Vibration Tests: a Brief Historical Background

Transcription

1 Sinusoidal Vibration: Second Edition - Volume 1 Christian Lalanne Copyright , ISTE Ltd Vibration Tests: a Brief Historical Background The first studies on shocks and vibrations were carried out at the beginning of the 1930s to improve the behavior of buildings during earthquakes. With this framework in mind, M.A. Biot defined the shock spectrum to characterize these phenomena and to compare their severity. The term shock spectrum has since been changed to shock response spectrum (SRS) in order to avoid any confusion and to clearly show that it characterizes the response of a (linear with one-degree-offreedom) system subjected to the studied shock. Vibration tests on aircraft were developed from 1940 to verify the resistance of parts and equipment prior to their first use [BRO 67]. Such tests became necessary as a result of: the increasing complexity of on-board flight equipment which was more sensitive to vibrations; improved performance of aircraft (and, more generally, of vehicles), to the extent that the sources of vibration initially localized in engines became extended substantially outwards to the ambient medium (aerodynamic flows). The chronology of such developments can be summarized as follows [PUS 77]: 1940 Measurement of resonance frequencies. Self-damping tests. Sine tests (at fixed frequency) corresponding to the frequencies created by engines running at a constant speed. Combined tests (temperature, humidity, altitude).

2 346 Sinusoidal Vibration The exciters which were used at the time were mechanical and the vibration was created by the rotation of off-centered mass. Shock machines, of standard impact, were developed shortly after. The table, guided by vertical columns, fell into a tub filled with sand. The shape of the shock created during the impact could be selected by fixing pieces of wood of particular form under the table of the machine The first electrodynamic exciters were developed [DEV 47] [IMP 47]. Their limited power made it possible to carry out only tests of sinusoidal vibrations. At this time, the first standards were written and used for the acceptance tests carried out on each material. The measured vibratory environments being in general of random type, the standards quickly evolved towards swept sine tests which made it possible to cover a broad range of frequencies in spite of the limitations of the exciters Swept sine tests were introduced to simulate variations in engine speed, or to excite all of the resonance frequencies of the test item, regardless of its value. Test severities resulted from measurements of the real environment taken on a category of carriers. The measured signals were filtered using square filters and the largest peak of the response of the filters was drawn on an amplitude-filter central frequency diagram. The group of points thus obtained was largely enveloped by straight line segments in order to define a swept sine test, with constant displacement at low frequency, then eventually with constant velocity and finally with constant acceleration. Thus, the standards proposed swept sine tests, which are often still specified today in certain documents. It was, however, understood that it would be better to apply random vibration tests, and it was attempted to specify swept narrow band random vibration tests, broad band random vibration tests not being possible because of the lack of power of machines. All of these studies were essentially completed for military applications. Similarly to today, shocks carried out on the shock machines were limited to simple shapes: half-sine wave, square (or trapezoidal) shock and terminal peak sawtooth shock. For convenience, and in order to reduce costs, the possibility of creating shocks directly with an exciter was studied. With the

3 Historical Background 347 test specimen remaining on the same machine for both shocks and vibrations, it was possible to gain much time Specifications and tests with random vibrations (introduction of jet engines, simulation of jet flows and aerodynamic turbulences with continuous spectra). These tests were highly controversial until the 1960s [MOR 53]. To overcome the insufficient power of such installations, attempts were made to promote swept narrow-band random variations in the frequency domain of interest [OLS 57] First publications on acoustic vibrations (development of jet rockets and engines, effect of acoustic vibrations on their structures and equipment) First acoustic chambers [BAR 57] [COL 59] [FRI 59] The specification of random vibration became essential and the possibilities of an exciter were sufficient to carry out broad band random vibrations. Studies were carried out to determine equivalences between random and sine vibrations. Missiles and also space vehicles and satellite launchers use many pyrotechnic devices, which enable them to use very precise time slots during the operation of equipment (separation between propelling stages, firing of an engine for example). These devices contain small amounts of explosives which generate very short, but locally very severe, high frequency shocks, which are propagated in the structures while attenuating and combining with the response of structures. The frequency contents of these pyrotechnical shocks thus increases with frequencies closer to those of the equipments. Their amplitude remaining still significant, these shocks can produce important malfunctions. In the 1960s some publications reported the new interest in these shocks, which were often regarded as not very severe because of the very high frequency. Following incidents, a very large number of works were published in the early 1980s, and this interest has continued until today, both to measure shocks, study the propagation, to attenuate or to filter them mechanically and to take them into account in the softwares used for the dimensioning of parts. With acceptance tests arriving late in the design/production process, in the event of problems with the behavior of the materials, it was preferred in around 1960 to carry out qualification tests before beginning the production of the products, using standards still defined without reference to the real environment.

4 348 Sinusoidal Vibration 1965 J.W. Cooley and J.W. Tukey s algorithm for calculating FFTs [COO 65]. Although the spectrum of shock is still criticized and not used to specify the shocks in the standards, it was very useful for severity comparison of several shocks in the absence of a more powerful tool. Some first attempts were made to try to control the exciters directly starting from a shock response spectrum, in order to be able to simulate shocks for which the SRS is difficult to reproduce starting from a simple shape shock Increasing number of publications on acoustic vibrations Tri-axial test facility [DEC 70]. Development of digital control systems The use of standards that superficially recreate the environment sometimes led to the creation of products which were too large for their environment, or sometimes to imaginary problems the material being designed more to resist the qualification tests than to resist real conditions of the environment. It was in addition often necessary to reduce the mass of the material to the maximum. It was thus necessary to dimension the material to resist, with a certain margin, its real conditions of use. This remark was at the origin of the development of a method transforming and epitomizing measurements of the mechanical environment into test specifications expressed in a simple form and with a reduced duration in order to reduce costs. This procedure implies: determination of the life profile of the products; searching for measurements corresponding to each condition of the identified environment; then, the synopsis of all the data collected in order to calculate the simplest possible specification from it, with a small number of tests of reduced duration if the real environment is of long duration; finally, the tests thus determined must be organized in order to ensure the best representativeness of the tests with the lowest cost.

5 Historical Background Extreme response spectra and fatigue damage spectra developed; useful in writing specifications (a method in four stages starting from the life cycle profile). Equivalence necessary during the synopsis is based on two criteria: the reproduction in the tests of the largest stresses created in the product when it is in its real environment (except duration reduction) and the reproduction of the fatigue damage related to a large number of stress cycles undergone by the material. These two criteria are the base of the extreme response spectra and fatigue damage spectra developed around 1975, unifying the methods of shock and vibration analysis. The application of this method supposes the exploitation of many measurements and the realization of calculations, which led to the development of software running under Windows and, associated with databases, under Unix Account taken of the tailoring of tests in certain standard documents (MIL- STD-810F [MIL 97] [GAM 92]): development of specifications on the basis of measuring the real environment. Versions of the standards MIL STD 810 D in the USA and GAM EG 13 in France, then NATO standards, have themselves evolved in this direction in years 1980/1985, requiring the test tailoring. However, only standard GAM EG 13 proposes and describes in its technical appendices the method by the equivalence of damages. At that time the MIL STD 810 standard explicitly authorized the use of the shock response spectrum to specify shocks Taking the environment into account in the project management (according to the R.G. Aero Recommendation). Test tailoring makes it possible to demonstrate during the qualification tests that the developed product will be resistant to its future real environment. These tests arrive late in the event of failure, since they oblige a resumption of the design of the object. This is why in around 1990 the concept of tailoring the product to its environment was introduced, which encourages taking into account the real environment through a step of tailoring at the very beginning of the project.

6 350 Sinusoidal Vibration Historical background. Overview of the main developments in the field of vibrations, shocks and standardization of tests