Consider any electronic or mechanical piece of equipment intended for military use. Stating that the equipment is “rugged” is not sufficient. There’s no neat category that defines what rugged means. It has to be quantified in terms of the laboratory test intensities at which the equipment has continued to operate satisfactorily. This means the whole gamut of climatic environments including high temperature, low temperature, humidity, altitude and depth. Also important are the dynamic, man-made environments of vibration, shock and sound pressure.

All those potentially damaging climatic environments “come at” hardware simultaneously from all directions. Take temperature, for example. In the “real world” and therefore in the lab, most if not all surfaces are warmed simultaneously or cooled simultaneously. Take altitude for another example. In the real world as in the lab, most if not all surfaces are exposed to partial vacuum simultaneously.

Last-Century Methods: Sequential Axis Shaking

The problem today is that while climatic test is by nature simultaneous, testing of dynamic vibrations and shocks often isn’t. Why should procurement agencies continue to condone last-century testing? Why should laboratories continue to waste time and money shaking test hardware first in its X axis, then its Y, then finally in its Z axis? That common sequential-axis testing, familiar to generations of test engineers and technicians, is time consuming and labor intensive. It’s detrimental on a number of levels. It means paying for three tests and paying for three fixtures. Further, it necessitates much potentially harmful handling. Worse, sequential-axis testing is not as effective nor as quick at finding product weaknesses as simultaneous multi-axis testing.

Two words—“simultaneous” and “multi-axis”—are very important when specifying a dynamic (vibration and shock) test for equipment intended for use aboard military and commercial land, sea and air vehicles. The new Test Method 527 (multi-exciter testing) in the late-2008 “G” revision to the venerable MIL-STD-810 was overdue, but certainly is welcome.

Land Vehicle Simultaneous Multi-Axis Shaking

Only single-axis-at-a-time shaking was possible with mechanical shakers prior to 1950, limited typically to 10 to 55 Hz. Wider frequency range (typically 10 to 200 or to 500 Hz) EH or electrohydraulic (servohydraulic) shakers also are single axis. However, automotive test engineers long ago combined three or more EH shakers for multi-axis shaking, replicating, for example, damaging railroad transport inputs. So did seismic test engineers, replicating multi-axis earthquake inputs to buildings.

In Figure 1 are multiple EH shakers creating realistic simultaneous multi-axis railroad transport vibrations on a laboratory vibrating platform. Nearly every automobile manufacturer has such a vibrating platform. Why? Because new automobiles suffered railcar-induced damage (surprises) en route to dealer showrooms. It is far cheaper to find railcar-induced weaknesses before a new model automobile goes into production.

Nearly every automobile manufacturer also has a setup in which multiple EH shakers drive the four wheels in a manner that represents various road inputs that relate to various road and off-road conditions and various vehicle maneuvers at various speeds. It is far better to find roadway-induced weaknesses before a new model automobile goes into production. Reductions in warranty expense far more than pay for testing.

Higher Frequency Simultaneous Multi-Axis Shaking

Testing to 2,000 Hz is desired for aircraft and missile hardware, also for engine-mounted hardware. This necessitated the development of ED or electrodynamic shakers. In operating principle these resemble electrodynamic loudspeakers. ED shakers are driven by power amplifiers under specialized computer control.

At relatively few U.S. military establishments, three or more ED shakers have been on-site combined for simultaneous multi-axis shaking. Figure 2 was taken at the Army Research Lab, Adelphi, Maryland, after two ED shakers were added. Earlier, with just one shaker, some field failures could not be replicated in the lab. After adding two more shakers, those field failures were replicated. More recent multi-exciter ED shaker systems include White Sands Proving Ground in New Mexico, Hill Air Force Base in Utah (Figure 3) and Keyport Naval Undersea Warfare Center in Washington State. Experience at those facilities led to the new Test Method 527 (Multi-Exciter Testing) mentioned earlier.

The system at Hill AFB in Utah involves eight ED shakers.  The four vertical-thrusting units provide thrust-axis translation, also pitch and yaw, to the aerospace load above them. The two pairs of horizontal-thrusting units (one shaker omitted) provide vertical and lateral translation. Note that each pair is deliberately misaligned in order to provide rotation to that load.

At those military facilities, individual ED shakers were contractor on-site combined, at considerable engineering development and much labor expense, over many months.

Multi-Exciter Systems

At least two Japanese firms are supplying factory-assembled arrays of three ED shakers to Japanese automobile manufacturers. One such system, Figure 4, was purchased by Spectrum Technologies, a commercial environmental testing laboratory at Redford, Michigan; their multi-axis shaking service is being used by Detroit-area firms that provide on-engine electronics and other hardware.

Over-the-road vibration data is acquired from three accelerometers, one sensing fore-and-aft motions, one sensing left-right motions, and one sensing vertical motions (almost always the most severe. Alternately, one three-axis (or “triax”) accelerometer can be used. X, Y and Z accelerometer signals are recorded, later edited and still later fed to a specialized computer that controls the motions of the X, Y and Z shakers that together drive hardware being tested. A three-shaker system will be approximately three times the price of a one-shaker system. But finding just one weakness in the lab, rather than having an influential customer find it later, in service, can more than pay for the system.

Is there a less-expensive way to get simultaneous multi-axis vibration? One approach is to use multiple inexpensive pneumatic hammers, on various compass headings, angled up into a softly sprung rather flexible platform that forms the bottom of a thermal chamber. Devices to be tested (DUTs) are mounted on to the top of that platform. They thus receive not only multi-axis vibration but also varying-temperature stress. A major drawback is lack of vibration intensity control, in the wide variation in vibration intensity received by the units being tested.

Breaking Some Hardware

Rugged hardware is not developed overnight. Early units must have failed, in service or in the test lab. Root causes of those failures must have been sought, found and eliminated, as proven by subsequent tests. Much can be learned from failures. Just passing a spec or a standard is not sufficient. The industry needs to go further and be surprised. It is far better to be surprised in the lab than in the field, perhaps in combat.

It costs little to “go beyond the standard.” When a system developer has already paid for the use of the shaker, paid for the fixture and installed your DUT (device under test) in the fixture, on the shaker, why not apply more force? Or extend the frequency range? Perhaps both, until something fails? Not only will root cause failure analysis identify the weak link, but it will also provide the data on the intensities at which the hardware survived.  

Equipment Reliability Institute
Santa Barbara, CA.
(805) 564-1260.
[www.equipment-reliability.com].