Encoder calibration, safety audit

Women in engineering? Yes and no Re: Articles and postings on the skills gap and particularly women in engineering. I am a chemical engineer and my 20-year-old daughter is a chemical engineering student at the University of Michigan. She said that she got an early interest in science from watching Bill Nye—The Science Guy on television.

By Control Engineering Staff September 1, 2008

Women in engineering? Yes and no

Re: Articles and postings on the skills gap and particularly women in engineering.

I am a chemical engineer and my 20-year-old daughter is a chemical engineering student at the University of Michigan. She said that she got an early interest in science from watching Bill Nye—The Science Guy on television. She is doing well even though her high school physics teacher decided to skip the thermodynamics chapters in their book. I hope he listened when I explained to him later that thermo is one of the foundations of chemical engineering.

My younger daughter is still in high school, but is not considering engineering. One of her math teachers probably helped discourage kids by her own attitude about the material she was teaching. When she was explaining matrix multiplication, she told the class, “You’ll never use this.” Some of the people who do use it are the people who design atmospheric simulators and the rich kids who program video games.

John Kreinbrink, Process Engineer

Encoder calibration

Q : At the aerospace company where I work, all CNC systems that have incremental encoders are calibrated regularly and adjusted as required. We have robots in some of our thermal spraying cells that are not calibrated, but their motion repeatability is not as critical. We are planning a new robot-based processing system where repeatability and accuracy at the same time are important. The robot uses absolute encoders and goes through self-checks on each start-up. Therefore, the manufacturer states that no calibration is necessary. Is this your view?

Kartik Shanker, P.Eng. Repair Development Engineer, Winnipeg, Manitoba

A : There are actually two issues here: process stability/accuracy, and machine safety. Depending on the encoder technology, regular calibration may not be necessary. You may, however, want to check system accuracy periodically in order to maintain the machine’s safety integrity level (SIL).

Many encoders, such as glass absolute encoders, use digital readouts of positions of mechanical structures (fiducials etched into the surface) that cannot be modified without damaging the encoder. That is, they do not “age” like most electronic components. Furthermore, they may be locked into position mechanically by keyways, etc., so they can’t slip out of position. In these cases, you may not need regular calibration to maintain process tolerances. The manufacturer is your best guide. Get it in writing, and also get a copy of the encoder’s calibration certificate. If there’s no calibration certificate, then any claims as to accuracy or repeatability are so much sales talk.

That, however, just covers the encoders themselves. They are not the only things that move. Robot arms can bend under repeated loads. Bearings can wear, becoming wobbly enough to violate motion tolerances. A good calibration program can uncover these problems as well.

Safety is another matter. Safety issues arise when talking about calibration because calibrating a piece of equipment will uncover most safety-related issues. Calibration engineers and technicians start their ministrations by doing a thorough safety check for their own self preservation. A calibrated piece of equipment is, within parameters set by the equipment’s original safety risk assessment, a safe piece of equipment.

Any robot—and that includes any moving machine under automatic control—is a potential safety hazard, and needs a thorough safety assessment before being deployed. ( Control Engineering has published many articles about machine safety and risk assessment; find them at www.controleng.com.) Part of the assessment should evaluate whether regular calibration is necessary to prevent a safety hazard.

A third concern arises when recommissioning a piece of equipment after maintenance, repair, or an extended idle period. Always have an experienced test engineer (which includes most maintenance engineers) conduct an acceptance test before letting any such piece of equipment loose in a factory. All kinds of things happen to moving equipment not in regular use, and during repair: Vital components disappear. Nuts, bolts, spacers that look like simple washers, and encoder disks get installed incorrectly. Things simply get bent or broken. Experienced test engineers know how to guard against unknown and unexpected failures on startup.

Simple things like knowing where and how to establish a safety perimeter, checking overhead for what might be damaged should a fire start, and knowing how to react when the unthinkable happens, are part of the test engineer’s mindset. Again, a system calibration check conducted during the acceptance test will uncover any such problems.

C.G. Masi, senior editor

Help support our program

As a high school science teacher who is the program director of my school’s FIRST robotics team, I would like to know how to get in touch with this group [of large industrial companies who support science and math in schools] to help support our program. I also host the FIRST Los Angles regional competition and can never get any of these big companies to step up and support the 60+ teams in the greater Los Angeles area. My e-mail is nmcintyre@chaminade.org .

Nancy McIntyre, nmcintyre@chaminade.org

Safety compliance question

Q : When a SIL4 system is required, and the system comprises various subsystems, is it sufficient to rely on the certification of the [subsystem and component] manufacturers, or does a user have a responsibility to audit the manufacturers in order to establish compliance? Is it necessary to engage an independent third party assessor?

Terry Vetier, taadvet@linkt.com.au Systems Engineer, Australia

A : Generally speaking, safety certified products carry a conformance level certificate issued by an independent third party organization. The safety level of conformance could be SIL, Cat., or a new designation call performance level (PL). The third-party organization (i.e., TUV, BGIA, UL, FM, etc.) evaluates and certifies compliance to standards like IEC 61508, EN 954, IEC 61800, NFPA 79, UL 508, etc. This evaluation and certification process takes months, is quite costly, and is generally accepted as valid.

Considering the application and whether the safety certified products have been applied properly is a separate process. Risk assessment guides and standards typically advise or require that the risk assessment be updated following the application of protective measures (including safety certified products) to establish verification that identified hazards have been properly mitigated to acceptable levels. This usually applies to the application and not the safety certified product used in the application.

Generally speaking, it is acceptable to file the component manufacturer’s safety compliance certificate as part of the overall documents maintained for reference. Additional information can be found on www.SafetyBase.com or you can contact me directly.

J.B. Titus, Certified Safety Professional j.b.titus@siemens.com Siemens, Norcross, GA