Closing in on cracks: a high-resolution digital camera provides fast, flexible imaging for photomacrography and microscopy.
I has long been recognized that digital cameras are faster than traditional film-based photography. The question among professionals in many fields has been whether the quality of digital images could be sufficient for a given application.
According to Ted Clarke, senior technical specialist in metallurgical failure analysis and tribology at Case Corp.'s Technology Center in Burr Ridge, Ill., and a noted expert in photomacrography, the answer to that question is yes. Clarke maintains that digital images are not only equal in quality to traditional images, they are also three times faster to acquire, less expensive, easier to distribute, and more useful as reference images for future analysis. "We are finding it's a big time-saver to acquire images digitally and then file the images on a local area network server," he explained.
PHOTOGRAPHY AT SMALL SCALES
A major part of Clarke's work at Case involves examining gears, shafts, and the like to determine how and why they failed. Case, which in 1996 had $5.4 billion in annual sales, is a manufacturer of farm and construction equipment.
Clarke produces detailed reports and photographic documentation of each item he analyzes. Along the way, he has developed expertise in photography at small scales, technologies known as photomicroscopy - photography through a microscope - and photomacrography, a term that refers to extreme close-ups using a single macro lens, usually at less than 50X magnification.
Three digital imaging systems have been used in Clarke's laboratory at Case. A scanning electron microscope and a low-resolution charge-coupled-device (CCD) video camera are on hand, but the dominant system used for gray-scale imaging - which comprises the greatest portion of the lab's imaging work - is a MegaPlus camera, model 1.6i/AB, from Kodak's Motion Analysis Systems Division. The camera uses a 9-by-13-mm CCD sensor to produce 10-bit digital gray-scale images measuring 1,534 by 1,024 square pixels.
MEETING QUALITY STANDARDS
In terms of quality, the basic issue is matching images on 4-by-5-inch instant film, and recording the same field size with the same resolution. "Industrial laboratories have standardized on 4-by-5-inch photomicrographs, and instant prints have largely replaced the glass plate and sheet film prints made years ago," Clarke explained. The reason, of course, is speed. The faster a laboratory technician can document a test and distribute the results, the faster a company can solve any problems.
Clarke's standard for 4-by-5-inch instant print resolution is approximately 6 lines per millimeter at a magnification of 500 times the numerical aperture. His experiments with the MegaPlus camera indicate a resolution of 5.6 lines per mm. The experiments were conducted using a Nikon 60-mm f/2.8 Micro Nikkor lens on a copy stand, imaging a version of the National Institute-of Standards and Technology's Microcopy Test Chart in a 3.5-by-4.5-inch field size (which corresponds to the image area of a standard 4-by-5 print).
Since then, he has found ways to increase the flexibility of the camera system. For larger field sizes, he mounts a Vivitar 28-mm f/2.8 close-focusing lens on the camera. The 60-mm Micro Nikkor yields magnifications on the camera's CCD sensor of 0.1X to 1.0X, which when enlarged are equivalent to 1X to 10X on 4-by-5-inch instant prints. Since no bellows mount is commercially available, he has machined a special mount from salvaged materials that allows him to add an Olympus bellows on the camera. With it, he can use a 100-mm f/6.3 Zeiss Luminar macro lens to yield 1.0X to 2.5X magnifications on the sensor; and with a 63-mm f/4.5 Zeiss Luminar macro lens he can get 2.5X to 5.0X on the sensor, for 25X to 50X magnification on a 4-by-5-inch print - as opposed to a maximum of 27X using the Polaroid.
"This gives us far more flexibility than we had in the past," Clarke said. "We can image small regions on large parts at high magnification, which we couldn't do before, when we were working with a view camera system."
Recently, the Gear Research Institute, which is operated by ASME and the American Gear Manufacturers' Association, asked Clarke to help with a study they were conducting on contour induction hardened gears. On these gears, unlike normal case-hardened gears, the hardening treatment is applied only to the outside of the gear teeth, allowing the main body of the teeth to retain flexibility. Normally it would be necessary to image the gear surfaces produced in the single-tooth bending fatigue test with an electron microscope, but the researchers were convinced that Clarke's photomacrographs told them all they needed to know, so microscopy was dispensed with.
IMAGE OUTPUT AND DISTRIBUTION
The MegaPlus camera, cabled to a Pentium PC, produces images that are immediately viewable on a 1,024-by-1,280-pixel monitor. The camera can be transferred in less than a minute from the copy stand to a nearby Zeiss Universal Microscope to capture images of even greater magnification.
The digital camera is particularly fast for producing photomacrographs. The photomacrographs at Case are typically of fracture surfaces and other relevant features, such as broken gear teeth. The metal surfaces are often highly reflective, and establishing the proper lighting and exposure to capture detail is much more difficult than in photomicroscopy. The digital system allows Clarke to quickly establish proper illumination, focus, and depth of field while viewing camera output in real time on the computer monitor. This results in time savings that average at least 75 percent. In the past, it wasn't uncommon to expose three or four sheets of instant film for every usable picture. The digital camera has paid for itself several times over, Clarke says, by eliminating waste and making the metallurgists and technicians more productive.
When a metallurgical report is complete, the marked-up images that illustrate it are loaded as bitmap files onto the Technology Center's server, where there is a two-gigabyte sector for images. From the server, they can be viewed by any Case employee with appropriate authorization. Noted Chief Engineer Gordon H. Walter, "We can send users a brief report by e-mail. Then they can pick up the images, look at them on their monitor, and download them if they need to."
In other settings, the 1.3-megabyte images might choke a local area network. But that hasn't been a problem at the Case Technology Center, where the network is designed to accommodate engineering drawings even larger than the MegaPlus camera's image files. The individual files and Microsoft Word reports with linked image files are sent internationally as well.
"Last week, we sent two of the 1.3-megabyte image files to France," recalled Walter. "That only took a couple of minutes. Until now, the only way we could move an image that fast was to fax it, but this is far superior to faxed images. With a fax, you lose image detail. But we can send images of a fractured part and show the small detail and origin of the fracture."
To print the images, the company is currently using networked laser printers with eight megabytes of memory and 600-dots-per-inch resolution. Network printers with higher resolution and gray-scale range could eventually upgrade the company's print capabilities. Clarke currently prefers viewing images on screen because the monitor images display more detail with greater contrast.
INTEGRATING IMAGES INTO THE WORK FLOW
Recently, Case has more fully integrated digital images into the reporting process. Reports with embedded digital image links are issued over the LAN so users don't have to seek out images separately on the Technology Center server. The department is also creating a reference library of microstructure images that have been captured digitally. Image Central software from Advanced Imaging Concepts in Princeton, N.J., is to be used to create the database with reference images and associated data. "The idea is to document different types of failures and keep the images in the system so people can go in and say, 'This is what a typical failure of type X looks like,' and compare it with the part they're examining," Clarke explained. Images will remain on the network over the long term for reference. "We're looking at a major improvement in the way test images are acquired and used throughout the company," Walter said.
Henry Baumgartner is a contributing editor to Mechanical Engineering.
|Printer friendly Cite/link Email Feedback|
|Date:||Jul 1, 1998|
|Previous Article:||RP industry's growing pains.|
|Next Article:||Current directions in meshing.|