Time to Failure Series Continued
Developing Your Motor Diagnostics Program Part 4: Seven Steps 2
Howard W Penrose, Ph.D.
ALL-TEST Pro, A Division of BJM Corp
Step 2 – select stake-holders for your system.
In any program, communication is key. Involve all aspects of the company and your motor system vendors in the development of the program. This involves the true generation of a partnership with the interest of your motor systems in mind.
When you involve vendors, and they know that you are testing and commissioning new and repaired motors, the vendor will pay more attention to what is being shipped to you. The reason is simple: Warranty returns and customer complaints cost the vendor more than identifying problems in the beginning, just as they affect your business. By making the vendor aware of your testing and test criteria, they are more likely to correct any problems before their product arrives at your doorstep. In addition, the vendor should also be able to work with you in the development of a root-cause-analysis.
Communicate training requirements and coordinate between departments. Many excellent diagnostic programs have failed due to miscommunication. Work with the group in selecting and reviewing technologies and testing requirements. Part of any successful program includes the ability to communicate between technicians and technologies in order to provide a comprehensive view of the motor system. (note: there is no ‘Holy Grail’ condition-based monitoring technology, at this time, just CBM systems that are part of the Reliability Professional’s toolbox.
Determine manpower requirements and work with the appropriate departments for the skills. You can provide all of the CBM equipment available, but if it is not used, or used effectively, then it is not a successful program.
Set return on investment requirements and success metrics for the program. Some of these metrics may include fault cost avoidance, repair cost reduction, production equipment availability, reduced cost per unit of production and even energy cost improvements.
Communicate and coordinate findings and corrective actions. This is vital in that if no-one is aware of the successes, then a successful program may be perceived as being unsuccessful. The communication should include all partners in the program as well as others within the company in order to promote the program and its use.
Time to Failure Series Continued
Developing Your Motor Diagnostics Program Part 3: Seven Steps 1
Howard W Penrose, Ph.D.
ALL-TEST Pro, A Division of BJM Corp
The Seven Steps Overview
There are seven basic steps to developing your motor management program. We will cover each of the steps each day:
Step 1: Know Your System
Step 2: Select Stake-Holders
Step 3: Selection of Equipment
Step 4: Training
Step 5: Developing the Program
Step 6: Calculate Return On Investment
Step 7: Promote the Program.
Step 1: Know Your System
1. Do not rely upon perception. It is human nature to associate with what we feel most comfortable with, such as mechanical problems. Or, you may think of the system, such as a critical fan system, while forgetting the fault, such as a winding failure, after time passes.
2. Rely upon examples of problems, including root-cause-analysis reports, both internal or from vendors.
3. Review paperwork, work orders, invoices, repair and supply vendor information for accurate system information.
4. Know the number and types of critical motors in your systems. Basically, what motors will effect production if they fail unexpectedly.
5. Know the total number and types of motors in your plant.
6. Determine the failure modes of the machines and systems in your plant.
7. Determine time for corrective action, repair costs and associated production costs for the critical systems.
8. Review your existing programs for success.
Time to Failure Series Continued
Developing Your Motor Diagnostics Program Part 2: Motor Failure Considerations
Howard W Penrose, Ph.D.
ALL-TEST Pro, A Division of BJM Corp
One of the first things that has to be considered when putting together your motor management program is that your motor system will not ‘instantaneously’ fail. This is a fact supported by consistent research projects on reliability, regardless of any negative commercial suggestions to the contrary. If faults occurred instantaneously, there would be no need for testing, predictive and preventive programs, industrial or reliability engineering, product warranties, assurances that equipment would be able to operate as designed nor would business survive the chaos of systems that failed without warning. In every single case, without fail, there is some condition that leads to a fault.
The good news is that all systems wear out gradually over time. This is viewed in reliability and industrial engineering terms as availability or resistance to failure. As equipment ages, its resistance to failure decreases as does the availability, both of which change as a natural log that consists of time and the mean time between failure (MTBF). This amount never quite reduces to zero until the equipment actually fails.
With all of this in mind, a motor management program can be developed that utilizes motor diagnostics, and other test systems, to evaluate, trend and estimate time to failure.
There are two basic types of motor system failures: Passive and active.
In a passive failure, winding insulation breaks down, winding contamination occurs, rotor fractures or voids and similar faults. As long as the motor is running, these problems may not be obvious until the fault becomes active.
In an active fault, the motor ceases to operate. It may stall or experience a catestrophic failure of some component of the system such that it can no longer perform its job.
The purpose of the conditioned based monitoring portion of your motor management and motor diagnostics program, you will be observing, using instrument measurements, the condition of the equipment. The time to failure estimation begins once degradation is indicated.
The proper implementation of your motor management program will nearly eliminate all unplanned failures, but not all. There will always be problems caused by improper operation, ‘lightning strikes,’ improper maintenance, sabatoge and other un-plannable problems which may quickly degrade equipment condition. These are not high risk conditions, in most environments, and, as such, should not be given high regard when managing your program. You need to consider just those conditions that you can manage.
In the next lecture, we will begin to review the seven steps to developing a successful program.
Time to Failure Series Continued
Developing Your Motor Diagnostics Program Part 1: Introduction
Howard W Penrose, Ph.D.
ALL-TEST Pro, A Division of BJM Corp
Unfortunately, there has been a number of negative discussions concerning semantics by those who do not understand the true purpose of ‘predictive maintenance,’ condition based monitoring, reliability centered maintenance and estimating time to failure. Their purpose is to promote negative attacks on successful technologies. It is disappointing that, for strictly commercial reasons, these few try to convince others that technology should be working backwards towards the dark ages of motor testing.
It is the purpose of this lecture series on developing your motor diagnostics program to provide positive materials on the development and success of programs and the potential to improve your bottom line.
I have been involved in system conditions since starting my career in electric motors as an electric motor repairman in the US Navy on board the USS Theodore Roosevelt, CVN-71 (Plank Owner). My family’s background (mother’s side) in electric motor repair back to 1905 (Great Grandfather and Grandfather with Westinghouse Canada rewinding large hydro-generators), and my father’s work with the environment in St. John’s, Newfoundland then as Environmental Division Department Head at Argonne National Labs, Illinois followed by his work in estimating volcanic eruptions using chemical analysis (including development of the instrumentation), has driven a deep-rooted passion in developing time to failure estimation techniques for electric motor systems.
My work in motor repair research started in 1992 with research on repairing motors for inverter application, testing and repair methodology, followed by motor management development, starting in 1994, and energy research, starting in 1994. The research continued with industrial energy and reliability program development during my time at the University of Illinois’ Energy Resource Center as a Senior Research Engineer and Adjunct Professor of Engineering.
At this point, I was struggling with the systems that were available for motor testing, energized or de-energized. I had negative experiences with motors not starting following surge comparison, and other high voltage testing, in my time in the Navy, as a motor repair specialist and field service specialist, which created awkward situations with customers. There was also something missing in my work with vibration, infrared and insulation to ground testing as I felt that I needed something more to assist in my research. Other than vibration, none of the technologies that I had worked with, nor the technologies that I had reviewed, to date, were able to produce the results that I had been looking for: To provide information that would allow an estimation of time to failure in an electric motor system.
In 1998, I was tasked with developing a program for promoting premium efficient electric motor retrofits for PG&E in California as the recognized motor system support expert for the US Dept of Energy’s Motor Challenge Program (Partner). In this funded project, I recommended the development to be in the direction of electric motor energy and condition analysis. The project manager recommended the review of motor circuit analysis as part of the project which was to include other technology reviews such as vibration analysis, infrared analysis, data logging and the use of the US Department of Energy’s MotorMaster Plus software. Not having much luck with MCA through this time, I returned to users of the technologies that I was aware of, and determined that they were not as effective as hoped. In particular, the issues included user friendliness, size, price and the amount of training necessary to utilize the systems.
In a twist of fate, the president of BJM Corp and owner of ALL-TEST Pro, contacted me for assistance with a technical support question. I got my hands on an ALL-TEST IV PRO 2000 instrument and in testing found that it met the project criteria for size, price and ease-of-use. Testing in the field and in a motor repair shop provided the necessary information to show that it produced the pass/fail criteria for motor condition that we were looking for. It also showed tremendous potential for estimating time to failure in motor windings, cables and rotors. In order to continue my passion for estimating time to failure in electric motors, I joined ALL-TEST Pro with the understanding that I would be able to continue my work in electric motor system reliability, motor-system management and time to failure estimation.
The work has continued with the BJM Corp, Pruftechnik and Dreisilker Electric Motors, Inc. funded modifications to the US Department of Energy’s MotorMaster Plus software program to allow motor circuit analysis and vibration test results to be entered. This allowed the energy and condition analysis and motor management work to continue. The addition of electrical signature analysis increased the ability of the system to perform a complete analysis and data collection (or data logging) for energy and condition analysis of systems. The selection of this system and the following findings finally satisfied the criteria of my career work on developing time to failure estimation capabilities as no other system or technology had succeeded to date. Third party work, comparisons, etc. also confirmed that industry agrees.
Continued work, starting with data collection and test limit analysis, found that MCA has been able to evaluate three phase motors and generators, DC motors, servo equipment, transformers (including transmission and distribution), hybrid vehicle motors and transmissions, wound-rotor and synchronous motors, coils, etc. The work also allowed for instrument users, ALL-TEST Pro research engineers and BJM Submersible Pump personnel to track and trend winding insulation faults and develop a system for estimating time to failure.
This research was absolutely crucial for the development of a successful motor management program. In particular, it’s proven capability to detect early winding degradation that leads to winding faults. This work, research and third party analysis is well-established.
Now that we have established this information, and history, we will begin the discussion of the development of motor management programs.
DEFINITION:
Motor Management: The combination of partnering with vendors and company departments, condition-based-maintenance testing, commissioning, repair standards, spare management, RCM and trending condition in order to reduce motor system related unplanned downtime in such a way that is non-intrusive and provides a significant return on investment.
Tomorrow we will discuss motor system failure and how to manage them.
Time to Failure Series Continued
Electrical Signature Analysis – Part 12: DC Motor Analysis
Howard W Penrose, Ph.D.
ALL-TEST Pro, A Division of BJM Corp
DC motor analysis can be a challenge with ESA, if you do not know what you are looking for. In today’s lecture, we will review a few key points in ESA analysis of a DC motor and drive.
In order to fully analyze a DC electric motor, you still require the ability to test AC voltage and current. When DC is rectified, a small amount of AC remains at the peak of the DC voltage and current, known as ‘form factor’ or ‘AC ripple.’ In an ESA FFT, there will be a dominant voltage line frequency and SCR frequency (# of SCR’s times the line frequency) with harmonics that will taper off. The line frequency harmonics should end by the first SCR frequency. Demodulated voltage and current should show a ripple-type waveform. A hall-effect transformer can be used to show absolute current values. Finally, all test results should be taken from the armature leads only, as the armature contains an AC current during operation, which will allow some load analysis capabilities.
One key difference between AC analysis and DC analysis is that the sidebands evaluated in AC do not exist. Instead, the results will be similar to those found in vibration, requiring FFT capabilities out to 5kHz to detect most faults outside of the motor.
A few DC Drive Faults
One problem that can occur in DC drives is a condition which I will refer to as ‘blow-by.’ In this condition, at least one SCR is shorted so that AC voltage is present on the DC side of the drive. This will result in a strong line frequency peak in current and demodulated current. Also, the line frequency peak will be close to the same value (in dB), or the same for extreme SCR faults, with harmonics across the complete FFT.
Loose connections will show strong line frequency harmonics at a value of approximately -60 dB, or less.
A few DC Motor Faults
A few common problems with DC motors include sparking due to raised mica between commutator bars and worn brushes. In both cases, the resulting sparks cause a raised noise floor around the SCR frequency peaks.
Unbalance will cause a low frequency running speed peak, just the same as found in an AC motor. It is important to note that this will also occur in cases of misalignment.
In our next lecture, we will discuss the analysis of Servo equipment.
Time to Failure Series Continued
Electrical Signature Analysis – Part 11: Driven Equipment Evaluation
Howard W Penrose, Ph.D.
ALL-TEST Pro, A Division of BJM Corp
Homework:
Contact me and request the ‘MCSA Pattern Recognition Guide Excerpt’ for general reading on this topic.
Driven Equipment Evaluation
Driven equipment frequencies can also be detected. We will cover belted, geared, direct drive and fans and impellors in this section.
Belts
In order to determine the frequencies associated with a belted system, there are several steps. In this example we shall identify the motor speed as 1760 RPM (29.33 Hz), one 4 inch (driver) and one 8 inch (driven) diameter sheave with a 40 inch center to center sheave distance.
Step 1: Determine the driven shaft speed by determining the sheave ratios. In this case, it will be Running Speed (RS) * (driver dia/driven dia) = 1760 RPM * (4 inch/8inch) = 880 RPM which is 14.67 Hz.
Step 2: Determine the belt speed by determining the belt length which is equal to: (center to center distance (C-C) * 2) + ˝((driver, inches * π) + (driven, inches * π)) = (40” *2) + ˝((4” * π) + (8” * π)) = 97.28”. Next, the surface (conveyor) speed can be determined by calculating the conveyor speed for either sheave. In this case, we can use the motor sheave and calculate (radius, inches * 2π * RPM) = 2” * 2π * 1,760 RPM = 22,117 inches per minute (IPM) or 368.6 inches per second (IPS). The belt speed can then be determined by taking the conveyor speed and dividing it by the belt length. In this case, 368.6 IPS / 97.28 inches = 3.79 Hz.
Gear Mesh
The driven shaft speed in a geared system is fairly straight forward to determine, as well as the gear mesh frequencies.
The driven shaft speed can be determined by multiplying the driver speed times the ratio of the driver gear to the driven gear number of teeth: Driven = 29.33 Hz * (20 Teeth/100 Teeth) = 5.87 Hz.
The gear mesh frequencies are determined by taking the running speed times the number of teeth. The value is the same for either geer: Gear Mesh = 29.33 Hz * 20 Teeth = 587 Hz. Sidebands around this CF would indicate gear mesh problems.
Driven Equipment – Blade Pass Frequencies
As with calculating the driven speed in a geared system and gear mesh, calculating blade pass frequencies for either fans or impellors is straight forward: The number of blades multiplied by the shaft speed.
In the direct drive system used in the previous examples, the operating speed is 29.33 Hz. A pump with six blades would have a blade pass frequency of 29.33 Hz * 6 Blades = 175.98 Hz.
If it was a fan with 12 blades using the pulley system in the belts section of the previous example, the blade pass frequency would be 14.67 Hz * 12 Blades = 176.04 Hz.
Faulty impellors or blades would be indicated at these frequencies and may include sidebands. The frequencies and harmonics remain the same regardless of how many blades have faults.