NIST & ASTM International
A Partnership that Benefits Industry
When the National Institute of Standards and Technology performs research and ASTM
If there is a single fact that characterizes the NIST commitment to industry, perhaps it is this one: 149 NIST scientists and engineers currently belong to ASTM International; among them they hold memberships in 398 technical committees and subcommittees. This level of commitment is not new. ASTM International has been developing voluntary consensus standards for 110 years, and for 107 of those years NIST has proven to be a reliable partner, not only through memberships but also through formal and informal cooperative programs custom-tailored to industry’s changing needs.
The NIST mandate is to develop and maintain the technical tools that U.S. industry needs to stay competitive and America itself needs to remain economically and physically secure. NIST’s support of ASTM, however, has had a dramatic impact on industry worldwide, and the organizations’ partnership — in all of its forms — continues to affect not only long-established industries but emerging industries as well.
Robots for the Real World
Faced with the growing challenges of terrorism and natural disasters, governments and industry are seeking new avenues for developing essential technologies to protect people and infrastructure and to respond to critical incidents. NIST and ASTM are providing those avenues. One example is in the field of urban search and rescue robots, a young discipline that attracts talented young scientists and engineers like Elena Messina and Adam Jacoff.
Messina is acting chief of the Intelligent Systems Division of NIST’s Manufacturing Engineering Laboratory. She explains that urban search and rescue teams currently do not use robots. “There have been attempts, but they’ve all been by researchers. Research teams brought robots to help search the pile after 9/11. They were used during Katrina and also with mudslides. Not a lot of success in general, but the potential is clear.”
The Department of Homeland Security saw that potential and came to NIST for help in developing standards for the technology. “We already had a lot of general expertise in robotics,” says Jacoff, an Intelligent Systems Division mechanical engineer, “and conducted readiness assessments of autonomous mobility for the U.S. Army.” NIST even designed test courses for US&R robot competitions sponsored by the Defense Advanced Research Projects Agency in 2000. Meanwhile, NIST’s Office of Law Enforcement Standards was working with the law enforcement and public safety communities on performance standards for bomb disposal robots.
“Participation is a critical factor in any activity,” says Timothy Brooke, assistant vice president, ASTM Technical Committee Operations, and staff manager for Committee E54 on Homeland Security Applications. “If you don’t have all the stakeholders at the table — and I’m talking about the government agencies, the industry, the users, the suppliers, the academics — you’re really not in the position to be as effective as you could be. Providing a forum where all these relevant individuals and experts and knowledgeable people come together, that’s what ASTM’s role is.”
ASTM formed a task group on robots applied to urban search and rescue within Subcommittee E54.08 on Operational Equipment, and the members got to work on designing test methods for various performance requirements. As work progressed, NIST found an additional and unique way to evaluate the test methods: incorporating them into an ongoing annual series of events called RoboCupRescue, where US&R robots from around the world compete to demonstrate mobility, sensory perception, mapping, remote teleoperation and even assistive autonomous capabilities. Today the RoboCupRescue arenas use almost all of the initial set of draft test methods.
“RoboCupRescue has helped exercise and refine the test methods,” says Jacoff, who chairs the RoboCupRescue Robot League. “It really is guiding robot developers toward answering first responder needs, and the winners are showing us the next wave of robotic capabilities.”
NIST has found a way to catch that wave by inviting the best-in-class winners of RoboCupRescue to participate in exercises that challenge the robots in real-world search-and-rescue environments. “We host these events at various FEMA [Federal Emergency Management Agency] training sites,” says Messina. “There’s one up the road at the fire and rescue training facility in Montgomery County [Maryland] and there’s Disaster City.” Disaster City is a training facility at Texas A&M University that includes environments ranging from collapsed buildings to train wrecks. Along with research robots, dozens of commercially available robots — typically developed for applications other than US&R — have engaged in the exercises, in which the FEMA US&R task force members operate them in realistic training scenarios and practice the draft test methods.
The response robot evaluation exercises have no spectators, just robot developers, their robots and emergency responders from across the country there to see what the latest technologies can do. “The developers want to come to these events,” says Jacoff, “because we are giving them unique access to this panel of responders at facilities that really allow them to test their robots. What encourages me most is how willing they are to have their robots fail in these extremely challenging environments. They’re clearly interested in seeing what their robots can and can’t do.”
“The beauty of this,” says Messina, “is that we’re getting the responders involved at the point when these technologies are emerging so they can help guide development toward addressing their requirements. We’re working on performance test methods, not design standards, so we’re not pinching off any development. Developers are free to solve the problem any way they want, and that’s what we’re seeing.” She laughs. “You could say that we’re all working unhindered by what we think we already know,” she says.
Creative Metal Work
“The most difficult issue right now for people in the metals sector,” says John Sieber, a research chemist in the Analytical Chemistry Division of NIST’s Chemical Science and Technology Laboratory, “is satisfying their auditors and accreditors that they have laboratory quality under control.” One issue is the lack of analytical tools to handle the steady stream of new alloys; industry is developing new alloys faster than NIST can develop Standard Reference Materials for them.
NIST SRMs have been at the core of U.S. industrial quality for more than a century. The institute’s catalog lists almost 400 active SRMs for industrial metals alone. That is an impressive achievement, but it still falls short of the metal trade’s requirements, especially since the advent of ISO 17025, General Requirements for the Competence of Testing and Calibration Laboratories, and Quality Assurance Systems. It is a persistent problem that has tested NIST’s and ASTM’s shared creativity for decades.
The logjam began around 1960, with foundries producing a whole range of new alloys, particularly of copper and aluminum. Then the aerospace industry created a huge new market for titanium alloys, and the nuclear industry needed alloys of zirconium. Thomas O’Toole, staff manager for ASTM Committee E01 on Analytical Chemistry for Metals, Ores and Related Materials, explains that by the late 1970s there were three technical committees turning to NIST for SRMs related to metals and ores, and NIST’s resources were insufficient to handle all the requests. “NIST was given way too many reference materials to develop,” O’Toole says, “and finally said, ‘We can’t do this anymore. Work on your priorities.’ ”
ASTM did, by creating a committee of its board to vet SRM requests, a function now handled by a subcommittee of Committee E01 on Analytical Chemistry for Metals, Ores and Related Materials. Then, in the 1980s, the two organizations signed a Memorandum of Understanding that supplemented NIST’s SRM efforts with industry funds, which are continuously reinvested in new and renewed SRMs. Under its Congressional mandate, NIST is required to recover its investment in SRM development through sales. ASTM’s investment is recovered the same way, through a surcharge on the price of each SRM. NIST returns the money to ASTM, and each year ASTM reinvests it in the program. “That,” says Sieber, “keeps some of the metals and ores projects going.”
Still, the metal sector’s need for reference materials far outpaces NIST’s ability to meet them. So Sieber, a member of ASTM Subcommittee E01.22 on Laboratory Quality, has sought increased involvement from both manufacturers and laboratories within the industry. The goal is for NIST to concentrate on providing key SRMs while supporting increased development of certified reference materials, CRMs, by commercial laboratories. CRM is the internationally accepted term for the highest quality reference materials. NIST uses the term SRM for its own CRM products.
Industry involvement has also led to cooperative successes in solving specific problems. “For a number of years some of these high-tech alloy companies were asking NIST for more values on a particular nickel-cobalt alloy SRM,” Sieber says, “and it was an expensive thing where we were going to have trouble doing it. So we went to a variety of companies, both manufacturers and labs that provide for third-party testing, and basically they did a cooperative analysis project for us.” NIST provided the samples of material, the private laboratories conducted the tests and returned their data, and NIST statisticians calculated the best estimation of the true value based on the data. “We put those numbers on our certificate as what we call reference values,” says Sieber, pointing out that these particular reference values are not a result of NIST analysis. “They are pretty good numbers, and they answer some of the questions that the manufacturers and their customers had. The manufacturers are very happy with it, and I think this could be a model for certain programs.”
Sieber and others see such creative solutions as at least a partial response to the high demand for reference materials in the metals and ores industries, and he feels strongly that ASTM is the place to find these solutions. “Having the manufacturers and the customers together making standards definitely has an advantage,” he says. But he also believes that the situation demands even more participation. “We have a long list of members in Committee E01, but only a small number of companies allow their laboratories to participate in studies to help make reference materials. There are a lot of resources out there that we’re just not using. If all these organizations banded together and worked at a higher level, we could get a lot more done.”
Hardness and Calibration
Samuel Low is a materials research engineer and NIST’s leading expert on Rockwell hardness, a measure of a material’s surface hardness, which is often n important indicator of durability. The Rockwell hardness test method, which consists of driving a tungsten carbide ball or conical diamond indenter into the material’s surface under a specified load and measuring the depth of the indentation, was developed in 1921 and has been employed around the world ever since, especially by manufacturers and buyers of steel. “It is a very fast test,” Low says, “takes about 15 seconds, and is fairly nondestructive. It’s the most used test for quality control in the industry.” In fact, ASTM standard E18, Test Methods for Rockwell Hardness of Metallic Materials, is among the most downloaded test methods in the ASTM catalog.
In the 1980s, however, a growing number of problems led U.S. industry to realize that the foundations of its Rockwell hardness test method lacked consistency. What followed were 20 years of intense discussion and technical effort that led NIST to develop a new SRM and ASTM Committee E28 on Mechanical Testing to issue a major revision of the test method in 2007.
The core problem was test machine calibration, which is done by testing standardized hardness blocks of known characteristics. Private laboratories produce the blocks, and a study at the time showed significant variations among blocks from different suppliers. “It was a chicken-egg thing,” Low says. “The labs had hardness blocks to check their machines, but there was no national standard to check the blocks against. And so errors were repeating themselves in cycles and being compounded.”
The consequences were serious. “Someone makes a material that is right on the edge of the acceptable tolerance range, which is okay,” Low says. “Then they send it to the buyer, and the buyer’s machine is slightly off, and that kicks the material outside the tolerance range. Suddenly there’s a dispute. There were lots of rejected material that should have been accepted and accepted material that should have been rejected. And that’s costly.” In addition, the drift in machine accuracy was pulling the U.S. out of alignment with the Rockwell hardness scales used in other countries.
In 1988 ASTM approached NIST to develop an SRM for Rockwell hardness: sets of hardness blocks of uniform, thoroughly documented characteristics to serve as a national reference to which all hardness blocks produced by other laboratories would be traceable. “When we issued that standard,” Low says, “it ended up shifting the Rockwell C scale [the scale applicable to steel] in the U.S., which brought us more in line with other countries, but it also caused a lot of headaches. People were afraid they would have to scrap their machines, which wasn’t true. Some people had to replace their indenters, which they probably had to do anyway, and do basic maintenance. That was about it. It took about a year and a half for everything to settle down, and now everything’s fine.” The result has been more consistency and fewer disputes throughout industry, both nationally and internationally.
With the calibration problem resolved, ASTM Subcommittee E28.06 on Indentation Hardness Testing, which Low chairs, moved on to a major revision of the E28 standard. After 10 years of balloting, it was issued last year. “The new version, in my view, states clearly what the old version intended,” Low says. “It does have improvements as well, but the big change is the format. It’s much clearer. It works much better.”
That’s proof that even a tried-and-true 85-year-old test can be improved.
A Hands-on Approach to Quality
According to Scott Orthey, staff manager for ASTM Committees C01 on Cement and C09 on Concrete and Concrete Aggregates, membership in these committees has more than doubled in the past 10 years. One reason is the surge of construction in developing countries. Another is intensifying awareness of the need for the U.S. to deal with its own aging infrastructure. While many industries have flocked to the quality assurance philosophy to maintain the integrity of their testing laboratories, the cement and concrete industries are sticking with what has worked best for them since 1929: the Cement and Concrete Reference Laboratory.
“The government had issues with the uniformity of cement used on federal construction projects back in the 1920s,” explains Ray Kolos, CCRL director, “and complained to ASTM that it needed better standards. It turned out that what was really needed was people to go around the country and make sure that laboratories were running the tests correctly.” ASTM and the federal government created CCRL to do just that and gave it a home at NIST, where it would have access to the Institute’s technical resources. “Right from the beginning it was a success,” Kolos says, “and that has continued ever since.”
Today, CCRL is headquartered in the Materials and Construction Research Division of the NIST Building and Fire Research Laboratory. The CCRL staff of 16 inspectors, all ASTM employees, travel to almost 500 cement and concrete testing laboratories a year, completing a full inspection cycle of 1100+ laboratories every 27 months. Inspectors arrive at each site with what amounts to a mobile measurement laboratory calibrated to NIST standards. Inspectors spend most of two days checking critical equipment dimensions and operating characteristics, observing technicians conducting ASTM test procedures and reviewing elements of the quality system covered by ASTM standards. At the end, CCRL provides the laboratory with an account of how well its operations meet ASTM standards requirements.
CCRL also conducts Proficiency Sample Programs to help laboratories monitor their performance between on-site assessments. “These are programs where we obtain bulk materials — hydraulic cement, concrete, reinforcing steel, fly ash, concrete masonry units — and process them to make homogenous samples,” says CCRL General Manager Peter Spellerberg. “We distribute the samples in pairs, ask the participating laboratories to perform a series of tests on them and return the test results for analysis.”
“The most important thing about the samples is that they have uniform test properties,” says Kolos. “The specific test properties of the material are determined statistically through the use of consensus values. Based on the number of standard deviations that an individual laboratory’s test results are from the consensus values, we determine a rating. And that rating gives each laboratory an idea of how well it’s doing.”
CCRL inspections and PSPs give industry laboratories invaluable information for maintaining quality, and they support ASTM and NIST standards programs. When an ASTM technical committee has a test method in draft status, CCRL can ask laboratories participating in its PSPs to conduct tests using the new or modified method in addition to the usual battery of tests. This gives the committee real-world feedback at the draft stage. When ASTM issues a new or modified test method, CCRL inspectors provide tutorials and answer questions in the labs they visit. “If we see that a number of laboratories are having a problem or are confused about a new procedure,” says Kolos, “we’ll note that and take it back to the technical committees. And we might make a suggestion about changing the wording to make it clearer.” Once the change is made, CCRL’s subsequent inspection results provide a follow-up to evaluate whether the change resolved the problem.
Orthey adds that CCRL is playing a role in the 225-plus interlaboratory studies that ASTM is conducting to provide precision statements for its test methods. “Because of CCRL, we have a large pool of laboratories to draw from,” he says. “And we use those laboratories for tests on materials other than concrete. Coal, for example.”
For NIST, CCRL is a source of materials and data. “We assist NIST by obtaining the materials they use for developing portland cement SRMs,” says Spellerberg. “We may also do some processing of the materials, and we certainly provide data from the Proficiency Testing Programs.” NIST and its industry partners have also used PSP data to develop computer models for BFRL’s virtual Cement and Concrete Testing Laboratory, a Web-based software package for predicting the performance of cement-based materials and reducing the number of physical tests required by concrete producers, engineers and researchers.
“We are at NIST,” says Kolos, “because we share the same goal: improving the testing of materials. In that regard we are a good fit, a partnership that has definitely worked for the betterment of the country and the industry.”
Robert Ausura is a writer and photographer living in Gaithersburg, Md.