DE10236735A1 - Motor vehicle vibrating test bed control signal generation method is such that, the damage causing conditions of the test bed closely match those during actual driving - Google Patents

Abstract

Method for generating a power density spectrum for a test-bed that is used for vibration testing of motor vehicle components, has the following steps: a load collective is generated during driving, from an acceleration time signal; from the load collective a power density spectrum is generated; the acceleration time signal on the test-bed is measured; a test bed load collective is determined from it; the two load collectives are compared to obtain a measure of existing damage or to determine the time before a given amount of damage occurs during driving.

Description

The invention relates to a Method for generating damage equivalents for driving Noise profiles for vibration testing of Vehicle components.

When driving a vehicle, the accelerations are measured with which individual vehicle components cause vibration vibrations are stimulated and which a certain damage, hereinafter referred to as the damage measure, on the vehicle component. there it should be noted that this damage is of course the Do not impair the function of the vehicle. One test, one damage caused with the same degree of damage, can be put to the test carried out become. For this, a shaker is controlled with a noise signal, that vibrates the vehicle components. The on the test bench resulting accelerations can be measured and evaluated.

Such a method of generation a noise profile is already known from WO 98/14765. To This method initially defines a frequency range for which the Noise profile should be created. Then power density spectra (PSDs) using Fourier transforms determined from the data of the driving operation. From these power density spectra becomes a single power density spectrum in the form of a staircase function generated that then smoothed becomes. With the help of a material Wöhler line and a frequency response determined using an FEM calculation for the Test optimal power density spectrum calculated.

This method is disadvantageous first, that the power density spectrum for testing on the test bench only very complicated from certain parameters from Wöhler lines and FEM calculations that usually either not available or very difficult to carry out, or can be determined can be determined. Second, with one Fourier transformation for each each discrete frequency only z. B. an average or maximum value of Amplitudes calculated the actual Distribution of the amplitudes is lost. This method described leads to relatively inaccurate power density spectra.

The object of the invention is a method specify with which a control signal for a test bench can be generated, which leads to essentially the same degree of damage as while driving.

This object is achieved by a Method according to claim 1 solved. Advantageous further developments are the subject of the dependent claims.

According to the invention, it has been recognized that the following structure is possible for a method for generating noise profiles which are equivalent to driving operation for vibration tests for controlling a test bench for motor vehicle components:
From the acceleration time signal of a vibration excitation measured during driving operation on the vehicle component to be tested, a driving operation acceleration load spectrum is first calculated at least for a defined frequency segment in a predetermined frequency range, indicating the frequencies of the magnitudes of the amplitudes of the acceleration time signal. In order to obtain a degree of damage on the test bench that essentially corresponds to the degree of damage from driving, several frequency sections with the same defined width can preferably be formed. This is possible by a prior narrow-band filtering of the acceleration time signal from driving. The driving mode acceleration load collectives from driving mode are then calculated by classifying the filtered signal, preferably a range pair classification. At the individual reference points in the frequency range, which result from the filter width and separate the frequency sections from one another, their frequency is thus specified as a function of the size of the acceleration amplitudes.

A power density spectrum for controlling, for example, a shaker, which is an element of the test bench, is then generated from the driving operation acceleration load spectra. A power density spectrum is formed as the control signal, for example from the maximum and minimum values of the amplitudes of the respective load spectra in the associated frequency sections. A separate calculation is carried out for each frequency section. The following formula shows a possible calculation of the power density spectrum, where fb stands for the filter bandwidth: PSD = ((⏐a Max ⏐ + ⏐a min ⏐) / 2) 2 / (10⋅f b ) [g 2 / Hz]

The shaker or the like is controlled with this control signal. The acceleration time signal generated on the test bench on the vehicle component to be tested as vibration excitation is measured. This measured acceleration time signal is filtered with the same filter as the acceleration time signal from driving. A test bench acceleration load spectrum is calculated from these filtered acceleration time signals in the predetermined frequency range for each defined frequency section indicates the frequency of the magnitudes of the amplitudes of the measured acceleration time signal.

Then a degree of damage is determined, that preferred after that later described method is calculated. Now through relationship building of the determined degree of damage Vibration excitation of the vehicle component from driving and of the determined degree of damage Vibration excitation of the vehicle component from the test bench operation the needed Test time to achieve the degree of damage driving on the test bench can be calculated. When the shaker is activated for the duration the calculated test time with the power density spectrum above all defined frequency sections in the specified frequency range away, turns up on the test bench essentially the same degree of injury that occurs while driving.

Alternatively, the test bench can also be controlled with a predetermined target test time, which is the same for all frequency sections, with which the same degree of damage occurs on the test bench as in driving operation for all frequency sections in the predetermined frequency range. In order to obtain this new power density spectrum, the power density spectrum calculated from the driving acceleration load spectrum has to be calculated with a factor x which can be calculated from the target test time t, which _{is the} same for all defined frequency segments in the specified frequency range _, and which can be used to calculate different test times t _fi for each frequency segment , be multiplied.

The degree of damage to the vehicle components can preferably be from the driving operation acceleration load collectives and the degree of damage to the vehicle components from the test bench acceleration load spectra be calculated using a suitable method. A fundamentally known to the expert possibility represents the damage accumulation hypothesis according to Palmgren mines.

To the power density spectrum yet The method can preferably be adapted more precisely to the driving operation performed iteratively become.

In the drawing is a preferred one embodiment presented the invention. It shows a simplified block diagram an embodiment of the invention

From an acceleration time signal measured while driving 1 the vibration excitation is after a narrow band filtering 2 with a filter bandwidth of, for example, 5 Hz over a predetermined frequency range (for example from 5 Hz to 500 Hz) for each frequency section with a width of 5 Hz, a driving mode acceleration load spectrum 3 determined. This shows the frequencies of the magnitudes of the amplitudes of this frequency range at the support point (5 Hz, 10 Hz, 15 Hz, ...) corresponding to the respective frequency section. In position A of a switch 4 becomes a power density spectrum 5 from the maxima and minima of the individual driving mode acceleration load collectives 3 created. With this power density spectrum 5 is in the test bench 6 controlled a shaker and the acceleration time signal generated on a vehicle component spanned in this test bench 1' the vibration excitation measured. From this acceleration time signal 1' again after a narrow band filtering 2 ' with the same frequency bandwidth for each frequency section cut a load spectrum, namely the test bench acceleration load spectrum 3 ' , created.

Then the switch 4 placed after B. For example, with the help of a standard Wöhler line or the component or material Wöhler line according to the miner elementary method known to the person skilled in the art, the degree of damage is determined for each frequency section 7 the driving operation acceleration load spectrum of the vibration excitation and the degree of damage 7 ' the test bench acceleration load spectrum of the vibration excitation is determined. The two damage measures are compared in a comparator 9 the test time required to achieve the degree of damage 7 generated from the driving operation by the quotient from the damage measure 7 from the driving operation and the degree of damage 7 ' is calculated from the test bench.

So that the test time is the same for all frequency sections, a target test time 8th specified to which the power density spectrum 5 should be adapted for all frequency sections. In a computing unit 9 a factor x is calculated for each frequency segment in the defined frequency range, with which the power density spectrum 5 must be multiplied to get a target power density spectrum from all frequency ranges 5 ' to generate. The factor x is calculated from the root of the quotient from the calculated test time to achieve the required degree of damage 7 for each frequency range t _fi and the desired test time t _to:

The process can iteratively loop for a very precise power density spectrum, which achieves the damage level from driving even more precisely 10 run through. It can ever However, a large number of further details may be designed to deviate from the above description without departing from the content of the claims.

Claims (5)

Method for generating a power density spectrum for a test bench for vibration tests on vehicle components, with the following steps: a. From the acceleration time signal measured on the vehicle component to be tested during driving ( 1 ) A vibration excitation load spectrum is generated at least for a defined frequency section in a predetermined frequency range ( 3 ) formed, the frequency of the magnitudes of the amplitudes of the acceleration time signal ( 1 ) indicates. b. From the driving operation acceleration load spectra ( 3 ) becomes a power density spectrum ( 5 ) to control the test bench ( 6 ) generated. c. The on the test bench ( 6 ) Acceleration time signal generated on the vehicle component to be tested ( 1' ) the vibration excitation is measured. d. From the on the test bench ( 6 ) acceleration time signal measured on the vehicle component to be tested ( 1' ) a test bench acceleration load spectrum is created for each defined frequency section in the specified frequency range ( 3 ' ) formed, the frequency of the magnitudes of the amplitudes of the measured acceleration time signal ( 1' ) indicates. e. For each defined frequency section in the specified frequency range, a comparison of the previously determined degree of damage ( 7 ) the vehicle component from the driving operation and the previously determined degree of damage ( 7 ' ) the vehicle component from the test bench operation the test time required to achieve the damage level ( 7 ) while driving on the test bench ( 6 ) calculated. Method according to claim 1, characterized in that for a degree of damage ( 7 ' ) the vibration excitation of the vehicle component on the test bench, which is identical to the degree of damage ( 7 ) the vibration excitation of the vehicle component should be in operation, the test bench ( 6 ) with the calculated power density spectrum ( 5 ) is controlled with the test time calculated for the respective frequency section. Method according to claim 1, characterized in that given a predetermined target test time (same for all defined frequency sections in the predetermined frequency range) 8th ) the test bench ( 6 ) with a target power density spectrum ( 5 ' ) is controlled, whereby the target power density spectrum ( 5 ' ) from the power density spectrum ( 5 ), the respective test times of the individual frequency sections and the target test time ( 8th ) calculated. Method according to claim 1, characterized in that the degree of damage ( 7 ) the vibration excitation of the vehicle component from the driving operation-acceleration load spectra ( 3 ) and the degree of damage ( 7 ' ) the vibration excitation of the vehicle component from the test bench acceleration load spectra ( 3 ' ) be calculated. Method according to claim 4, characterized in that the degree of damage ( 7 ) the vehicle component from the driving operation-acceleration load spectra ( 3 ) and the degree of damage ( 7 ' ) the vehicle component from the test bench acceleration load spectra ( 3 ' ) can be calculated using the miner elementary method against a standard, material or component Wöhler line.