|
Equipment
Reliability Institute |
|
Do you measure static or dynamic deformations (changes in dimension - strain)? Quite likely you bond strain gages to the structures you investigate. Larry Shull warns readers about some of the ways to get wrong data. The reader's question: How do I test a shelf item? We invite our readers to keep on sending their questions. One of our specialists or Wayne will respond to them. Best wishes, |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ******************************* | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Reliable
Strain Gage Installation
During the heyday (1960-1973) of the Saturn/Apollo Program there was a term used and understood by all involved, FIVE NINE RELIABILITY. Interpretation: each and every part on any component of the entire Saturn assembly was proven to be 99.999% reliable under full scale loading conditions. This, in effect, means that 99,999 times out of 100,000 tests each part had functioned as designed. Of course, every part was not tested that many times, but its performance was proven to meet these standards. This proven reliability was the major reason for the superb performance of the entire Apollo program. Quality control was about hardware, not about paper work as it is today. Five nine reliability is a thing of the past, based on material and equipment available on today's markets. Automobiles must be worked on before they can be driven off the show room floor; new toasters, out of the box, fail to perform when plugged in for the first time; buy five shirts off the rack, when opened at home four have defects, some serious enough to return the shirt. This holds true for almost anything on the market, mechanical, electronic, material, and so on. We have generations of younger people who accept this as the "norm", never having experienced anything better. Poor performance is expected from almost any endeavor. This attitude is carried to the test beds of mechanical items. Failure of sensors used to determine performance is shrugged off as "nothing unusual!" If a strain gage, or other type transducer fails to provide the expected answer the sensor is assumed "bad". Until the introduction of the computer into the mechanical test world, the strain gage was considered the most precise component in any measurement system. The manufacture of these tiny sensors was tightly controlled and installation on a test item was a thing of great pride for the technicians involved. Strain gages have performed flawlessly in many different environments, static, dynamic, deep cryogenic temperatures, outer space, and so on infinitum. The secret of this success has always been the great care taken in selecting the proper gage, adhesive, protective coatings, solder, wiring, and most important, installation techniques. Properly installed strain gages will survive almost any environment short of catastrophic hardware failure directly beneath the gage itself.
Reliability of the strain gage is directly related to techniques used to install each gage. There are no short cuts. Certain rules apply to each individual gage as it is removed from the package, prepared for bonding, positioned, clamped and cured. Preparation of the surface where the gage is to be bonded must meet certain standards to assure a good bond. It must be smooth, free of any scratches, nicks, other disfiguration, and almost medicinally sterile. The adhesive to be used must be fresh, properly mixed and settled. It should be applied to both the bond side of the gage, as well as the position where the gage is to be located. Air dry criteria must be met before the gage is placed in position. If a strip of tape is used to handle each gage it must be of a certain quality; any old tape will not do. There are other similar rules that must be observed if the gage system is to perform as expected. Violations, even minor things, can lead to disastrous results. As an example, at one time a very good technician decided he would concoct his own solder flux, and did so with out telling anyone. Things went well as long as his gages were used within a short time after installation. Came the time the gaged part was put in storage for about six months, waiting for test equipment to be fabricated. Eventually all was ready, the test item was on the test pad, strain gages and other sensors connected and checked. At the time of installation all gages, approximately 300, had checked out at the normal 350 Ohms. Now these values varied from zero to infinity, and anything in between. This is a classic example of what corrosive flux does to electrical circuits over time. Hundreds of similar examples could be cited, extending over the past 30 years or so. Inexperienced and untrained personnel, trying to install gages have accounted for a good number of such instances. Excellent technicians have also contributed, as in the example, trying short cuts, not paying attention, and using chemicals on new materials, without first checking for compatibility. Some nstallation Procedures are absolutely wrong, but must be adhered to because that is what Quality Assurance uses to certify the situation. There is another situation that arises from the instrumentation used to condition and record the information from a strain gage. A properly installed gage will respond, almost to the molecular motion of the atoms, to anything that occurs in the surface immediately beneath the gage grid, integrating the motion over the grid length. No motion outside this area will be monitored. The gage will respond to any strain, be it axial, bending, torsion, temperature effects in the gage, thermocouple effects in the solder joints, magnetic effects on gage and wiring, and so on. All of these things cause changes of resistance in the gage, which has no way of knowing what, where, when, why, and who caused the change. Methods have been devised to sort things out, canceling unwanted information, while enhancing wanted information. To do this type compensation requires considerable knowledge of, not only the gage behavior, but how the part will behave under given loads, what temperatures are expected, and any other effects mentioned. A thorough and complete understanding of the Wheatstone bridge is necessary. Thus, strain gage systems can be designed and installed that will produce valid information required by Engineering, while eliminating unwanted signals. This, as a rule, requires four (4) gages to make one measurement. Such methods were used successfully for years, with excellent results. With the introduction of the computer into the mechanical test world, engineers wrongly decided that arithmetic will provide all the right answers without "wasting" the extra gages. Computing bridge circuits are rarely used today, especially in stress analysis where they are badly needed. The problem, as mentioned above, is that the gage responds to any and all environmental changes. Some of these can be sorted out by using sum and difference methods as would be the case for axial and bending loads applied to a part having at least three gages located around the part at a particular station. However, if the test item is a cylindrical vessel, loaded in compression, and gages are located only on the outside, or only the inside, bending strains can directly cancel compression strains under the gage. Another situation is caused by strain gage temperature changes during mechanical loading of the item. Strain gages are excellent temperature sensors, and unless this is considered during data reduction, serious errors can occur. The point of this short dissertation is that strain gages, properly installed, are five nine reliable in any situation short of catastrophic failure of the test item immediately beneath the gage grid. Poor installations result in inexact data and failed gages. Not understanding gage behavior in all environments can result in improper data interpretation. Not accounting for temperature, or other extraneous effects, can also lead to misinterpretation. The strain gage itself is rarely at fault. Larry Shull is the president of Strain Gage Advice, Inc. and also one of ERI's specialists. Please click here to access his bio. Also, Larry provides a short course in Engineering Applications of Strain Gages. Click here to see the course outline. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ******************************* | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Fifth Edition
- Harris' Shock And Vibration Handbook This fifth edition supersedes the previous four editions, the first having appeared in 1961 and subsequent editions in 1976, 1988 and 1996. Gaps between editions have been 15, 12, 8 and now 6 years; one might infer that changes in dynamics are occurring more frequently as the decades pass. This edition is coauthored and edited by Allan G. Piersol of Woodland Hills, California.
Chapters 1 through 9 deal with basic theory and have only minor editorial changes, additional references and/or revised references. At least minor changes were made in all other Chapters. Major changes to individual Chapters are indicated in the Chapter listing which follows. Some of the Chapters by deceased authors were revised and updated by the editors, retaining author credits in recognition of the authors' significant contributions to shock and vibration technology.
I'm sure your favorite engineering bookstore can obtain the Handbook for you. Or send an e-mail to customer.service@mcgraw-hill.com. Or FAX: the McGraw-Hill Bookstore Attn: Mail Order Department, 212-512-4105. Ask for: Harris' Shock & Vibration Handbook 5/e ISBN: 0071370811 $150.00. On a personal note, I had the honor and pleasure of knowing Charles E. "Charlie" Crede in New England in the 'fifties and later in Southern California in the early 'sixties. In 1962, on Charlie's recommendation, I invested in the three-volume Harris & Crede Edition 1. Charlie's autograph (in Volume 1) is dated October 1962. Unfortunately, Charlie died shortly thereafter. Fast forward 22 years. I carried that Volume 1 to a New York City meeting with Dr. Cyril M. Harris who, on September 16, 1984, added his autograph. Now, treasurers of first editions, what am I offered? Wayne and publisher Eve Mattingley hope to meet many TEST readers at the Annual Technical Meeting of the Institute of Environmental Sciences and Technology, April 28 to May 1, 2002, at Anaheim, California. Wayne will convene a meeting of Working Group DTE013 on "Vibration and Shock Test Fixturing", Sunday, April 28 at 3 pm, location to be determined. On May 1, Wayne will speak on "Planning your Vibration/Shock Test Facility". That is part of chairman Jeff Schutt's session on "Test Facilities". Also on May 1, Wayne will co-chair (with Karl-Friedrich Ziegahn from Germany) a session on "Test Methods". Details of the IEST meeting may be found at http://www.iest.org/estech/estech.htm. Wayne Tustin, ERI's president, can be reached at tustin@equipment-reliability.com or at 805/564-1260. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ******************************* | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Questions our readers have asked... This section of our newsletter was created for you, reader! Feel free to send questions or suggestions to the webmaster. They will be either responded by Wayne Tustin or forwarded to one of our specialists. Here is the question for this issue: A: Your only recourse is to clamp your product onto the vibrating table of a shaker or onto the platform of a shock test machine. See Figure 1.
Often the "clamped to" element is an intermediate aluminum or magnesium flat plate, perhaps an inch thick, which in turn is bolted to your shaker or shock test machine. Why? Because the existing mounting points on your shaker or shock test machine are probably not at the locations you need for your particular DUTs (devices under test). You will need to drill and tap holes, typically 3/8-24, at appropriate locations close to where the DUTs will be located. The vertical members, Figure 1, can be several suitable lengths of "running thread". Cut these from 3/8-24 or other threaded rod . Tighten a 3/8-24 nut on each vertical member, atop the lower horizontal bar (if you use it - see later note) so that the lower bar and vertical member remain attached to the intermediate plate. The horizontal members, Figure 1, might be ½ inch thick x 1 inch wide bar stock. Whereas Figure 1 shows only one overhead "holdown" bar, larger DUTs might require two overhead bars. If the bottom surfaces of your DUTs can lay flat on your intermediate plate, you might not require the lower bars. Atop the "holdown" bar, tighten 3/8-24 or other nuts, so that the DUTs don't move while the test is underway. Wayne Tustin, ERI's president, can be reached at tustin@equipment-reliability.com or at 805/564-1260. |
Are
you a test engineer? Are you involved in shock and vibration testing
and measurement? I think you'll be benefit from hearing about the
new fifth edition of the famed 
We
will not go into the quality control during the manufacturing of
strain gages. It will suffice to say that inspection is 100% visual,
and a resistance check. Other parameters are determined statistically,
under tight control. Every strain gage sold is guaranteed out of
the package. What happens to it during or after installation is
the user's responsibility. 
My
purpose in writing this review is to honor the editors/organizers/compilers
(as well as the authors). What a monumental task, to coordinate
the work of 53 authors! Pages are only numbered within sections.
They add up to 1416 pages, including a 231-page index. 
|
Last chance
to enroll!
|
|
|
|
Contaminants
and Moisture can disrupt your electronics
|
|
|
New date for Optimizing Electronic Circuit Card course
|
John Starr, a specialist from ERI, will teach "Electronic Circuit Cards - Applied Vibration" in August 19-21, 2002. Click here to register for the course. |
|
Vibration and Shock courses coming up
|
Wayne Tustin will teach the following short courses in vibration and shock measurement, analysis, calibration, testing, HALT, ESS and HASS: Santa Barbara, California, February 11-13, 2002 Huntsville, Alabama, Livonia, Michigan, Baltimore, Maryland, Newport, Rhode Island, May 22-24, 2002 Santa Barbara, California, Additional courses: |
|
New web
sites coming up
|
|
|
|
Message
Board
|
|
|
|
Announcements
|
|
The 20th Aerospace Testing
Seminar Wayne's COTS presentation Fuel Cell Technology Seeking teacher |
| Think about |
Don't Let Your Career Depend Upon Your Employer Do you need some specialized, career-advancing training? And been told by your boss that "There's no money in our training budget."? Your career and your lifetime earnings are too valuable to let that deter you. Tell (don't ask) your boss that you will be away from ____ to ____, participating in that training, at your own expense. THAT will get his attention + the attention of higher management. |
| Check our Glossary |
Check our Vibration and Shock Glossary. We have just added new words and their definitions. This list evolved from Wayne's 50 years of work experience and it is constantly updated. |
|
Contact
information
|
|
Wayne Tustin tustin@equipment- Webmaster webmaster@equipment Web sites Copyright © 2000-2001 Equipment Reliability Institute. All rights reserved. |
|
Free
Newsletter
|
|
Recommend Format Previous issues Unsubscribe |