· Table of Contents
Infrared System Specifications - What does it all mean?Colin Hockings
NDT Supervisor Qantas Airways
The selection and purchase of many NDT instruments requires some consideration of the instrument's capabilities and suitability for the proposed applications. Many such instruments require a significant capital investment, which must be substantiated. Management, finance and purchasing departments rely on the NDT person to provide quality information about the selection of equipment and an understanding of the technical details. Thermographic imaging equipment is no different. It could be argued that the novelty and complexity of these instruments makes selection decisions more involved than for other more familiar NDT items. This paper attempts to present the main technical specifications of typical thermographic imaging systems and convert these into real performance considerations for use in the field.
Having decided that a thermographic (infrared) inspection will provide the kind of information which will satisfy an inspection need, the next decision is to select a thermographic camera. The technical specifications are lengthy and full of abbreviations and jargon. A full comprehension of the meanings and implications of the specifications is essential to making a correct equipment selection. What follows here is not an exhaustive report of all thermographic system specifications. The prospective purchaser may well need to consult material such as reference texts and the literature as part of a thorough thermographic system evaluation. Furthermore, this presentation assumes that the reader has a basic understanding of the thermographic inspection method.
When considering system sensitivity in the application of IR imaging, the first question is how small an object needs to have its temperature measured? Secondly, how small an object needs to be discriminated in the image? Then add a third question, at what distance? The following descriptions may enable the reader to temper the salesman's enthusiasm with cogent argument on the technical merit of his equipment.
Long wave (6-14 um) or short wave (3-6 um)? It is frequently suggested that short wave systems are more susceptible to non-relevant thermal interference when used in outside situations. This assertion is based on the fact that the sun's emissions are more influential on the shorter wavelength cameras. Long wave cameras are not free from the sun's effects but are considered to be generally less susceptible to this interference. All thermographically imaged scenes are a combination of a variety of radiation frequencies. This distribution depends on the temperature range across the scene, the material of the emitters and their surface condition. No one factor such as the presence of the sun should be considered in the absence of other influences when deciding on long or short wave equipment. Long wavelength detectors are more suited to lower temperatures (< 80°C), but short wavelength infrared radiation is more energetic and hence a smaller amount of radiation will be more detectable. Many false images caused by the influence of the sun or other heat sources can be analysed by changing the angle or position of the camera and noting the behaviour of the image. This is very much an experience based skill and would be expected to develop quickly in a Level 2 thermographer.
|Fig 1: Typical transmission of 300m ground level atmosphere.|
The inspection distance, the medium through which the inspection takes place, its components and their percentage, and the ambient temperature, are all factors which need to be considered when deciding if a short or long wave camera is right for the application. Water is a very good absorber of infrared radiation. This suggests two considerations for the prospective purchaser. Firstly, that thermographic examination of objects under water is not effective and secondly, because the atmosphere contains significant amounts of water vapour, accurate temperature measurements can only be determined using a system which can compensate for humidity and distance. When considering thermographic testing over long distances, it should be noted that other common atmospheric gases such as CO2 can also absorb specific wavelengths. For airborne systems such as search and rescue, the wrong choice could have life-threatening implications.
Surface condition is significant because it is only from the surface of an object that the IR radiation is emitted. The surface condition largely determines how emissive an object will be. The ideal emitter (radiator) being one that emits 100% of the radiation.
Emissivity is a comparative figure, a ratio, of how good a surface is as an emitter of infrared radiation, compared to that of an ideal radiator (known as a black body). Emissivity correction is essential for any thermographic imaging system required to provide accurate temperature measurement. For example, if an object is at a higher temperature than background and has a low emissivity, it will be detected as being at a lower temperature than it really is. Emissivity correction will compensate for this measurement error. Emissivity variations can however be used to advantage when distinguishing between surfaces at the same temperature but having different emissivity characteristics.
|Fig 2: Effects of emissivity on imaging heat patterns.|
For low emissivity materials such as polished aluminium, the reflected radiation may well dominate and obscure the true heat state of the object. The cold sky for example reflected from an object may be misinterpreted as the true temperature of the object. To overcome such problems it may be possible to apply a benign coating to increase the emissivity and reduce reflections. Water soluble matt black paint such as used as non toxic children's paints is very effective. Some other alternatives commonly found in NDT shops is "stripable" background lacquer used for colour contrast magnetic particle inspection and spray pack developer used for dye penetrant work. Both products will increase the emissivity however a cost comparison with other products should be made if large quantities are needed.
The IFOV figure for a thermographic camera may be the most simplistic way to specify the spatial resolution of the camera. It can be expressed as the angle given by the ratio of the projected size of a pixel on the smallest detectable target divided by the distance to the target and usually described as parts of a degree or as milliradians. Generally, the smaller the IFOV figure, then the better the camera for a given total field of view. It is clear that the distance factor in this relationship becomes very important when determining if a camera will resolve the image of a small object at a relatively large distance. But this is resolution only and not accuracy. The IOFV capabilities of a thermographic camera may well find a small hot or cold spot but not necessarily measure its temperature accurately.
|Fig 3: Instantaneous Field of View (IOFW) for a square detector where WD is the dimension of a single detector element and FL is the effective focal length of the system optics. The IOFW represents an angle normally expressed in milliradians, and AP is the area projected.|
Measurement Field of View is sometimes referred to as IFOVmeas. It quantifies the resolution of the IR camera in temperature measurement terms. It is the angle given by the ratio of the target measurement spot size divided by the distance. Like the IFOV, it is also expressed in degrees or milliradians. As would be expected, the MFOV will be larger than the IFOV because the camera will need more imaging data to accurately measure temperature.
The spot size ratio (SSR) is number which expresses the maximum distance the camera can be from a target of a given size and still maintain temperature measurement accuracy. A SSR of 200:1 indicates that at a distance of 200 metres the camera will accurately measure the temperature of an object of one square metre in size. The SSR figure is the inverse of the MFOV and is a useful number to remember in the field, but not always quoted in the manufacturer's specifications.
High sensitivity systems can discriminate very small changes in the level of heat energy. This may be more important in a specific application than accuracy in temperature measurement. The sensitivity of an IR system is usually expressed in temperature terms, such as ± 0.5°C.
NETD takes account of the system's electronic noise, which for a given signal will vary depending on the temperature of the object. The NETD figure is derived from the temperature difference, which would be attributed to a signal equal to the electronic noise at a particular temperature. ASTM E 1543 - 94 describes such a test. As an example, if a 1°C temperature difference is equal to a 100uV signal and the internal electronic noise is 10uV, the NETD is 0.1°C. Valid only at the temperature measured. Another way of expressing it is that it is the temperature difference which produces a change in signal equal to the system RMS noise. It may be useful to request the NETD at each end of the working range of the IR camera as well as the temperatures expected to be encountered during inspections. Usually NETD becomes smaller as temperature rises. When the subject of interest is very small then the spatial resolution of the system must also be considered.
|Fig 4: Representation of NETD of 0.1°C.|
The MDT of an IR camera is exactly that. It is determined by assessing two large objects of reducing temperature difference until no distinction can be made. ASTM E 1311 - 89 (93) describes such a test. It should be noted that the MDT is only valid at the specified assessment temperature. MDT may be lower than the NETD because when viewing the image, the human brain should be capable of integrating visually the system noise to what is effectively a lower level of noise. An example of this ability is the way a television image can be presented as relatively coarse pixels but the brain will identify the objects in the image correctly.
It may not be necessary to measure to small fractions of a degree, but better sensitivity in a system will result in a clearer distinction between objects and hence aid in the interpretation of images.
Thermal resolution can be expressed as the Slit Response Function (SRF) in terms of a subtended angle or the number of resolvable elements in a longitudinal line. An assessment is described in ASTM E-1213 - 92, Standard Test Method for minimum Resolvable Temperature Difference for Thermal Imaging Systems. The SRF may be different in the vertical and horizontal planes. The % SRF indicates the size of a slit when the temperature shown is that % of true. A 50% SRF will not be acceptable if accurate temperature measurements are to be made. 90% or more would be needed. Systems may be similar at 50% by quite different at 90%, so it is useful to know the temperature at which the SRF is quoted when comparing systems.
If temperature measurement is not as critical as imaging, then the Minimum Resolvable Temperature Difference (MRTD) may be more useful. As with MDT, to assess this parameter the IR imager views small slots of equal width whose temperature is gradually changed to that of the background. The temperature difference at which the slots are no longer discernible is the MRTD. As the slot width increases the system resolution will approach the MTD figure.
Accuracy in temperature measurement is important in many applications but when quoted in specifications it is usually determined by imaging a relatively large object. It is presented as a temperature range of, say ± 2°C or a % of the temperature measured, or a combination of both. For system comparison it is essential to know the accuracy of the figures, the temperature scale and the object distance at which the accuracy is quoted. Temperature measurement may also require emissivity consideration and under some conditions will produce intolerable errors. Long wave band systems are more susceptible to errors due to inaccurate emissivity values.
Zoom is a useful feature in any camera but it may not improve the performance. optical zoom reduces the field of view with an apparent reduction in distance to the object and hence better resolution. Electronic zoom magnifies the existing electronic image and all its noise. The result of electronic zoom is that the individual pixels appear larger and no increase in resolution is obtained.
one useful application of electronic zoom is to use it to ensure good focus. Some viewfinders produce an image, which is quite grainy to the eye. Focusing the optics while in zoom mode makes focusing easier. Then the operator returns to the normal view assured that the image clarity is optimised.
Reducing the sensing element size in the detection array does not ensure better resolution, because the output signal from a smaller element is decreased accordingly. Such reduction will affect accuracy and sensitivity. An increase in exposure time may be necessary to compensate for a reduced detector size. Therefore the system will be slower to image the scene. For static analysis this is not important but where movement of the camera or the object is unavoidable, or where the change in heat state is rapid, the inspection may be ineffective.
Fig 5: Electronic zoom showing original, times 2 and times 4.
Fig 6: optical zoom showing original, times 2 and times 4.
Some IR cameras can be used with a variety of lenses for telescopic and microscopic applications. It is essential that the camera "recognises" the lens which is fitted, either automatically or by operator input to ensure temperature measurement accuracy. The standard lens supplied with the camera may not allow imaging closer than about 600mm (24 inches). If confined spaces or long viewing distances are encounted, additional lenses will be needed.
With a 16° lens for example, an IFOV is quoted as 1.2 milliradians. If a 4° lens is fitted instead, the optical magnification is four times that of the 16° lens. IFOV has been effectively improved to 0.3 milliradians. Such improved sensitivity levels may be required for inspections from long distances.
Filters are usually an interchangeable part of the optics of the IR camera. Special application cameras may incorporate permanent filtration. Filters are transparent to specific wavelengths. To understand what filter will work with the camera it is necessary to understand the spectral response of the camera (not just the detector), and the IR emissions from the object of interest. Filters selectively pass certain wavelengths of IR radiation to the detector in the camera. The desired effect of a filter is to restrict the unwanted wavelengths which will obscure the required information, and to pass only the desired wavelengths. Some applications may require custom made filters.
After identifying the IR wavelengths which are to be detected, ensure that these are in the camera's response range. It should be noted that filters will reduce sensitivity and can significantly affect accuracy. Also ensure that the camera is designed to be used with the filter chosen. Some typical applications requiring filters would be measuring the temperature of glass or measuring temperatures through flames.
The frame rate is the number of times the scene is viewed to produce a frame, i.e. several fields are integrated to produce a frame. If the frame rate = the field rate then no integration is necessary. When comparing IR systems, frame rate is more significant. Slow frame rates will produce "streaky" images even if the camera is moved slowly. Another term which appears similar but describes a different function is the Frame Repetition Rate. This is the time taken for the infrared imager to scan and update every picture element in the detector, expressed as frames per second.
Eyepiece or flat screen. An eyepiece is usually a better option in the field because it is unaffected by stray ambient light which can significantly reduce contrast. However a flat screen can be acceptable if shielded from reflections and is more easily viewed by a second person during the inspection. The additional consideration for the eyepiece viewer is that when concentrating on viewing into the eyepiece, the operator is less likely to be aware of changes to his surroundings. This may have safety implications for the operator.
Focal plane array (FPA) detectors are quite non-uniform in their response to incident radiation. To correct for this, most cameras have a non-uniformity correction (NUC) capability. To maintain accuracy in temperature measurements NUC should be done after changing temperature ranges, if the camera temperature changes, or if the lens is changed. Cameras with automatic NUC capability will of course be easier to use in the field.
Doubtless, the complexity of specifications can be an obstacle to understanding IR systems. However this should not be a reason for making a wrong choice. Any time spent in researching and consulting with existing users will be well spent. IR cameras are very clever devices, have great potential to pay for themselves many times over when applied to condition monitoring of expensive plant, and are becoming easier to use and afford. over recent years they have earned a place in the NDT "toolbox" along with the more established methods.
|© AINDT , created by NDT.net|||Home| |Top||