Nondestructive testing (NDT) is nothing new. It's used routinely in the aerospace industry, and its use is growing in automotive and other industries. Getting maximum benefit from NDT, however, is going to take an attitude adjustment.
The first step, perhaps, is to understand that NDT is not just eddy current testing, or magnetic particle testing, or X-ray testing. It never ceases to amaze how many designers and engineers out there equate NDT with eddy current or magnetic particle testing, and nothing else. In fact, NDT encompasses a number of other technologies including dye penetrants and ultrasonics. Each of these technologies is designed to ferret out surface and/or subsurface discontinuities.
The fact that there are a number of different NDT methodologies available leads to another point: not all NDT technologies are equally good at finding a specific defect. For example, it's not unusual for us to get a request to use X-ray inspection to locate surface cracks. While it's certainly possible to do so, other factors, such as part geometry, may work against X-ray inspection. If the surface crack is not in a plane perpendicular to the X-ray beam, it may not appear on the film. Eddy-current or magnetic-particle inspection would likely provide better results.
Designers need to consider up front what defects are likely to occur during the manufacturing process, and what test and inspection methods are best suited to detect them. Then they must incorporate that information into the specifications.
Most manufacturers are quick to accept the need for gages to confirm the part geometry specified on a print as part of the manufacturing process. That's not the case when it comes to incorporating methods for validating structural integrity, however. Why? We believe it's because NDT is viewed as expensive, and not a necessary cost of doing business. That's an attitude that involves serious risk.
This perception of NDT derives in part from how it is typically employed. Frequently, one, two, or maybe 5% of the pieces in a production run are sent to an NDT laboratory for testing. Because the cost to setup and run the required tests is distributed over a relatively small lot size, the price per piece can be high, depending on the tests. And, while statistically you're likely to have discovered any defects by testing a representative sampling of parts, you can't be certain you've caught them all.
Just as often, NDT inspections are done on an emergency basis. A defect is discovered in product that has already been shipped. Now the remaining product must be retrieved and inspected in panic mode. The cost of transportation, and the cost of running the tests on an emergency basis, plus any compensation due the customer can be very expensive. So expensive, in fact, that the cost of one such episode could more than pay for 100% part NDT inspection for a year. In the extreme, we're familiar with three instances where the cost was enough to put the manufacturer out of business.
It doesn't have to be that way. Take a look at the aerospace industry, for example. Their attitude is that the potential liability resulting from the failure of a single defective part more than justifies the cost of programs calling for 100% inspection of critical components using NDT. In effect, it's cheap insurance.
There's another related misconception about NDT that says it's not suitable for high-volume operations. That's simply not the case. While it's true that many NDT laboratories are not set up to do high-volume work there are others who can keep pace. We've handled production runs as high as three-million pieces.
In high-volume applications, NDT leads to choices and trade-offs. Is it more cost-effective to do the testing in-house or outside? Establishing an in-house capability may require a substantial capital investment, plus staffing and training costs--but offers total control. Outsourcing the work eliminates tying up capital, but also involves relinquishing some control.
A third option offers installation and maintenance of NDT equipment at the manufacturer's facility plus operator training, in return for a per-piece processing fee. You avoid the capital expense of purchasing the equipment, yet retain control of the inspection process.
This article was first published in the December 2004 edition of Manufacturing Engineeering magazine.