Equipment Reliability Institute
ERI News - your reliability newsletter
February 2005 - volume 18

Hello, readers -

The big item in this issue is a very well-written piece by friend Jeff Schutt. Be sure to read Jeff’s first paragraph, at least. The manpower situation he describes may exist in your test lab.

Below that is an introduction to random vibration, by myself. It’s aimed at newcomers to electronics reliability, including aerospace, automotive (and occasionally naval) reliability. What does “random” mean? How is it useful in simulation for environmental test? How is it useful in stimulation for finding weaknesses? Why is multiaxis (at same time) better than historical “sequential axis” testing? Etc. Do you care to discuss these topics by phone or e-mail?

At bottom is another collection of practical test lab advice, by Bob Renz.

In our “short news” section (right column) we announce two new message boards: climatics and electronics cooling. Also upcoming training in data acquisition, in electronics packaging and in vibration and shock testing and related topics. Ask to see Chapter 1 of my new vibration test. Many people make good use of ERI’s links to test-oriented firms; perhaps you can suggest additional links. Some meetings and a new O’Connor text are announced. Sooner + better + cheaper; at best you’ll get 2 out of the 3. Hiring designers.

And don’t forget to join me on Feb 10th for the free “g²/Hz” lecture event starting at noon PST.

Best wishes,
Wayne Tustin

Common Sense Approach for Good Testing
by Jeff Schutt

There are many pitfalls to avoid, especially for the entry level engineer, when planning and conducting tests. Today’s turn-key test equipment enables engineers and technicians, with limited testing experience, to conduct tests relatively quickly and efficiently. However, while computers and menu driven software have enabled testing laboratories to become increasingly more efficient, as far as testing productivity is concerned, a void has been created. This void is the lack of in-depth knowledge of testing that exists in many laboratories. The downturn in military business, corporate downsizing, and attrition have resulted in many of the test industry’s best engineers leaving for greener pastures. Consequently, many entry level engineers and technicians are being thrust into the position of learning and understanding the nuances of testing.

It is one thing to go through the motions and to set-up and program a test. However, it is another thing to understand the rationale for conducting the testing and the details of conducting the testing. It is still another thing to be able to interpret the data and to extract meaningful conclusions from the data. Consideration must be given to numerous variables, in order to conduct an effective test and to obtain the desired results. This article will present these issues and will outline a common sense approach for good testing.

A good test begins at the planning stage. Understanding the test objective, and the selection of appropriate test procedures and requirements, are considerations that should be given as much emphasis as the actual testing itself. Test tailoring, when appropriate, and when permitted, should be employed. Certainly, test tailoring should not be thrust upon a neophyte test engineer. Test tailoring is something that should be left to knowledgeable, experienced testing experts. The testing industry is aware of the fact that most degreed engineers coming out of college are not ideally suited for work in a typical industrial or commercial test lab. Much of the knowledge required to work in a test lab, as is the case with many other jobs, can’t be taught in school. Certainly, there is a lot of on the job training that must be done, and experience that can be gained, in the test lab. Valuable knowledge and experience about a particular test program, or about testing in general for that matter, can be gained by being involved in the test planning process. Some elements of the test planning process are outlined in Table 1.

One could argue that planning for testing is the most important phase of testing. There are numerous considerations that must be made, so that a test can be conducted without incident and as efficiently and effectively as possible. Improper planning for a test could result in costly mistakes, in terms of time and dollars.

A good part of the test planning process is often left to engineering managers and contracts people. However, the test engineer must be kept in the loop on issues pertaining to the test planning process. By keeping engineers in the loop during the test planning process, appropriate feedback can be provided regarding issues such as test feasibility, time, cost, and effectiveness. This feedback loop is extremely helpful to engineering management and contracts people, when planning testing. This feedback loop is also extremely helpful to the test engineer, when preparing for testing.

Once a test has been properly planned, it is important to prepare for the testing. Even before samples for test are received, a great deal of work must be accomplished. It is important that everyone involved understand the test objectives. It is also necessary to determine what data is to be gathered. Data traceability is one aspect of testing that should not be overlooked during the planning and preparation stages of testing. All samples should be uniquely numbered and all data should be traceable to those samples. Data should also be traceable to a particular step or condition in the testing program, during which the data is to be taken. In addition to traceability, the baselining of samples is something that should not be overlooked. Baselining of samples involves, at a minimum, a detailed visual exam, to identify the starting condition of the test samples, prior to the point at which they are placed into test. Baselining of samples can also involve parametric, functional checks of the test sample’s operation and performance. By knowing how a sample performs “Out of the box,” prior to the application of adverse environments, the test engineer can have data that can be used for comparison purposes later on in a test program. Appropriately identifying samples for data traceability and baselining the samples are two important, preliminary steps that are necessary for a good test. Some general parameters that should be recorded for data traceability appear in Table 2.

In order to demonstrate the nuances of a typical test program, a specific example of a vibration test will be utilized. One might think that conducting a vibration test is as simple as following a recipe in a cook book. However, there are so many ingredients that can go into a vibration test. It is critical to specify each of the ingredients (parameters) and to understand when and how to use the ingredients. Some of the technical considerations for conducting a good vibration test are listed in Table 3. Issues such as accelerometer location, fixturing, test levels, monitoring requirements, and others all must be considered, when developing and conducting a vibration test. Failure to specify and understand these issues could result in an improper test being conducted. Vibration testing is especially sensitive to variability from one test to the next, because of issues like accelerometer placement and fixturing. By varying these parameters, a completely different test could be conducted, even if the same test levels are to be applied. It is absolutely essential that the test engineer understand the effect that these parameters can have of the outcome of a vibration test.

The Institute of Environmental Sciences and Technology is in the process of formalizing a number of recommended practices that should help the test engineer better understand the nuances of conducting a good vibration test. While there are many standards and specifications available, which describe various vibration environments, there is a lack of guidance available to the test engineer in the form of recommended practices. These recommended practices are being generated to help test engineers correctly implement the numerous standards and specifications that exist. Some of the recommended practices that the IEST has developed, and is developing, are listed in Table 4.

The above dialog and associated figures present some of the details to which the test engineer must pay attention, for good testing to result. However, there is much more to conducting a good test than what is mentioned above. The test engineer has quite a bit of “behind the scenes” activities, which are also necessary for good testing to result. For instance, every laboratory should have a system of procedures and policies for dealing with the myriad of issues that the laboratory confronts on a daily basis. Procedures and policies for the operation of equipment, calibration of equipment, verification of equipment, maintenance of equipment, data gathering, root cause analysis, and a host of other issues should be developed to help insure that an appropriate testing infrastructure is in place for good testing. A list of typical issues, which laboratory procedures and policies should address, is included in Table 5. These types of procedures and policies help to insure consistency and uniformity in the approach to conducting testing. These procedures and policies also help to ensure that equipment is well maintained and calibrated. These types of procedures and policies are generic enough, so that they apply to almost all testing that is conducted in the laboratory. Many such procedures and policies are part of the ISO 9000 and/or ISO Guide 25 requirements. Therefore, if a laboratory is to be ISO 9000 registered or ISO Guide 25 certified, it is critical to have such procedures and policies in place.

Conducting a good laboratory test takes more than button pushing to spit out a number. It takes engineering talent and experience. It also takes effective planning, diligent set-up and data gathering, and comprehensive reporting. It also requires that a system of procedures and policies be in place and be followed, so that a consistent and uniform approach to testing occurs. In addition to engineering talent, experience, procedures, and policies; the laboratory needs one other key element for successful testing, and that is common sense. Knowing what your test data is telling you, documenting it, and reacting to it appropriately are the key common sense elements to good testing. Engineering talent, policies and procedures are needed to facilitate the use of common sense.

References
• “Considerations for Vibration and Shock Testing of Electronic Assemblies,” Jeff Schutt, Trace Laboratories - Palatine, IL 60067, 847-934-5300, originally published in “Electronic Packaging & Production,” December, 1996, Vol. 36, No. 13, Randolph D. King, Publisher

• IEST-DTE-RP-012 “Dynamic Data Acquisition and Analysis Handbook,” Institute of Environmental Sciences and Technology, 940 E. Northwest Highway, Mount Prospect, IL 60056, 847-255-1561

• “Trace Labs Quality Manual”, June 30, 1998 Rev., Trace Laboratories-Palatine, IL 60067, 847-934-5300

• IEST-DTE-RP-013 “Shock and Vibration Fixturing”, DRAFT, Institute of Environmental Sciences and Technology, 940 E. Northwest Highway, Mount Prospect, IL 60056, 847-255-1561

• IEST-PR-RP-001 “Management and Technical Guidelines for the ESS Process”, DRAFT, Institute of Environmental Sciences and Technology, 940 E. Northwest Highway, Mount Prospect, IL 60056, 847-255-1561

Jeff Schutt at the time of writing this article, in 2002 was General Manager of Trace Laboratories, an independent, diversified, ISO Guide 25 accredited testing facility at Palatine, Illinois. Trace specializes in
providing a wide range of testing services to the computer, military,
automobile, aerospace, telecommunications, medical, and other industries. Jeff is now with Underwriters Laboratories at Northbrook, Illinois.

Mr. Schutt is Past President of the Institute of Environmental Sciences and Technology (IEST). To contact Jeff Schutt, email him at schutt@ameritech.net.

Reprinted with permission from TEST Engineering & Management, December/January 2002-03, pages 8-9.© Copyright 2002, The Mattingley Publishing Co., Inc.

(back to the top)

Random Vibration Testing and Screening
by Wayne Tustin

Let’s discuss (1) some of the engineering situations where in-service random vibration creates difficulties. And discuss (2) how random vibration testing and screening help make hardware (including electronics hardware) become more rugged, more capable of withstanding random vibration in service.

Random vibration aboard rockets
Many early (pre-1955) rocket launch failures have been blamed onto random vibration during launch and ascent. Since 1960, random vibration tests have widely been adopted. Almost all launches nowadays are successful, Figure 1.

Automotive random vibration
At about that same time, the automobile industry discovered that road and off-road inputs to land vehicles were also random. (At the time Figure 2 was taken, the test vehicle was usually accompanied by a cable-connected van full of recording apparatus. Nowadays data acquisition would be little more than a laptop computer on a passenger seat in the test vehicle.)

Let the reader imagine that a triaxial accelerometer is mounted on his automobile’s suspension. Those three signals are recorded, as in Figure 3, the upper three traces (lateral, fore-and-aft and vertical), plotted as acceleration magnitude vs. time.

FFT or Fast Fourier Transformation of those signals results in the three plots grouped at bottom. They show PSD or Power Spectral Density in units g2/Hz. Note that the spectra are fairly flat from about 1 to 20 Hz; we sometimes call that “white” random vibration, equal “power” per unit bandwidth. The spectra extend further, to about 200 Hz.

FFT? Fast Fourier Transform?
Computer use of this transform is a major topic in ERI (Equipment Reliability Institute) short courses and distance learning. We don’t dwell on the details of how a computer transforms time-domain signals (as viewed on an oscilloscope) into frequency-domain information (as viewed on a spectrum analyzer. But we do dwell on the value of FFT to those analyzing vibration and those controlling shakers.

Random vibration testing
Automotive vibration testing typically extends to 200 Hz, whereas aircraft and rocket vibration testing extends to 2,000 Hz. Vibration tests can be combined with climatic (such as thermal) environmental tests by combining a climatic test chamber with a shaker, as in Figure 4. This is about as far as we get on Day #2 of a three-day course. The next material commences Day #3.

Random vibration for ESS
All that testing discussion sets the stage for discussion of HALT (highly accelerated life testing), ESS (environmental stress screening) and HASS (highly accelerated stress screening).

In 1979, NAVMAT document P-9492 demanded in 1979 that electronic equipment suppliers conduct ESS. New shipboard and aircraft gear had too many early-life mortalities. Future gear had to be more reliable in order to (1) increase fleet readiness to fight, to (2) lessen skilled repair manpower needs and to (3) reduce the number of spares. Further, P-9492 promised manufacturers that their at-final-assembly failures would drop, increasing their profits. P-9492 stated that historic “burn in” (hot soak) and single frequency-at-a-time sine vibration testing did not effectively identify incipient failures. P-9492 asked for rapid thermal ramping, alternating with episodes (on shakers) of random vibration; this would be much more effective in identifying weaknesses. At-factory rework and field repairs would be greatly reduced. Those benefits have generally been achieved. By 1981 the US Army and Air Force made similar demands.

Combined environments
ESS was speeded and handling (with attendant chance of damage from frequent connects and disconnects) was greatly reduced by combining thermal ramping with random vibration, as in Figure 4.

Need for Multiple Axes
But not all incipient failures are revealed by such widely-used single-axis electrodynamic shakers. Common practice is to randomly shake ___ minutes in the DUT’s (device under test) X axis, then ____ minutes Y and ___ minutes Z. It is faster, cheaper and more effective to simultaneously shake products in all three axes, as is possible with the three shakers of Figure 5.

Lesser investment
There presently are few labs with multiaxis shake capabilities. A big factor is cost: not only three shakers but also three power amplifiers and three controllers.

Is there a less expensive approach? Fortunately, yes. Consider several pneumatic (operating off plant compressed air, but not synchronized) repetitive shock hammers, attached to the bottom of a softly sprung platform, in Figure 6. All hammers point upward but along various compass directions. The platform to which they are attached can move with six degrees of freedom, N-S, E-W, roll, pitch and yaw.

Figure 6 shows such a platform forming the base of a thermal-ramping chamber. Note ducting that directs blasts of conditioned air (very hot, alternating with very cold) air over or through the flight hardware DUTs.

Unfortunately, there is little control over the resulting vibration spectrum, nowhere near the control possible with electrodynamic shakers. And there is relatively little low-frequency forcing available with pneumatics.

Acoustical forcing (not shown) can shake relatively thin structures, such as printed wiring boards. Spectral control is good.

Random vibration contributes to reliability
We have glimpsed the value of random vibration in testing developmental prototypes (where we try to simulate in-service vibration. We have also glimpsed the value of random vibration in post-production screening of newly produced hardware, in stimulating subassemblies and assemblies

Wayne Tustin, ERI's president, can be reached by e-mail or phone (805) 564-1260. Read more about Wayne at ERI's website.

(back to the top)

Test Lab Musings (part 7)
by Robert L. Renz

Mount one or more white-boards in your vibe lab -they make an easy way to post a shaker schedule or to post set-up data.

Take the time to label every accelerometer lead with the channel number - or else plan to pull and wiggle cables as you try to see which accelerometer goes to which charge amplifier. Electrician’s wire numbers also look a lot better than pieces of masking tape.

Before you take a photograph of the EUT on the shaker for the report, pick up the area first so the photo doesn’t show lab notebooks, test cables, tape, or tools in the background.

Use the “When in doubt, save it” approach when you set up the automatic report format in your shaker software program. Include the run time log, the profile, the shaker data, the channel information, and all the rest that you “never” need for the report.

Robert L. Renz of General Dynamics - Advanced Information Systems at Bloomington, Minnesota.

(back to the top)

Don’t miss the free g²/Hz lecture event on February 10th!

What? “g²/Hz “, a one hour interactive presentation, using PowerPoint® slides and video clips.

Who? Led by “Mr. Random Vibration”, Wayne Tustin.

Why? Because relatively few people understand the common PSD unit of measure, g²/Hz.

Where? On your computer monitor or (if a group) your video projector.

When? Commencing at noon PST (1:00 MST, 2:00 CST, 3:00 EST), Thursday, February 10th.

Click here to find out how to participate.

Have you questions about humidity (condensing or non-condensing)? Temperature? Altitude? Salt fog? Fungus? About measuring any of the "climatic" environments? Or about conducting "climatic" environmental tests?

Direct your questions (or observations) to ERI's Climatics message board. Sometimes the Climatics board moderator, Steve Brenner, himself, will
assist you. Steve is ERI's
climatics test specialist.

Plan 1: Let's have Chuck Wright come here to teach a few days.

Plan 2: Let's send individuals to Chuck's Bohemia (Long Island) New York course March 7-9.

If we train 8 or more, plan 1 is less expensive and it's focused on our needs. If we only train 2 or 3, plan 2 is less expensive, and our people will learn outside class from their classmates.

If you would like to request a free sample of Chapter 1 - "What are vibration and shock?", from Wayne's new book "(...) Random Vibration and Shock Testing", please visit our website. Fill out the quick form and submit it to us. We will then e-mail you a PDF file of Chapter 1.

Someone once called ERI the "Yahoo.com of environmental testing". He was referring to our links to other businesses. Please visit our links page, where you will see a listing of categories, arranged alphabetically, from "Accelerated Test Consulting" to "Vibration Test / Shaker Systems".

Clicking on "Commercial Environmental Testing Laboratories", for example, will take you to a listing of company names, firms that offer testing. Clicking on any of these names will take you to that firm's home page. Some of these firms operate numerous laboratories.
When you contact these firms, please mention that you found them on ERI's links page. If you'd like to ask for someone by name, drop an e-mail to ERI, or phone. If you happen to know of a firm that should be listed here, please tell ERI.

ESTECH 2005
The 51st Annual Technical Meeting and Exhibition, Estech 2005, will meet on
May 1-4, 2005, at the Hyatt Regency Woodfield, in Chicago.

CEEES Conference
"Methods and Benefits of Environmental Testing and Engineering" will meet in Germany, on May 11 and 12, 2005.

New Book released
The "New Management of Engineering" book from Patrick O'Connor is now available. Visit his website for more details.

"Modern vehicles are loaded with microprocessors, CAN networks, sensors galore and software ties it all together. Not all of this software runs safety-critical systems like the anti-lock brakes or stability control system. Some of it does pretty mundane stuff like controlling the sound system, or heated seats. In my case, my Chevrolet truck has a "memory" feature that remembers the seat position, mirror placement, radio settings, and the fan speed and temperature on the climate control.

Trouble is, this whole system routinely goes wonky, smashing me and the seat up against the steering wheel and going brain dead on the other settings. Chevrolet has only recently, sheepishly revealed to me that there is a software bug somewhere in this system, and as of September 2004, no workaround has been identified. For now, I'm stuck with a seat with bugs that routinely needs to be rebooted. Funny, huh? Good thing it's not a safety-critical system."

Editorial in COTS Journal, November 2004.

Where to find good designers? Some of our best designers have a test background ...... they've worked as test engineers, deliberately applying various stresses to hardware. They've seen hardware fail. Many times they can look at DUTs and see weaknesses ..... can predict how and where failure will occur.

Designers are well advised to ask test lab people to comment on their new designs, well before building a prototype.

Design managers are well advised to consider experienced test engineers (as opposed to recent graduates) when design openings occur.

ERI - Equipment Reliability Institute
1520 Santa Rosa Ave.
Santa Barbara - CA - 93109
Tel: (805) 564-1260
Our fax number:
(805) 966-7875

Wayne Tustin tustin@equipment-
reliability.com

Webmaster webmaster@equipment
- reliability.com

Websites
http://www.equipment-
reliability.com

http://www.vibrationand
shock.com

Copyright © 2000-2004 Equipment Reliability Institute. All rights reserved.

Subscribe
If you would like to subscribe to ERI News, go to either website, fill in the form "Free Newsletter" and hit the Submit button. Subscribe now!

Recommend
If you enjoy reading ERI News and want to recommend it to a friend, just hit "forward" on the menu of your e-mail program or tell your friend to subscribe at our website.

Previous issues
Missed the previous issues? It is not a problem. Just visit our newsletter archives section and find all ERI's News issues.

Unsubscribe
If you do not want to receive ERI's quarterly newsletter, please reply to the newsletter message with "remove" as subject or click here.