Ben Shank, an engineer at Thermotron Industries, reveals how the company’s new classic shock methods for crash testing, could prove more controllable and realistic in comparison with IEC specifications
The engineer reviews recent efforts to make classic shocks used in crash testing more controllable and realistic in comparison with IEC specifications
What is the classic shock method?
Classic shock pulses are used on electrodynamic shakers (as well as with other technologies) to model rapid impacts such as door slams and vehicle crashes. Unlike a crash sled, which can bring the whole assembly up to speed slowly before impact, shakers are limited to a few inches of total displacement. This means forces must be applied before and after the main pulse to keep the armature close to center. Unfortunately, these pre- and post-pulses can significantly affect the damage response of the device under test as measured by the shock response spectrum (SRS).
What are the issues associated with this?
Depending on the resonant frequencies of the device under test, which are often unknown, many of the pre-main-post pulse sequences in use right now represent significant over- or under-tests compared to the ideal reference. Sometimes this is necessary given the shaker system involved, but often we conduct these less realistic tests out of habit. Those who cannot readily adopt a pre/post-pulse optimization method can gain much of its benefit simply by being aware of the problem. Especially when the specification does not tax the shaker system, selecting the reference with the highest allowable displacement will (usually) greatly improve fidelity to the specified SRS damage.
How have you made classic shocks more controllable and realistic?
First we need to define ‘controllable’. Any spec that calls for a classic shock pulse will specify a percentage of the peak acceleration that the pre- and post-pulse must not exceed. Automotive specifications tend to refer back to the IEC 60068-2-27 spec, which calls for +/-20% limits on both the pre- and post-pulse. If anything outside the main pulse is beyond those limits, that shock is technically out of spec.
So we want to back off on the pre- and post-pulse acceleration, both for realism and controllability. But that inevitably requires higher displacement, which eventually creates its own control problems. Shortly after joining Thermotron full-time, I wrote a kinematics script to optimize the pre- and post-pulse acceleration for any specification by considering both the requested pulse and the limits of the shaker. Although it started as a learning tool, we are releasing a controller upgrade this summer with adaptive shock pulses integrated into pulse selection. This relieves the shaker operator of having to decide which of multiple references is best for their specific combination of pulse and shaker system.
Is this method of testing easy to conduct?
Shaker testing is very easy to conduct compared with crash sleds and drop tables. Often the most time-intensive part of a shock test is bolting the device to the fixture. (This is not to say that setting up a shaker for realistic testing is trivial, but that tends to be done once and an expert can be called.)
Tests can be set up through an intuitive user interface or simply opened from a saved file. Once set up, the shaker performs a series of small pulses to determine the frequency response of the whole system and then can perform one pulse or thousands of pulses in quick succession without further input from the operator.
The new methods proposed require either that the test creator balance closeness to the ideal SRS against shaker limits or use an algorithm to do the balancing. Adaptive shock and similar algorithms should be essentially transparent to the operator.
What is the next step for this work?
We are constantly learning new things, and no doubt the adaptive shock algorithm will steadily improve, but mostly the next step for this work is to use it. The number of electrodynamic shakers in use in the automotive industry, as well as other industries, is rising steadily.
We want to make sure that novice and veteran users are performing the best tests they can with as little headache as possible. Professional pride is certainly at play here but, more pragmatically, we all want to live in a world where competitively-priced products survive the environments for which they were designed.
November 2, 2016
