.
The properties of guided waves and examples of their dispersion curves (which
refer to the relationship between the velocity and the wave frequencies) are given in
Figs. 12.5 and 12.6 for pipe and plate geometries, respectively. In pipe, the guided
waves exist in three different wave modes: longitudinal (L), torsional (T), and flexural
(F). In plate, they exist in two different wave modes: longitudinal that is generally
called ‘Lamb’ waves and exists in symmetric (S) and antisymmetric (A) modes; and
shear horizontal (SH).
12.5. Examples of dispersion waves of various guided wave modes in pipe (for
114-mm-OD, 8.6-mm-thick pipe). The numbers in parentheses indicate the order
of the wave mode.
12.6. Example of dispersion curves of various guided wave modes in plate (for
6.35-mm-thick plate). The numbers after the letter (0 and 1) indicate the order of
the wave mode.
Although the properties of guided waves are complex, with judicious selection and
proper control of wave mode and frequency, the guided waves can be used to achieve
100% volumetric inspection of a large area of a structure from a single sensor
location.
Guided waves can be developed at a single location on a component and then travel
long distances from the source point to monitor the condition of that component.
The guided wave fills the entire volume of the component, and any change in
cross-section of the component, such as a weld or corrosion/erosion, usually reflects
or scatters the guided wave so that the change in cross-section is detected. Depend-
ing on frequency and material conditions, guided waves can travel as far as 150 m or
more from one source, and defects such as corrosion scatter back the guided waves,
thus giving an indication of corrosion locations.
There are two types of guided wave systems based on the physics of generating
the guided waves. One type uses a large number of piezoelectric transducers that
are pressure coupled to the pipe. This technology was developed and distributed
by Guided Wave Ultrasonics Limited21 and requires a mechanical fixture to squeeze
the transducers onto a pipe. This fixture can be expensive and can only be used on
cylindrical-type geometries such as pipe. The defect detection sensitivity is estimated
to be 5 to 10% of the pipe cross-section.
The other type uses a thin ferromagnetic strip (approximately 0.15 mm) that has a
high magnetostrictive property and is bonded to the pipe21−28 and is referred to as
magnetostrictive sensor (MsS) technology. The ferromagnetic strip and excitation
coil are relatively inexpensive and can be used on piping, plate, bar and a wide
range of geometries. However, time for bonding the ferromagnetic strip onto the
geometry is required.
Generally speaking, the MsS approach can be used as an inspection mode to detect
defects on the order of 3 to 5% of the component cross-section, but if the sensor
can be left in place for long periods of time (most feasible with the ferromagnetic
strip concept) and data are collected at various times during its life, defects as small
as 0.5% can be detected. However, guided waves tend to be poor at quantifying the
corrosion and should really be used as more of a screening and locating tool.
Guided waves have been used to inspect various types of pipeline, on land as well as
offshore. An example of an offshore application is the inspection of risers. Offshore
pipeline failure statistics have been collected for more than 30 years and illustrate
that the riser predominantly fails as a result of corrosion. The consistent wetting
and drying in the splash zone, combined with defects in the coatings, are the
usual contributors to the problem. A guided wave approach using torsional waves
can be deployed to detect significant reduction in wall thickness. This allows the
detection of wall-thinning areas in the riser without removing the outside coatings.
For detecting and monitoring corrosion under insulation, the MsS can be placed
on a pipe with many welded sections as shown in Fig. 12.7, with the data shown in
Fig. 12.8. The directionality of the guided wave from the MsS is controlled so the
wave can be transmitted to the right of the sensor or to the left.
Presently, the MsS technology uses the T-wave mode primarily for piping inspection
that is generated and detected with the thin ferromagnetic (typically nickel) layer
approach. Reasons for this practice include: (1) the fundamental T-wave mode is
not dispersive and, therefore, no consideration is necessary for possible dispersion
effects that exist in the L-wave mode; (2) the T-wave MsS has fewer effects than
other extraneous wave modes and, therefore, it gives better signal-to-noise ratio
and its data are easier to analyze; (3) the T-wave does not interact with liquid
inside the pipe and, therefore, is far superior to the L-wave24 for inspection of
liquid-filled pipes; and (4) the T-wave MsS does not require heavy bias magnets
and, therefore, is much easier and safer to handle than the L-wave counterpart. The
disadvantage of the T-wave MsS is the requirement for direct physical access to the
pipe surface for bonding of the thin ferromagnetic layer. Therefore, to apply the
T-mode MsS for bitumen-coated piping, for example, the coating must be removed
beforehand, whereas the L-mode MsS can be applied without removing the coating.
However, the advantage of the T-mode MsS greatly outweighs the disadvantage and,
consequently, the T-mode MsS is used primarily for long-range piping inspection.
Table 12.1 summarizes the capabilities and limitations of the present guided wave
technology for long-range piping inspection in bare pipes. The effects of pipeline
geometric features and other conditions such as coating and liquid inside pipe on
the inspection capabilities are also summarized in Table 12.2.
Table 12.1. Capabilities and limitations of guided wave systems for pipe inspection
Item Capabilities/limitations
Spatial resolution 3 to 7 inches (7.5 to 17.8 cm), depending on
frequency and mode
Pipe material Any material
Pipe size Up to 60 in. (152 cm) diameter for MsS technolo-
gy, approximately 32 in. (81.3 cm) for piezoelectric
system and less than 0.75 in. (19 mm) wall thick-
ness
Inspection range 100 ft (30 m) or greater, depending on coating and
for aboveground piping
Detectable defect type Isolated corrosion pits and circumferential cracks
Minimum detectable defect size 2 to 5% of pipewall cross section
Defect location Axial location within ± 2 in. (5 cm), depending on
operational frequency and mode
Defect characterization Limited to rough estimation of circumferential
cross-section
Long term monitoring Only with MsS system because cost of MsS sen-
sors are much less than piezoelectric sensor belt
Features/conditions Effects
Flange/valve Prevents wave propagation; forms end point of
inspection range
Tee Causes a large disruption in wave propagation
and limits inspection range up to that point
Elbow Causes a large disruption in wave propagation
and limits inspection range no farther than the
elbow region
Bend Has negligible effect if the bend radius is greater
than three times the pipe OD; if the bend radius
is less than the above, behaves like an elbow
Side branch Causes a wave reflection and thus produces a
signal; no significant effects on inspection ca-
pabilities
Clamp Causes a wave reflection and thus produces a
signal; no significant effects on inspection ca-
pabilities
Weld attachment Causes a wave reflection and thus produces a sig-
nal; if the attachment is large (such as pipe shoes),
can reduce inspection range
Paint Has negligible effects
Insulation Has no effects unless the insulation is bonded
to the pipe surface, in which case the inspection
range will be shortened due to higher wave atten-
uation
Coating Has negligible effects if the coating is thin (e.g.
fusion-bonded epoxy coating); thicker coating
(e.g. bituminous coating, polyethylene coating)
increases wave attenuation and shortens inspec-
tion range
Liquid in pipe No effect on torsional-wave; significant degra-
dation on longitudinal-wave
General surface corrosion Increases wave attenuation and shortens inspec-
tion range
Soil
If pipe is buried, the surrounding soil greatly
increases wave attenuation, and the inspection
range is significantly shortened
12.9. Illustration of the flat MsS guided-wave probe for structural health monitoring.
The coil is placed directly on the nickel layer.
12.11. Monitoring data showing the defect signal increasing as the simulated
corrosion increases from no corrosion to 0.2 in. by 0.2 in. (5.1 mm × 5.1 mm) by
approximately 0.05 in. (1.3 mm) deep.
6 Conclusions
The use of an FPGA-based system architecture has allowed savings in both numbers
of discrete components and the space required for the control elements in the mul-
ti-channel Teletest® long range ultrasonic system. This system has a soft processor
implemented on the FPGA which has the potential for it to be reconfigured to meet
the requirements of different ultrasonic testing applications. The performance of
the multichannel ultrasonic testing system has been improved in terms of increased
speed, reduced size and reduced power consumption. Figure 8 shows the completed
Teletest® FOCUS+ prototype system and its internal assembly of the first prototype
unit. This product has been commercially successful globally since it was firstly
released to the market in May 2011 for the improvements made by the new FPGA
embedded system design over the previous system.
Figure 8. Teletest® FOCUS + Electronic Unit and Internal Assembly of the First
Prototype Unit
Fibre-optic sensor
Featuring light weight, wide bandwidth, good compatibility and immunity to elec-
tromagnetic interference, long life and low power consumption, optical fibres are
widely used as sensing devices, particularly for capturing static or quasi-dynamic
strain. However, application of optical fibre as a Lamb wave sensor in the ultrasonic
range is limited (Staszewski et al., 1999; Culshaw et al., 1998; Levin, 2001; Pierce et
al., 1996) because of the low sampling rate of the optical spectrum analyser (OSA).
This difficulty has recently been circumvented by using a fibre Bragg grating (FBG)
filter connected with a photodetector. The major drawback of using an optical fibre
sensor in collecting Lamb wave signals is that, since the measured response is of
a three-dimensional nature, careful analysis of the output is necessary to extract
axial composition. The accuracy of measurement is often influenced by the strong
directionality of optical fibre (the Lamb wave signal captured by an optical fibre
sensor perpendicular to Lamb wave propagation can be 100 times lower in amplitude
than that collected by a sensor paralleling the direction of wave propagation).
> Read full chapter
Sensor Technologies
Alan S. Morris, Reza Langari, in Measurement and Instrumentation (Second Edi-
tion), 2016
Because ultrasonic energy is emitted at angles other than the direction that is normal
to the face of the transmitting element, a problem arises in respect of energy that
is reflected off some object in the environment around the measurement system
and back into the ultrasonic receiver. This has a longer path than the direct one
between the transmitter and receiver and can cause erroneous measurements in
some circumstances. One solution to this is to arrange for the transmission-time
counter to stop as soon as the receiver first detects the ultrasound wave. This will
usually be the wave that has traveled along the direct path, and so no measurement
error is caused as long as the rate at which ultrasound pulses are emitted is such
that the next burst is not emitted until all reflections from the previous pulse have
died down. However, in circumstances where the direct path becomes obstructed by
some obstacle, the counter will only be stopped when the reflected signal is detected
by the receiver, giving a potentially large measurement error.
Sensors
Sensors convert a physical variable into an electrical signal. Sensors are able to
detect thermal, optical, chemical, mechanical, and electromagnetic information and
translate it into electrical signals mainly through differences in electrical resistance,
capacitance, and inductance. Sensors based on changes in electrical resistance are
the most common, such as piezoresistive sensors. Piezoresistive sensors are active
sensors that change resistivity as a response to strain, and require a power supply and
an electrical current to function. The resistive change is obtained by physical changes
such as stretching, bending, pressure, deformation, and friction. In contrary, passive
piezoelectric sensors generate electrical current as a response to deformation with
no use of batteries; instead, intrinsically piezoelectric materials are used to generate
electricity from movement.
Actuators
Actuators convert electrical energy into physical manifestations. Actuators are always
active and include electroactive materials that generate different types of outputs in
response to an electrical field and current, such as sound, temperature, movement,
and light. Several electronic systems making use of power source can be classified as
actuators, such as electrical heating resistances, where heating is generated from the
conductive wires under the application of a high electric potential. Other examples
are electrical muscle stimulators and mechanical actuators used in robotics. An LED
is a common example of a light actuator. LEDs normally contain electroluminescent
layers sandwiched by two electrodes (anode and cathode), thereby generating cur-
rent. The electric current passing through the electroluminescent material causes it
to emit light (photons).
Microprocessors
The microprocessor usually performs the processing, storage, and communication
of data. A microcontroller can work in digital or analogue form. Data processing
includes the use of software and hardware. Software is the logical instructions and
operations in charge of analyzing, organizing, interpreting, processing, and com-
municating the signals with the use of specially developed algorithms. Hardware is
all the physical parts or components of a system, including memory, interfaces, and
integrated circuits.
Power supply
The power source provides the voltage needed for the electronic systems to function.
The most common type of power source is the battery. Other sources of power
include photovoltaic (PV) cells, fuel cells, thermoelectricity, and piezoelectricity. A
battery is an electrochemical device that converts chemical energy into electric-
ity. The most commonly used batteries are lead acid, alkaline, nickel–cadmium
(NiCd), nickel–metal–hydride (NiMH), lithium ion (Li-ion), and lithium polymer
(Li-polymer). Capacitive batteries store energy in an electric field. For the capacitive
batteries, density of storage is limited so they require high volume or surface.
Rechargeable batteries have a larger energy storage density, but their lifetime is
limited. Supercapacitors, on the other hand, provide an unlimited life cycle; however,
their energy storage density is low.
Instead of batteries, energy can be harvested from the environment under the form
of heat, light, and motion. Energy from motion can be achieved by piezoelectric
materials, or by electromagnetic (EM) induction. A fuel cell is an electrochemical
device that converts hydrogen and oxygen into water and, in the process, produces
electricity and heat. A thermoelectric generator is a semiconductor device that
converts heat into electricity based on the electrical power that is generated when
a temperature difference is established across the two ceramic plates. Solar energy
can be converted into electrical energy using PV cells (solar cells). A solar cell is a
semiconductor device in which solar energy of certain wavelengths can be absorbed
to generate electric power.
Interface technologies
Interface devices are used for transferring information between devices and the
wearer as well as between the wearer and the environment. Interface devices can
be used for input or output information using sensors and actuators. Devices for
input information include buttons, keyboards, voice recognition, and writing pads.
An output interface is used for presenting information to the wearer, such as vibra-
tion interfaces, audio interface, voice synthesis, and visual interface (liquid-crystal
displays LCD, LEDs, fiber optics display FOD, electrophoretic display). The LCD is
neither flexible nor lightweight and offers poor angle visibility. In contrast, elec-
trophoretic display and polymer light-emitting diodes (PLEDs) provide high contrast
and brightness, require less power, and are flexible.
Introduction
Mohammad Jamshidnezhad, in Experimental Design in Petroleum Reservoir Stud-
ies, 2015
Mineralogy and grain size: Quartz and calcite are the most common minerals in
reservoir rocks. Trace minerals are often present as individual grains or as cement.
Grain size and sorting can vary considerably; however, reservoir quality tends to
decrease with decreased grain size. Accordingly, very finely grained rocks (such as
shale) tend to have sealing properties.
Porosity: Usually petroleum rock pores are filled with connate water and hydrocar-
bons. Porosity is the ratio of pore volume to bulk volume and is usually reported as
percentage. Two porosity values are usually measured: total porosity and effective
porosity. Total porosity is the fraction of rock bulk volume that is void, whether the
individual pores are interconnected or not. Effective porosity is the ratio of connected
void space to rock bulk volume. It is the effective porosity that reservoir engineers
are interested in and, in almost all cases, the porosity measured in the laboratory is
the effective porosity.
In the sonic log, the time required for a sound wave to travel through one foot of
formation is measured. This transiting time is then correlated to the porosity.
In the formation density log, the bulk density of the reservoir rock is measured.
Using bulk density, matrix density and average density of fluids filling the formation,
porosity is evaluated.
The neutron log is sensitive to the amount of hydrogen atoms in a formation. In the
neutron log, a neutron source is employed to measure the ratio of the concentration
of hydrogen atoms in the material, to that of pure water at 75°F. This ratio (called
hydrogen index, HI) is directly related to porosity.
Permeability: The second main property of a reservoir rock, after porosity, is per-
meability. Porous medium is not sufficient for a reservoir rock—the pores must
be connected to each other. Permeability is a measure of the rock’s capability to
transport fluids. Primary work on permeability was done by Darcy in 1856. Darcy’s
law is formulated as:
(1.1a)
The unit of permeability (k) is the darcy, which is the permeability of a rock trans-
porting a fluid of 1 cp viscosity (μ) with a velocity (U) of 1 cm per second and 1
atmosphere pressure drop (P1−P2) along a rock of 1 cm length (L). Permeability of
most hydrocarbon reservoirs is much less than one darcy, so the unit of millidarcy
(0.001 darcy, abbreviated “md”) is usually applied.
For laminar gas flow through porous media, Darcy’s law is shown as:
(1.1b)
The first method is by using well testing data. In a well test, by changing the flow rate
of a well, variation in well bottom-hole pressure is recorded as a function of time.
The flow rate of a well is changed by increasing or decreasing the rate. The pressure
change is analyzed by plotting the recorded pressure and its derivative versus time.
The two most common tests are the buildup and drawdown tests. In a buildup test
the well is shut in after a period of production and then its pressure is measured.
In a drawdown test the pressure is measured in a well that is open after a period of
well shutting in.
The second method is measuring the permeability in a core laboratory. A known gas
(air or nitrogen) is injected into a core (or a plug) at controlled velocity and then the
pressure drop is measured. Using Darcy’s law (Eq. 1.1b), permeability is calculated
and it is then extrapolated to the zero value of the reciprocal pressure (1/p) to estimate
liquid (oil or water) permeability.
In the presence of more than one fluid, permeability is referred to as the effective
permeability. In this case, the ratio of effective permeability of any phase to the
absolute permeability of the rock is called the relative permeability (kr) of that phase.
Darcy’s law for multiple phase flow through the rock is formulated as:
(1.2)
Shape and size of grains, lamination, cementation, fracturing and solution are the
factors governing the magnitude of porosity [Tiab and Donaldson, 2004].