Thermal Imaging from Thermal Survey Ltd
What
is thermal imaging?
Principles of infra-red radiation.
Equipment types.
Image interpretation.
Survey Techniques.
What is Thermal Imaging?
Infra-red radiation is electromagnetic radiation, as are radio waves and visible light. Electromagnetic radiation ranges from low frequency, long wavelength emissions such as the Rugby Radio Clock at 60Khz (5000 metres) to X-rays whose frequency may exceed 1020Hz, a wavelength of 3 x 10-12 metres. All these emissions share the same free space velocity of 2.99 x 108 m/s.
It is important to remember
that the thermal imager produces a picture totally made up of detected
heat with no visible light content whatsoever. Infra-red imaging, by
detecting small radiant temperature differences at high spatial resolutions,
presents an excellent way of monitoring the energy efficiency of buildings,
pipelines, plant & equipment and electrical power distribution.
These are by no means the only uses for thermal imaging equipment however,
such equipment has been used in applications as diverse as airborne
animal counting and mineshaft detection.
Principles
of Infra-red Radiation
Infra-red radiation is emitted by every object in the universe, so long as that object is above absolute zero, (-273°C).
The distribution of radiation
with wavelength, according to the temperature of the body, is shown
in the black body curves below. (A black body being a theoretical perfect
radiator with an emissivity of 1.0)
This distribution of radiant energy follows Planck's Law as shown in the figure. Max Planck being a Nobel prize winning physicist of the early 20th Century.
From these curves, we can clearly see that the frequency at which maximum radiation occurs changes depending upon the actual temperature of the object. This is of use when determining the wavelength of infra-red radiation that should be studied to provide an image of the object.
In practice, the total energy
emitted by the object is of interest. This total emission is the integral
of the curve above and is defined by the Stefan-Boltzman Law:
Although in practice, many factors influence the detected signal intensity. We can see that there are two major factors which must be taken into account when considering thermal imaging of an object.
The absolute temperature of the object which defines the wavelength at which maximum (but not all) radiation occurs, in addition to influencing the amount of total radiation.
The emissivity of the object which defines how much radiation will be emitted from the object. The emissivity of an object can also cause other complications which will be looked at later.
Infra-red lies past the red end of the visible light spectrum and for
imaging purposes can be regarded as the wavelengths covered between
1µm and 20µm. (micron = µm = 1 x 10-6 metres).
Infra-red in the 1µm region is generally used for non-imaging applications such as short range remote controls or basic intruder detection systems.
There are several areas across
the infrared wavelength spectrum in which the absorption of radiation
by the atmosphere renders these wavelengths unusable for imaging applications.
This extreme atmospheric absorption is caused mainly by carbon dioxide
and water vapour present in the atmosphere.
This leaves us with two bands of Infra-red radiation that are transmitted through the atmosphere well enough to enable imaging to take place. These are the 3 - 5 µm and 8 - 14µm bands.
By examining black body (Planck)
curves, we can see that a radiator at ambient temperature would radiate
most effectively in the 8 - 14µm band whereas a hotter object
such as a furnace would emit the greater amount of its radiation in
the 3 - 5 µm band. When considering thermal imaging equipment,
the anticipated temperature of the object under examination should be
used to give an indication of the most suitable band to use. It should
however, be remembered that most radiators will be emitting radiation
in both bands so that images may be produced in either band.
Equipment Types
In both of the imaging bands considered, there are two types of detector that may be used to convert the incoming infra-red thermal radiation to an electrical signal suitable for processing into a pictorial output; Thermal Detectors and Quantum Detectors.
Thermal detectors rely on a change in material characteristic caused by absorption of infra-red energy. The most common type of thermal detector uses the pyro-electric effect in which the temperature change of the element causes a change in the charge present on the device electrodes. These are the types of detector element used in fire and intruder detection systems. They have the advantage of not requiring cooling and are also used as part of PEV imaging systems.
More sophisticated imaging systems tend to use quantum detectors, these are semiconductor devices in which incident radiation excites excess carriers, proportional to the radiation intensity.
The most common semiconductor used as a quantum detector is Cadmium Mercury Telluride (CMT), this has the advantage that its composition can be adjusted to give maximum sensitivity at either 3 - 5 µm or 8 - 14µm.
The signal output of a quantum detector is so small that it would be swamped by noise generated internally to the device at room temperatures.
Since noise within a semiconductor is partly proportional to temperature, quantum detectors must be operated at low temperatures. CMT detectors should be operated at -80°C when operating in 3 - 5µm modes and to -193°C when operating in the 8 - 14µm band.
This cooling requirement is a significant disadvantage in the use of quantum detectors. However, their superior electronic performance still makes them the detector of choice for the bulk of thermal imaging applications.
There are several different ways of cooling the detector to the required temperature.
Bulk liquid. In the early days of thermal imaging, liquid nitrogen was poured into imagers to cool the detector. Although satisfactory, the logistical and safety implications led to developments into HPPG and thermal transfer technology.
HPPG. High Pressure Pure Gas can be used to cool a detector to the required temperatures. The Joule Thomson effect is the reduction in temperature of a gas when it rapidly expands from a high to low pressure. The gas is passed via a pipe coil to an orifice (typically <100µm in diameter), the gas rapidly expands and undergoes a rapid loss in temperature. The waste gas passing upwards past the incoming coil cools the incoming gas further until the cooler has provided a rapid cool down to design temperature. Suitable gases are Nitrogen, Oxygen, Air and Argon. In general use, Pure air is the most common gas used due to the relative simplicity and low cost of producing suitable volumes of gas. Particular care must be taken regarding the purity of the gas used in a Joule Thomson cooler. Since the orifice is so small, any particulate contamination will block the cooler, as will the formation of ice if there is any water vapour in the gas. Therefore, suitable filtration must be provided at all stages. A typical thermal imaging facility will clean and dry the air during the charging of bottles and the air will also be passed via a filter assembly when being subsequently fed to the detector.
Mechanical cooling systems are also in use. These have the logistical advantages of freeing the imager from the requirements of carrying high pressure gases or liquid nitrogen. They do however, have a number of disadvantages when compared to a Joule Thomson system such as higher noise level, electrical interference, longer reaction time, increased power requirements, additional control circuitry and they have, in the past, gained a poor reputation for reliability. Modern split cycle Stirling coolers have overcome (or reduced to acceptable levels) these disadvantages and are now coming into widespread use in commercially available imagers.
There are a number of different ways in which imagers operate, these
can be roughly classified into PEV, Staring Array and Scanning systems.
A Pyro-Electric Vidicon (PEV) is a variation of a conventional vidicon camera tube. A plate of pyroelectric material is placed at the front of a vidicon tube, this effectively forms a variable capacitance with its characteristics varying according to the incident infra-red radiation. The plate matrix is scanned by an electron beam and the resultant impulses amplified and processed into a video signal. These are uncooled devices and although in service for basic applications are no substitute for cooled quantum detector based systems. Since the pyroelectric effect depends upon a change in incident radiation, they have to be constantly moved to produce an output. In practice, this is achieved by using a mechanical optical 'chopper' to interrupt the thermal radiation scene. They have limited spatial resolution due to thermal spreading (conduction) within the cell matrix on the front plate of the vidicon.
Staring Arrays, as the name implies, consist of a matrix of detector elements. These elements are often manufactured from Cadmium Mercury Telluride or Platinum Sillicide. The entire scene is focused on this array, each element cell then provides an output dependent upon the infra-red radiation falling upon it. These types of imagers have the advantages of not requiring delicate thermionic devices (such as the vidicon) or sophisticated scanning optics. However, at the moment, although a number of commercial imagers do use this technology, there are practical limitations in producing an array with a large enough number of elements to match the resolution achieved by scanning systems. This is an area of imaging where significant development is currently taking place, particularly for midrange commercial applications.
The bulk of high resolution (military grade) thermal imagers use scanned optical techniques. They use a cooled CMT detector which is scanned across the image in a number of formats.
In the simplest form, a single element could be scanned along each line in the frame (serial scanning). In practice, this would require impossibly high scan speeds so a series of elements may be used. These may be scanned as a block, along each line. This cuts down the scan speed from having just a single detector but the scan speed and channel bandwidth requirements are still high. It does however, give a good degree of uniformity. The frame movement can be provided by frame scanning optics or in the case of linescan type imagers, by the movement of the imager itself. This type of imager is often used in aerial applications where the detector element(s) are scanned along the same line, whilst the forward movement of the aircraft provides the relative frame movement. These imagers often provide a digital or photographic output rather than a CCIR video signal. Problems of non-linearity may be introduced by lateral movement of the aircraft.
Another method is to use a number of elements scanning in parallel (parallel scanning). These have one element per line but scan several lines simultaneously, this can give rise to poor uniformity. However, frame scan speeds are lower.
A frequently used compromise is to use a serial/parallel matrix. This provides acceptable uniformity in conjunction with realisable bandwidths and scanning speeds.
Each of the above methods has its advantages and disadvantages. They are all in use in modern thermal imagers.
Another type of CMT based detector is the SPRITE. This again is a cooled detector which requires scanning optics. SPRITE (Signal Processing In The Element) takes the place of several serial elements. The processing that would have been done external to those elements now takes place due to semiconductor biasing & doping within the SPRITE element itself.. This has the advantage of reduced encapsulation lead-outs, less signal processing circuitry and an improved signal to noise ratio. Several SPRITEs may be used in a parallel scan to further improve efficiency. Most of the modern top quality imagers now available, including some of those used by Thermal Survey Ltd, use SPRITE technology. Multi-element SPRITE systems may still be regarded as restricted military technology.
When considering the optical requirements for thermal imagers, it is important to consider the optical material used. At infra-red imaging wavelengths, glass becomes a complex radiator and cannot be used to transmit radiation. There are many materials with suitable infra-red properties but these are often of restricted use due to physical limitations.
Germanium has become the most popular material as it is now readily available in large sizes with good optical characteristics. A wide range of protective coatings exist and it is in almost universal use for standard imaging applications. Its only major drawback is that it becomes opaque above 100°C, making it unsuitable for high speed aerial applications.
Zinc sulphide, zinc selenide,
sapphire and magnesium fluoride are also used in certain applications.
Image Interpretation
Most thermal imagers produce a video output in which white indicates areas of maximum radiated energy whilst black indicates areas of lower radiation. Most cameras have the facility to invert this video so that black relates to maximum radiation and vice versa.
This video output is recorded onto high quality, broadcast standard video tape on site. The resultant tape can then be analysed by Thermal Survey's image processing computer systems. The image is also available for viewing whilst filming is taking place. In this way, a Client Engineer can often plan remedial action at the scene.
The original black/white signal contains the maximum amount of information, certainly more than the eye can distinguish. However, in order to ease general interpretation and facilitate subsequent presentation, the thermal image can be artificially colourised. This is achieved by allocating desired colours to blocks of grey levels to produce the familiar colourised images. This enables easier image interpretation to the untrained observer. Additionally, by choice of the correct colourisation palette, the image may be enhanced to show particular energy levels in detail. For example, the operator can choose a palette to highlight cryogenic temperatures or by selecting another palette, objects at high temperatures.
As mentioned above, the amount of infra-red radiation emitted from a surface depends partly upon the emissivity of that surface. For this reason, extreme care is needed if using an infra-red imager to give accurate temperature measurements within an image. By far and away, the main benefit of thermal imaging is obtained from qualitative rather than quantitative use. Infra-red non-contact thermometers do exist but they all require accurate assessment of surface emissivities if the result is to be meaningful.
When interpreting infra-red images, remember that the image is comprised purely of radiated thermal energy. The effects of the sun, shadows, moisture and subsurface detail must all be taken into account as described below.
Often with infra-red building surveys, the item looked for, or the problem to be diagnosed, is not immediately apparent. Bear in mind that the imager is looking at the radiation emitted from the surface. The imager does not have the ability to see below the surface as such; however, the radiation from the surface is often influenced by subsurface detail such as buried conduit, cracks, wall ties etc. which all effect the thermal characteristics of the adjoining material. When conducting aerial surveys, sub-surface detail becomes even more apparent with buried pipework (hot or cold) being clearly visible because of their effect on the surface temperature and emissivity.
In the same way, air flow can often be detected by its cooling or heating effects as it enters or leaves the building structure. Moisture can often be seen as a result of cooling of the surface material by evaporation. If a wall is subject to dampness, the resulting image will show an uneven response due to the varying degrees of evaporation. It is sometimes possible to follow the path of water ingress through the building structure in this way. This does however, mean that surveys should not be carried out in the rain or whilst the building structure is wet as misleading results will result.
The following factors should also be borne in mind:
The effect of solar gain on the thermal structure of a building can lead to confusion. In general, infra-red surveys are carried out sometime after sunset so that all such effects have dissipated from the structure. However, this is not always possible and the position of the sun relative to the building should be considered. In this case, shadows falling on the building or shadows that have been on the building, can also have an appreciable effect on the thermal radiation viewed.
When looking at a large area, the emissivity of various surfaces must be considered. Most materials found on the surface of buildings will have a relatively high emissivity (~ 0.95) but there will still be noticeable differences in the perceived image due to a change in surface material. This can be overcome by a detailed knowledge of the building under investigation. When imaging surfaces such as metal or glass, special care must be taken. Polished metal surfaces tend to reflect infra-red radiation in the same way that they do visible light. Hence, an apparent 'hot-spot' may be a reflection of a hot object some distance away from the area under investigation. Such anomalies can be detected by moving the imager around so that the reflective angles change.
Glass is predominantly opaque to infra-red radiation (particularly so at 8 - 14µm) and in most cases, the image will be dominated by reflection. Hence, in ground floor windows, a reflected image of the survey team will often be noted and in upper floor windows the reflection of the cold sky temperature will be apparent. Glass is a selective radiator with an emissivity which fluctuates markedly with wavelength. These examples serve to emphasise that the radiation properties of the target materials being surveyed need to be known. Are the surfaces blackbody, grey body or selective radiators?
Survey Techniques
The primary considerations for all survey activity are the environmental conditions.
If looking at buildings, as per the majority of ground surveys, a temperature differential must exist between the inside and outside of the building. In this case, in the event of an insulation defect, warmth will be seen leaving the building structure if viewed from the outside. If the survey is being conducted from the inside, conduction from the external cold air in the vicinity of a defect will be noticed.
To achieve this differential, surveys are most often conducted in the winter months when the outside air temperature is at a minimum and the buildings are heated. The exceptions to this are refrigerated buildings which may be surveyed during summertime to achieve maximum differential.
As mentioned previously, areas of dampness will give an uneven thermal response. This may be confused with defective areas of insulation so care should be taken to avoid surveying when walls may be damp. (Unless of course, the object of the survey is to identify areas of dampness).
Wet ground, snow or frost will give rise to misleading survey images so care must be taken if conducting surveys during such periods.
Very much the same conditions apply to aerial surveying. Additionally, the cloudbase must not extend below the survey height since the water vapour in the clouds makes them opaque to infra-red. Care must be taken to ensure that survey flights are not made in excessively windy conditions. If the wind is too high, the effects of wind-chill will be seen around the edges of buildings and the image quality may be poor if the aircraft has difficulty remaining on a stable track heading. The exact threshold speed will depend upon aircraft type and the nature of the images required, but would be generally around 15 knots.
For aerial surveying, the imager may be mounted in a camera hatch of any suitable modified observation aircraft. A typical Thermal Survey survey would be flown at 610 metres altitude above ground level. This gives a field of view swathe 420 metres wide. The survey area is then divided up into parallel tracks 300 metres apart. This gives a degree of overlap to allow for wind gusts, aircraft roll or positional error.
The resultant video output from the camera is fed to a time/date generator which superimposes a time/date stamp on the video signal. This can then be used to cross reference the images with the tracks plotted on a map.
Used correctly, infra-red thermal imaging is a valuable tool for evaluating the conditions of buildings, plant & machinery. They are of use in diagnostic, quality control and work prioritisation roles to name but a few.
Phone: 01704 560072
Fax: 01704 562700
(If dialling from outside the UK, use +44 1704xxx)
E-mail
