| by John
DeGaspari, Associate Editor |
Productivity
earns the bread and butter in the manufacturing business, so plant managers
have to push as much product as they can out their doors in the shortest
time possible. But goods also have to meet quality standards. After all,
customers are more likely to remember bad experiences than good ones,
so products that are not up to snuff can taint a company's reputation.
And product recalls and make-goods on warranties are expensive.
Among the more sophisticated methods available to manufacturers for keeping
their processes in check and minimizing bad products are tests using such
phenomena as ultrasound, eddy currents, or X-rays. They can test parts
for interior cracks and flaws without breaking anything.
 |
| Strides in ultrasonic testing
may eventually automate checking spot welds on auto assembly lines.
Here, welding robots assemble body-in-white parts at a Ford plant. |
Collectively, test methods of this sort are known as non-destructive
evaluation, which has long been a standard practice in the power generation,
nuclear, and aerospace industries. Automotive suppliers routinely use
non-destructive methods to test all their safety-critical parts. Some
observers say that non-destructive testing is likely to become more important
to the auto industry as it moves to lighter-weight materials.
Non-destructive evaluation is the subject of a good deal of research,
which may soon make it easier than ever to integrate testing into assembly
lines, and so may take the technology to new corners of industry. Increased
sensitivity, higher computing power, and better imaging are improving
the quality of information and ease of use. Testing is not limited to
the manufacturing cycle. Software simulation tools have been developed
to allow design teams to consider part inspection early in the design
process.
SMARTER SENSING
R. Bruce Thompson, director of the Center for Non-destructive Evaluation
at Iowa State University in Ames, points out that advances in sensing
technology have improved understanding of how energy behaves in the microstructure
of a material. Overall, sensors are getting steadily better and providing
higher-fidelity information, and the presentation of data is becoming
increasingly easier for people to interpret.
The center at Iowa State is studying methods of non-destructive testing
and developing techniques for using them. It is part of the National Science
Foundation's Industry/ University Cooperative Research Center program,
a cooperative effort of industry, academia, and the government to support
technical research.
Improvements in non-destructive evaluation are providing a varied set
of tools, Thompson said. In ultrasonic sensing, for example, there is
better understanding of how grain size and defects influence ultrasonic
waves, he said.
Phased arraysin which a piezoelectric vibrating element is divided
into many separate elements that can be driven independentlymake
it possible to create complex patterns that provide information from many
angles during a single scan. A similar idea, applied to eddy currenta
non-destructive technique that scans a coil over a part and measures the
electrical impedance for telltale changes caused by cracksprovides
detailed information about a part, Thompson said.
DESIGNING FOR TESTS
One of the developments at Iowa State is a set of modeling tools to help
determine if an investigative technique will be adequate. "In the
traditional manufacturing process, you figure out what the material is,
how resistant the material is to manufacturing, and how to make it,"
Thompson said. "But typically nobody worries about how to inspect
it."
A small start-up company, NDT Technologies, based in Charlottesville,
Va., commercializes software for use with non-destructive testing techniques.
Joe Gray, X-ray group leader at the Center for Non-Destructive Testing
and a co-founder of the company, said modeling tools exist for three non-destructive
techniques: ultrasound, eddy current, and X-ray, which are used widely
by industry. Work on software for use with magnetic techniques is in progress.
Part geometry always affects inspection techniques and it's a challenge
to make sure that you are getting the quality inspection that you need,
Gray said. The models give information about waves and currents transmitted
through different materials and geometries. The software uses a computer-aided
drawing as a starting point, along with the definition of defects and
properties of the part that would govern the waves of a particular non-destructive
test. Designers could use the models to weigh cost tradeoffs in comparing
ways to form a part.
 |
| IR thermography checks the adhesive
bond of truck bed panel. |
A simple, inexpensive casting method, for example, may result in a porous
part. When that is the case, a processor has two alternatives: either
use a more expensive casting process that puts the part under pressure,
or use the less expensive process, but combine it with a non-destructive
evaluation technique. Both processes will get good products out the door,
but it is not always clear beforehand which is the less expensive course.
The simulation allows the user to consider the kinds of defects to expect,
to evaluate a particular inspection process and its cost, and to compare
the results to the cost of using the more expensive casting process without
inspection, Gray said.
"You really need to make those decisions before you go down the
road of making prototypes," Gray said. "We have made significant
progress in developing these kinds of inspection simulations, which allow
you to let the inspection of the part be part of the optimization of the
process."
U.S. BONDS
The U.S. Department of Energy has a goal of using of lightweight materials
in automobiles as an energy-conservation measure. But to introduce new
materials successfully into automaking, efficient testing methods for
welds and bonds must be developed.
A standard practice in the auto industry today is to test spot welds manually
with a pry test, in which a screwdriver-like device is used to check the
integrity of the welds. Periodically, whole assemblies are pulled off
the assembly line for a teardown, which is labor intensive, time consuming,
and expensive.
Deborah Hopkins, a scientist who leads the Industrial Partnerships Group
for the Engineering Division at Lawrence Berkeley National Laboratory
in California, has been researching non-destructive techniques to test
welds and adhesive bonds. Hopkins works in partnership with the Non-Destructive
Evaluation Steering Committee of the United States Council for Automotive
Research, an organization sponsored by Ford, General Motors, and DaimlerChrysler
to foster cooperative research in transportation technology.
Hopkins has been working with adhesive-bonded parts made of composites
and aluminum since 1997, and has expanded her research into automated
testing of spot welds. Her group is studying phased-array ultrasonic techniques
for testing.
 |
| Ultrasonic welding of aluminum
is demonstrated in a lab. |
Hopkins said that phased-array ultrasonics can scan electronically without
physically moving the probe, and can be focused to deliver more energy
into the part for higher resolution. She added that resolution is also
improving as the probes get smaller and the frequencies higher. Hopkins's
group is developing a portable system that it plans to test in a manufacturing
plant by the end of this year or early next. So far, the group has been
applying the phased-array ultrasonic technique to galvanized steel that
is used in conventional vehicles, but plans to test the technique with
aluminum and high strength steel in the coming year.
Hopkins's group is also working with infrared thermography to characterize
defects in welds and gaps in adhesive layers. IR thermography works by
sending heat into the piece and watching the thermal response. The thermal
technique has some important advantages. It does not need a coupling agent
between the sensor head and the part, such as a gel or water column that
is required by ultrasonic techniques; it also has a large field of view.
On the other hand, interpreting thermal imaging requires special skills.
Still, Hopkins said that thermal imaging has industrial potential for
specific applications, particularly processes, such as applying coatings,
that use heat. Thermography is particularly attractive for those processes
because applying additional heat to the part is unnecessary.
A research group at the University of Windsor in Ontario is also developing
ultrasonic techniques to evaluate spot welds. Roman Maev, who holds the
university's industrial research chair in applied solid-state physics
and materials characterization, has been working on the use of sound waves
to look into materials and display the results as an easy-to-read image
on a computer screen.
Robotic spot welding is subject to many uncertainties, said Maev. Variations
in electrical current can result in welds too large or too small; if a
technician neglects to change the copper tip at the end of an electrode,
the quality of the weld could be affected. Commercial ultrasonic systems
are available to check spot welds, giving the operator a go/no-go signal
on an oscilloscope. But, according to Maev, welds with high-strength steel
would benefit from more nuanced reporting.
 |
| Thermographic image of spot welds
potentially may replace teardowns as a way of independently verifying
ultrasonic inspection data. |
Maev's group claims it has developed a portable ultrasonic imaging
device, weighing just seven pounds and operated with a touch screen, that
provides the user with an actual image of the size, shape, and internal
structure of the weld. The imaging equipment is integrated into the welding
tool, and is able to provide an instant read-out of the quality of each
weld during production. The device uses a 20 MHz signal, a resolution
fine enough to detect cracks as small as 100 micrometers.
Maev's team has built prototypes of the device, which has gone
through testing at DaimlerChrysler, where technicians without extensive
training successfully operated the test unit during the evaluations. Plans
are to manufacture it commercially through a spin-off company.
WHEN AND WHY
Thompson at Iowa State said that he saw a great deal of interest in non-destructive
evaluation from the automotive industry 1980s and sees signs of renewed
interest today. Even so, non-destructive evaluation is sometimes a tough
sell to auto makers.
Emmanuel Papadakis, a consultant on non-destructive testing, supervised
a research and development group at the Ford Motor Co. for 15 years, until
1987. In his view, non-destructive testing has a place in the manufacturing
process wherever it can be proved effective in safety-related cases. In
other cases, it depends on the cost of testing, the cost of letting bad
parts into the field, and the percentage of flawed parts that would be
produced over time.
According to George Harmon, non-destructive laboratory coordinator at
the DaimlerChrysler Technology Center in Auburn Hills, Mich., safety-critical
parts, or inexpensive parts that would cause a great deal of damage if
they were to fail, require 100 percent inspection. But non-destructive
inspection for the sake of quality can be difficult to justify because
there is always a cost.
On the other hand, low-cost inspections, with a payback of three years
or less, are more palatable, he said. Resonant frequency, for example,
which vibrates a part to check stiffness, is one example. It is fast,
automated, and has been used by DaimlerChrysler suppliers to check connecting
rods and bearings.
 |
| Seven-pound ultrasonic imaging
device developed at the University of Windsor for DaimlerChrysler
provides an image of the size, shape, and internal structure of spot
welds. |
Non-destructive evaluation can be a valuable tool for troubleshooting
and correcting processing problems. Harmon suggests that non-destructive
evaluation techniques can, in many cases, be used to gain control over
the process. Ultrasonic sensors placed inside the electrodes of resistant
spot welders, for example, can provide real-time signals to keep supplying
current until sufficient liquid puddle is produced to make a good weld.
Before his retirement, Gil Chapman was the senior manager of advanced
materials at the Liberty and Technical Affairs Group, which focused on
advanced concept vehicle development for DaimlerChrysler Corp. In the
early 1980s he worked as a researcher at the Ford Motor Co. and developed
non-destructive testing techniques to solve quality problems in automotive
plants. Chapman developed a low-frequency ultrasonic inspection technique
to test the adhesive bonds on Ford's large L-series trucks.
Adhesive bonds between the fenders and hood and their underlying supports
were coming apart on a large number of trucks once they went into service.
Dust, humidity, and curing problems neutralized the adhesive before the
parts could be glued properly. Chapman's technique used a commercially
available low-frequency sensor to test if the adhesive made a good bond.
The method was fast, simple, and did not require a liquid couplant to
transmit the waves. The manufacturing people liked the approach, which
was eventually incorporated at Ford as Test Method FLTM BU 17-1.
Non-destructive testing can help pinpoint problems before too much value
is put into a part on a production line, Chapman said. But it also can
test a process: "The closer we get to the process itself, the quicker
we can identify what the problem is," he said.
Daniel Hutchinson, an engineer with General Motors' Controls, Conveyors,
Robotics, and Welding Group, said the company is interested in using non-destructive
testing as an in-process monitoring tool and as an audit tool. Reducing
the amount of scrap from destructive testing is one of the company's
objectives in using non-destructive testing for spot welds, he said. The
company uses ultrasonics on its plant floor to check the quality of welds,
and increased computational power has led to an increase in testing.
Harmon at DaimlerChrysler also sees non-destructive evaluation as an enabler
of new materials. More stringent fuel economy and emissions requirements
have stimulated the increasing use of materials such as aluminum, magnesium,
and plastic.
"There is a tradeoff between weight and strength," Harmon
said. "As you reduce the strength of material or assembly, you
need to be that much more concerned about the quality of the individual
component."
George Mozurkewich, a technical leader in Ford's Research Laboratories
in Dearborn, Mich., said that Ford is involved in a NIST program to develop
ultrasonic welding to join aluminum alloys. Resistive welding used for
steel does not work as well for aluminum, which requires much more electricity
to generate enough heat. In Mozurkewich's view, automation could
open the doors to wider use of non-destructive evaluation in the automotive
industry.
"I can envision tests on line that are robotically operated, that
don't interfere with the smooth flow of production, and that automatically
take care of the logistics of data tracking," he said.
Overall, manufacturers are acknowledging the value of selected non-destructive
evaluation techniques, and researchers are getting better at adapting
non-destructive evaluation techniques to the plant floor through portability,
fine resolutions, and smart software.
home
| features | breaking
news | marketplace
| departments | about
ME | back issues |
ASME | site
search
© 2004 by The American Society
of Mechanical Engineers
|