Jim Clark 0000-00-00 00:00:00
<b>Computed Tomography Goes Mainstream</b> Learn what you need to know about four dimensional coordinate metrology. By now most of us in the metrology world have seen beautifully rendered images from computed tomography (CT) data. We see images of bones, fossils, dinosaurs, teeth and of archaeological treasures, and hear about various universities that use CT systems. But perhaps we are left wondering, is this a viable metrology tool in terms of capabilities and cost, or is it just an experimental toy? Don't be fooled. CT's accuracy, ability to measure internal and external dimensions simultaneously without destroying the part, and the insight it provides through the additional fourth dimension of material density, are rapidly making it the must-have tool of quality. <b>ACQUIRING A COMPLETE VIEW ON INDUSTRIAL SAMPLES</b> Like any new product that has spent its formative years in the hands of scientists and engineers, there is a lot of confusion about how it all works. After all, it involves areas of physics, mathematics and computing not encountered on a day-to-day basis. This serves as an introduction to CT and to some of the techniques, concepts and components of the system that one will need to know to use a CT service or purchase a CT system. CT is fundamentally a very simple process. Place an object on a rotation stage and mount it between an X-ray source and a detector. The detector acquires simple 2-D radiographic images (think visiting the doctor's office with a broken bone) of the object as it rotates. After the object has rotated through 360 degrees, take the 2-D X-ray images, send them through a piece of software called a reconstruction algorithm and then, after a short period of processing and calculating, the result is available. The result is a 3-D volumetric map of the object, where each element is a 3-D cube (voxel) which has a discrete location (x,y,z) and a density ( ). Not only is the external surface information known, such as with a 3-D point cloud from laser scanning, but internal surfaces and additional information about what is in between the surfaces from the fourth dimension density is provided. <b>A FEW SIMPLE RULES OF THE ROAD </b> CT works very well if four simple rules are followed: 1. Enough X-ray power or flux to penetrate and get through the object is needed. The object cannot absorb or stop all the X-rays; some X-rays have to get through to the detector. 2. As the object rotates it has to stay inside the field of view cone created by the X-ray source and the detector. 3. The smaller the X-ray spot size, the better the resolution. 4. The closer one can get the object to the source (and maintain rule 2), the better resolution and accuracy obtained. What happens if the rules are not followed? Well sometimes it is not disastrous; it will depend on the materials and geometry. The results might not be optimal and there might be a few artifacts in the data, or the resolution might not be what was wanted, but the data might still give what is needed. CTKEY ELEMENTS </b> The X-ray tube is at the core of a CT system. The industrial X-ray sources do not use any radioactive material. Several different tube designs exist, but essentially an X-ray source consists of a tube or cylinder in which there is a filament (think light bulb) at one end, along with a high voltage cathode and anode, a magnetic lens and a metal target (normally tungsten). Since the X-ray tube and target combination directly impact three of the CT rules, it is very worthwhile taking time to think about the design of the X-ray source. Take note: if buying a system, ask whether the X-ray source is an open or closed tube source. Open tube sources, which allow one to replace the filament on a regular basis, typically have a much lower cost-of-ownership than closed tube sources, which have to be replaced when they fail, normally at great cost. For a typical source in reflection mode a current is applied to the filament, which causes it to heat up and emit a stream of electrons. The electrons are repelled by the cathode and attracted to the anode by the high voltage field. This field accelerates the electrons up to 80% of the speed of light toward the end of the tube. Before they leave the tube, the electron beam is focused onto a target material, typically tungsten, using an electro-magnetic lens. The electrons slam into the target and the sudden deceleration of the electrons as they interact with the target pretty much just heat it up. Indeed that is what happens to 99.3% of the energy in the electron beam; it goes into heating up the target. The remaining 0.7% produces X-rays that are generated in a cone beam from the target. These X-rays emanate from the region where the electron beam hits the target. This spot, or more correctly the size of this spot, is referred to as the X-ray spot size. In general, the higher the voltage applied, the more power is in the beam, and consequently more power is transferred to the target. The more power on the target, the larger the X-ray spot size, and the more X-ray power produced. <b>REFLECTION OR TRANSMISSION MODE?</b> Reflection mode is the most common mode because it can be cooled more effectively, allowing more power in the beam to be delivered before the target overheats and damages, thus allowing More X-ray flux to be generated. In a reflection target arrangement, the closest that one can get the object to the X-ray source (remember rule 4) is normally around 5.5 to 7.5 millimeters. While this might seem close for very small objects, it means one is giving up valuable pixels on the detector, or magnification. To increase the magnification, a transmission target can be used instead. In this mode, the electron beam is focused into and through a metal target that is mounted on the vacuum tube window, causing the X-rays to be transmitted from the material. While such a target cannot handle as high a power as a reflection target, as the cooling is much less efficient (typically 5-10 W for best resolution), it has the benefit of a smaller spot size (rule 3) and higher magnification, as one can get within 0.5 millimeter of the target (rule 4). There is often a balancing act between X-ray power and spot size. Another form of target, called a rotating target, helps with this balancing act. In the standard reflection target, the problem of limited power and spot size stems from the heating of the target. The target just cannot take the high levels of power, and quite literally holes are punched into the target. With a rotating target (imagine a spinning CD) the electron beam falls on a moving surface so the power in the electron beam can be increased considerably without incurring damaging heating effects. The power in the electron beam can be doubled without significantly degrading the spot size (rule 3), while the X-ray intensity can increase three to five times, allowing for either denser or larger objects to be measured (rule 1), or for objects to be measured quicker (detector integration times reduced). In any case, the information recorded is due to the attenuation of the sample. Denser samples will attenuate the X-rays more. So far, the X-ray sources mentioned have been microfocus X-ray sources. Be aware that most system suppliers only offer microfocus sources up to 225kV, while more powerful sources in their offerings are minifocus. Minifocus sources produce more X-ray flux, which is great (good for rule 1), but the spot size of these minifocus sources is orders of magnitude bigger compared to microfocus sources (bad for rule 3), greatly reducing the accuracy of the data. There are only a few vendors offering microfocus sources up to 320kV or even 450kV. Don't be mistaken: a microfocus source is needed to acquire accurate and detailed CT data for most high-accuracy industrial CT applications, and a supplier's spec sheet might not distinguish between the two types of sources. <b>SEEING X-RAYS </b> The imaging devices used to view and measure the transmitted X-ray radiation can greatly influence X-ray image quality and CT results. There are three principal types of detectors used: onedimensional linear detectors, image intensifiers attached to CCD cameras and amorphous silicon panels. One-dimensional linear detectors were used at the inception of industrial CT. These devices work by having a line of small scintallators which generate light when bombarded with X-rays. The scintallators are backed by a row of photodiodes that capture this light. Such a device only captures one slice at a time as opposed to an entire twodimensional image. One-dimensional detectors are still useful when measuring materials where there is a lot of scattering (think very dense materials or single crystal turbine blades). Recently, curved onedimensional linear diode array detectors have been introduced as a means of removing path length difference errors found in linear systems. Also, such a curved array means that each scintallator is radial, and so the length can be increased to improve the detection efficiency. With the advent of cone beam CT a number of years ago, vendors incorporated another detector type combining an image intensifier and a CCD camera. This was a great leap forward in data capture timing and for real-time 2-D radiography. A disadvantage of the image intensifier/camera combination is that image quality is somewhat lacking. Today, most high-end industrial CT is done using amorphous silicon panels. The size of the panel is very important because it can determine the size of the object that can be measured with CT. As the object is rotating in between the source and the panel, the maximum swept object width is only 65% to 75% of the panel width. For example, using a commercially available 16 inch by 16 inch panel, the maximum object width is around 12 inches. Some suppliers go beyond this object size limit by applying panel shifting. For example, by taking multiple scans using different panel positions, some customers are able to scan parts with widths up to 39 inches. <b>HOW ACCURATE IS CT?</b> How accurate is CT? Like optical systems that typically suffer from not having their accuracy readily defined by national standards agencies, it is often a complex issue that does not easily translate into standard coordinate measuring machine (CMM) or 2-D vision terms or techniques. X-ray and CT are no different. The size of the object, the materials used, the geometry, the detector used and the spot size can all come into play. The good news is that there is one simple, general rule to calculate accuracy. Essentially, the accuracy of an industrial microfocus CT system is approximately . Of the voxel size. The voxel size is determined by the diameter of the object being measured, divided by the number of pixels on the Detector which is being used to measure the diameter. The full calculation also involves the X-ray spot size, but for most industrial objects, this is a small correction < 1%. Recall rule 4, where there was concern about how close one could get the object to the source. The closer one gets the object to the source, the more detector pixels that are used to measure the object. The more pixels used, the smaller the voxel size, and the higher accuracy and spatial resolution. <b>THE FUTURE</b> X-ray and high-accuracy micro CT technology has evolved over the past 10 years to the extent that it is a mainstream metrology technology. It has accuracy and resolution, speed, flexibility and unimaginable detail, and it provides fourth dimension insight through density mapping. And importantly, the price point is now low enough to make it competitive with other techniques. Applications are diverse and growing constantly across the automotive, aerospace, energy, medical and consumer sectors, dealing with plastics, metals and exotic materials. Using a CT volume helps manufacturers trace both visible and invisible material and geometry-related flaws that in most cases cannot be identified using other nondestructive testing methods. Software tools enable the analysis of the volume against the CAD model, either via direct volume to CAD comparison, or through geometric and dimensioning tolerance measurements. Jim Clark is vice president strategic marketing and business development X-ray products at Nikon Metrology (Brighton, MI). For more information, call (810) 220-4360, e-mail firstname.lastname@example.org or visit www.nikonmetrology.com. <b>QUALITY ONLINE</b> For more information on computed tomography, visit www.qualitymag.com to read the following articles: "Faster Computed Tomography" "Measure Completely and Accurately" "What You Can't See Can Hurt You"
Published by QualityMagazine. View All Articles.
This page can be found at http://digital.bnpmedia.com/article/Quality+Test+%26+Inspection/515675/48771/article.html.