Closed-Loop Control in Hardness Testing
For more than 75 years, the most common hardness-testing instruments in industry used dead weights to apply the test forces. Benefits of using dead weights include their low cost and the relative ease of manufacturing them to the degree of accuracy required by commonly used test methods. However, their use also presents a problem; that is, the force must be applied to the test piece through some type of small indenter, such as a diamond or ball indenter. The transfer of a dead weight force to the tip of a small indenter is difficult to accomplish, especially a force as great as the 150 kgf, or 1471 N (330 lbf), used for a Rockwell C scale (HRC) hardness test, for example. Because of the large size of a 150-kg weight, it is necessary to design the testing instrument using smaller weights and levers to magnify the force to the required levels. However, the levers require pivots, guides and other friction-producing elements that induce errors. Instrument manufacturers have done an excellent job trying to control these sources of error, but any friction point in the system has a negative effect, which increases during use.
It also is difficult to control applying the weights in a dead weight system. To apply the test force, the weight has to be moved, and stopping it quickly without overload and oscillation creates a problem. Many older testers use dashpots to control applying the weight. However, dashpots are prone to serious variations in test results due to seal wear and temperature changes, so they have been replaced with a motor in many newer instrument designs. While this eliminates some of the dashpot problems, the desire to perform tests quickly makes the motor speed critical, as force overshoot and oscillation are frequent problems.
Open- and closed-loop systems
Instruments that use dead weights typically are open-loop systems. In other words, forces are applied based on the calculations of the weights, lever ratios, etc. The manufacturer performs an initial calibration to make sure the forces applied are within tolerance by using an independent measuring device (typically, an electronic load cell), and they usually are never checked again during the life of the instrument. It is assumed that forces remain consistent as each test is performed. While dead-weight systems have proven to work very well in many applications, including hardness testers, there always has been a performance level that typical (i.e., affordable) dead-weight systems could not surpass due to the inherent problems (Fig. 1).
During the 1950s, closed-loop systems came into common use on tensile testing instruments. Closed-loop systems are different from open-loop systems in that they have a means to electronically measure the force being applied during every test, and then feed (or loop) the information back to the control system. The control system is designed to use the feedback to adjust the force application mechanism to apply only the desired force (Fig. 2). These systems work so well that today all electronic tensile/compression testing instruments use closed-loop control exclusively.
Closed loop better for hardness testers
A closed-loop system can constantly measure the test force being applied, and also the components used in a closed-loop system inherently lend themselves to a much simpler design than a dead-weight system. As previously mentioned, dead-weight systems require levers, pivots and other friction-inducing components to function efficiently (Fig. 3). The indenter, the only part of the system in contact with the test sample, is detached from the weights themselves, separated by the levers and pivots, etc.
In contrast, the main component of a closed-loop system is a strain-gage load cell. This compact, lightweight device provides an electronic output proportional to the force applied to it. Load cells are available in many different shapes; therefore, it is possible to design a hardness system having the indenter attached directly to the load cell (Fig. 4). In this design, sources of error between the indenter and the test force are eliminated. While this design uses actuators to apply the test forces, which have bearings and sliding surfaces that can introduce friction, it also isolates the negative influences above the load cell so they do not affect the critical test force.
For example, if friction in the actuator is so excessive that the desired force is not applied to the indenter, the load cell will not indicate the correct force. Therefore, the system will abort the test rather than produce an incorrect result. In this way, the system is constantly checking itself to make certain that only the correct test forces are applied to the indenter. The mass of the actuator can easily be controlled because of the feedback loop.
A common way to measure the performance of a hardness tester is to use GR&R (gage repeatability and reproducibility) techniques. In this method, an attempt is made to quantify the performance of measuring instruments by comparing variations from an instrument with the total variations allowed for the part that is being measured. The result is a value (in percent) that indicates how much of the tolerance is being used up by the instrument. The smaller the percentage, the better the instrument performance. Typically, users of this method want to obtain GR&R results of 10% or less. However, 30% is acceptable in some situations. Hardness testers frequently fall into the 30% category because they typically don't perform that well, and variations within the sample consume a percentage that is difficult to quantify.
Depending on the age and design of a hardness tester, GR&R results from typical dead-weight Rockwell scale testers normally range from 12% to 25%. Under tightly controlled ideal conditions, a 10% target has been achieved. These results, however, do not reflect every day reality. Under the same conditions, a properly designed Rockwell tester using a closed-loop system can routinely achieve GR&R results of less than 7%. Tightly controlled units have achieved results as low as 2%. (A value of 2% is considered the lowest attainable percentage due to the nonuniformity of the test samples.) In addition, closed-loop systems have proven to be more stable, which increases the reliability of the test data.
Benefits of increased performance
Required system accuracy depends on how important test results are. If you are just trying to verify whether a part has been heat-treated, 10% of GR&R improvement might not be important. However, if you are working to specific tolerances, any reduction in the uncertainty of results can save money by minimizing the possibility of either rejecting a good part or accepting a bad one. Having a better knowledge of the hardness value offers the possibility to make adjustments to manufacturing processes to achieve the most economical operation.
Today, "uncertainty" is the buzzword for persons performing calibrations. For example, persons working to ISO Guide 17025 must provide an uncertainty statement with most calibrations performed, including hardness. A logical extension is that customers might someday request an uncertainty statement with every test performed. While this will be a difficult value to accurately determine, the calculation will be significantly influenced by the performance of the hardness tester. The better the hardness tester performs, the lower the uncertainty will be.
Hardness testers using closed-loop systems currently are available for Rockwell, Vickers, Knoop and Brinell testing in a variety of test force ranges. Load cells typically have force range limitations of 100 to 1. In other words, if the lowest force is 10 kg, the highest force is 1000 kg. This generally is a greater range than most dead-weight testers provide. Additionally, a closed-loop system has the capability to allow the use of any incremental force within the usable range. By comparison, dead-weight systems are restricted to the discrete weights. Some of the newer load cells can exceed the 100 to 1 limitation.
The inherent flexibility of closed-loop systems also is a benefit. Because the force application process is controlled using a microprocessor, test cycles can easily be changed. Not only is this feature desirable for special testing requirements, but also it guarantees that a tester can easily be modified to meet any new or revised test method. This could be very helpful, for example, if you are interested in having a Rockwell tester that can match the time cycles used on the new NIST (National Institute of Standards and Technology) Rockwell hardness standards.
Closed-loop systems are proving to be the preferred method to perform a wide range of hardness tests. A closed-loop system should be considered if having a hardness tester that provides the best possible hardness test results is important.
SIDEBAR: Closed-loop control improves GR&R
A unique electronic closed-loop control system design uses a precise in-line test system, which offers a highly accurate, repeatable tester. The load cell is centrally located between the indenter and the depth measuring system, forming an error-free, precision test axis and eliminating conventional mechanical linkages and friction. With the indenter attached right to the load cell, preliminary and total test forces are applied with unparalleled accuracy and repeatability (Fig. A). In addition, with the high-resolution optical linear measurement device located directly above the force system, precise, repeatable Rockwell hardness test results are achieved on any Rockwell scale.
Tester repeatability, variation and operator influence is determined by means of a GR&R (gage repeatability and reproducibility) test. GR&R is a statistical calculation that compares the ratio of tester and operator variation to an overall process tolerance (typically 4 - 6 HRC). A GR&R study determines how much of a process tolerance is being used up by variation in the tester (Fig. B). Low GR&R values (in percent) indicate a high-performance tester.