A New Era for Hardness Testing

After more than a decade
of work, NIST has released hardness reference standards
that are in line with
international requirements.

by Dirk Dusharme

Until recently, the United States didn't measure up to the rest of the world when it came to Rockwell hardness testing standards. Where other countries had standard hardness test blocks traceable to national standards, the United States had none. When other countries did intercomparisons of national standard reference materials and created country-specific

For more information

NIST Standard Reference Material Program

Building 202, Room 204
Gaithersburg, MD  20899
Phone: (301) 975-6776
Fax: (301) 948-3730

Web: ts.nist.gov/ts/htdocs/230/232



100 Barr Harbor Drive
West Conshohocken, PA  19428
Phone: (610) 832-9585
Fax: (610) 832-9550

correction factors, the United States was not officially included. And, while there was close agreement between European and Japanese hardness test results on standard reference materials, the United States has for years been known to produce lower hardness test numbers, particularly in the higher, 50 to 70 HRC range, where the difference from the Europeans or Japanese ranged from 0.5 to 1 point.

A new ASTM Rockwell test standard that may be released next year, coupled with the new NIST hardness test blocks released last year, will put the United States on track with the rest of the world.

The Rockwell scale, which was developed in the 1920s by U.S. instrument maker Wilson Instruments, has been utilized by U.S. industry for decades. A handful of companies manufacture hardness test blocks, which more or less tracked with each other. Rockwell tester manufacturers and users understood how to calibrate their equipment and, for the most part, mitigate differences between the various hardness test blocks.

Even on the international level, the differences between U.S. and international blocks were well-known and understood, and manufacturers and customers both within and outside the United States handled the differences with little problem.

However, with the growing use of ISO 9000 and ISO/IEC Guide 25 requirements of traceability to national standards, a potential trade barrier began to loom.

 "Countries could say that they would only accept products from countries whose testers were traceable," acknowledges Sam Low, materials research engineer with the National Institute of Standards and Technology, and program manager for the Rockwell Hardness Standardization Program. "In practice, this is normally not a problem . except in hardness," where the United States had no standard, he adds. "Some companies would not buy U.S. hardness test blocks because they were not traceable to national standards."

So, in the late 1980s, industry, through the American Society for Testing and Materials, got NIST involved. NIST ran intercomparisons of all U.S. commercial test blocks and diamond indenters. At the higher HRC scales, the blocks varied from each other by as much as one point, says Low.

In 1991, NIST purchased a hardness standardizing machine from the Italian national metrology standards laboratory and, after making several improvements, developed the first set of standard test blocks. A comparison study showed that they closely match their European and Japanese counterparts. Last year, NIST released the first set of hardness standard reference materials (SRMs), providing test block and tester manufacturers, and tester users a  traceable standard. Currently, only the HRC scale blocks are available.

But the SRMs don't complete the story. The United States still needs a hardness testing procedure that requires not only the use of the NIST SRMs, but also is more in line with international requirements, explains Low, who chairs the Traceable Hardness Standards task group. The group is working on revising ASTM's E18 Rockwell hardness testing standard. The existing E18 procedure for performing Rockwell tests is no longer adequate from either a national or international perspective. The new standard could be released some time next year, says Low.

The proposed changes in E18 could cause significant but, hopefully, short-term headaches for Rockwell tester users. To help customers understand the changes and to prepare for them, Ed Tobolski, general manager of the Wilson Instruments division of Instron Corp. and a member of the committee revising the E18 standard, made available in October a brief description of the proposed changes to all users of Rockwell hardness testers.

"People are not aware of the new     standards in general," says Tobolski, explaining the purpose of his letter. "They normally don't pay attention to it until it hits them in the face." If users are prepared for the changes, the transition won't be too tough, he adds.

In his letter, Tobolski outlines seven key areas of the proposed changes to E18 that he believes will have the most impact on Rockwell hardness tester users.


Direct verification

Tests have found that 90 percent of testers in use today do not apply the correct test forces or measure penetration depth correctly, says Tobolski.

The old, indirect method for verifying the correct operation of a Rockwell tester involved simply measuring a standard test block and verifying that the tester gave a value within the tolerance of the block. The feeling being that if the test read correctly, then the force and depth measurement systems must be OK.

"The problem with this method was that the person doing the indirect verification would adjust the tester to match the block," notes Tobolski. This "block chasing" could lead to incorrect test results on the customer's parts, he adds.

Tom Farrell, Mitutoyo test equipment product manager and E18 revision committee member, agrees. By mixing and matching test blocks and indenters from various manufacturers or by adjusting the tester, a tester can be made to read test blocks, he says. It was just such tweaking and the lack of standard reference materials that caused a one-point jump in hardness test results during the 1960s, claims Farrell. Testers chased blocks that, across industry, had been slowly drifting upward, rather than question the accuracy of the blocks or the indenters they were using.

Direct verification, which involves verifying the accuracy of a tester's test forces and depth measurement system, will address this part of the equation, claims Tobolski.

This test will be required once every three years under the proposed standard.

Adjustable load testers

This part of the standard is aimed at testers designed to allow operators to adjust load levels.

"For those who manufacture dead-weight testers, this is a sacrilege," says Farrell, adding that adjustable force indenters were meant to adjust for differences in test blocks or indenters when calibrating a tester.

However, this type of calibration, under the new standard, should no longer be needed, acknowledges Farrell. "If the load is correct and the depth measurement device is calibrated, then the only variation is the test block and the indenter," he points out. Because the test block will now be standard, due to the new NIST SRMs, any variance would be in the indenter. Hand-selecting an indenter will then be a straightforward process, claims Farrell.

The impact of this section of the proposed standard should only be felt by companies that make or use variable-load machines. These machines would need to be modified so that the loads are fixed, eliminating operator tampering.

NewAge Industries manufactures adjustable load testers. While gross load adjustments should not be allowed, small adjustments are still required to offset cumulative errors from the indenters, load cells and frictional forces, says Giancarlo Mazzoleni, technical director at NewAge and also a member of the E18 revision committee.

He points out that there are only two places to adjust the offsets: the load and the electronic display. He questions why loads should be fixed but the displays can be adjusted; he views Wilson's advocation of fixed loads as a competitive move. Wilson does not make adjustable load testers. "Wilson has been pushing this change in order to disallow companies that manufacture adjustable force testers," claims Mazzoleni. "All this does is shift the adjustment from the load to the electronic display."

Despite the disagreement, recent NewAge testers have an electronic load lock that satisfies the fixed load requirements of the proposed E18. In addition, those machines will give an error if the operator tries to test at a different load, says Mazzoleni.

Direct measurement of diamond indenters

This requirement may be the most controversial of the proposed changes. It requires that the angle and tip radius of each diamond indenter be measured and documented. While E18 had always specified measurement, there was never an implicit requirement that the indenters actually be measured, says Low. With the proposed standard, measurement is a must.

"This is extremely hard and costly to do correctly," argues Tobolski, who is against including this in the standard due to the difficulty involved. "Even NIST, when they were setting up their new equipment for the standard, had a difficult time." After selecting a dozen diamonds that met the specifications, the diamonds did not perform the same during a performance test, says Tobolski. The correlation between a hardness measurement and the diamond's shape, polish, angle, radius and position of crystallographic axis is not well-understood, he adds.

Farrell, on the other hand, agrees with the requirement and argues that once you have a batch of diamonds that meet the measurement specification, it is easy enough to sort them during a performance test, where the tester and the SRM are known.

"We do know that if you have a group of indenters with the same geometry, they are more likely to measure the same in a performance test," concurs Low. He stresses that NIST's Precision Engineering Division is working on developing a highly uniform indenter that would be both interchangeable and reproducible.

Test reproducibility issues aside, Low points out that another good reason for adding the measurement criteria to the standard is to bring E18 in line with international standards such as ISO 6508, which does call out indenter measurement.

A simple go/no-go projection method of indenter inspection may prove adequate for commercial use, says Low. More precise measurement might be required only at the national laboratory level.

The impact, if precise measurement is required, could mean a 25- to 35-percent increase in the cost of the indenter, predicts Tobolski.

Carbide balls

When ball indenters rather than diamond indenters are used, the new standard calls for carbide balls rather than steel balls. Carbide balls cost more but last longer. This also brings E18 into alignment with new ISO standards. The change may result in an increase in hardness readings on harder materials, according to Tobolski. The magnitude of any change is now being reviewed.

NIST-traceable test blocks and indenters

In the proposed standard, when available, all test blocks must be traceable to the new NIST SRMs when doing indirect verification by the calibrating agency and daily verifications performed by users, imparts Tobolski.

NIST-traceable indenters must also always be used, when available.

The reasons for this are obvious. If all Rockwell testers are correctly set up and verified using standard hardness blocks and standard indenters, then all Rockwell testers will produce similar results--something that frequently is not true.

Test cycle change

The one part of the proposed changes that could cause problems for some users is the new test cycle time. The old standard called for a maximum of a three-second dwell time for the major load, with no minimum specification. Some automated testers have dwell times as short as one-half second. The new standard will require a four-second dwell time, plus or minus two seconds.

The reason for the new requirement is that faster dwell times don't allow for material creep once the total test force has been applied, explains Low. The material creep rate is fastest at the beginning of the cycle, and machines set up for short dwell times may not give good repeatability.

There has also been some squawk     from industry over this requirement, says Low, but he adds that the fears are mostly unfounded. Actually, in response to industry concerns early on in the E18 revision process, the committee broadened the tolerance from one second to two seconds.

"It's been concluded that nearly all machines should now be able to meet the proposed testing cycle," asserts Low. "We're not trying to make it impossible for companies to do; we're trying to improve the test."

However, he acknowledges that for manufacturers utilizing automated testers with short dwell times, the new standard may cause some problems. "Some companies do hardness testing full-time," says Low. "Increasing the amount of cycle time will cut the amount of product they can test during a shift."

Accreditation of calibrating agencies

Currently, anyone can perform calibrations on a hardness tester, test blocks or indenters, says Tobolski.

The new standard will require all test block and indenter calibrating laboratories to be accredited to ISO/IEC Guide 25 by a recognized accreditation body utilizing ISO/IEC Guide 58. Fortunately, this impacts only the half-dozen or so test block, indenter and Rockwell standardizing tester manufacturers in the United States. Originally, some committee members were pushing for accreditation of the people who do the field calibrations (direct and indirect verifications) as well, but backed off in the belief that market pressure would eventually force accreditation in this area.

One impact that may concern some test equipment manufacturers is that there may not be auditors qualified enough to assess hardness testing, says Farrell. He believes that probably no auditors currently working for the accreditation agencies are technically capable of conducting an accurate hardness calibration audit.

Tobolski disagrees, pointing out that he knows auditors qualified to perform the tests and that, in any case, only five or six labs need to be audited.

The issue is further muddied because any experts in this rather esoteric field that would be hired by an accreditation agency would probably have come from one of the hardness tester manufacturers. Because that list of names is small, it is unlikely that Mitutoyo, for instance, would want someone known to be affiliated with a competitor to audit their site, says Farrell.

Overall impact

Once machines have been calibrated to the new standard and a company has the new SRMs and indenters, the remaining impact for some companies will be to adjust specifications and prints to the new hardness standards. In the C50 to C60 ranges, users may see an increase of 0.5 to 1 point, reports Tobolski.

"The easiest thing to do is to change the hardness spec for the part," he says. He also suggests that companies show both old Rockwell numbers and new NIST-traceable numbers on their prints and test reports. Showing both numbers will help decrease any confusion between supplier and vendor in terms of what is being supplied. It could even help educate those unaware that a new standard exists. This would only be until the new E18 is released some time next year, notes Farrell. The standard will dictate how to document results.

After more than a decade of work, NIST has released the United States' first Rockwell hardness standard. It provides a standard reference for all hardness tests performed in the United States, one that closely matches international standards. Come next year, if the E18 revision process goes as hoped, there will be no "old standard" or "new standard," there will be only the NIST standard. At that point, companies had better be ready.

About the author

Dirk Dusharme is technology editor of Quality Digest.

[QD Online] [Project Cost] [Color] [Excel] [Gaging] [Hardness] [Gage Guide]

Copyright 1999 QCI International. All rights reserved. Quality Digest can be reached by phone at (530) 893-4095. E-mail: Click Here