Methods for testing noise of machinery and equipment--vibration method for measuring noise

1. Introduction

The most common method for measuring the noise of machinery and equipment is to measure the sound pressure level with a sound level meter. However, in many occasions, this method that people are very familiar with seems powerless. For example: in the machine room where multiple machines are running, when it is necessary to measure the noise of each machine; or when it is necessary to detect the noise of each product one by one on the production line of finished products, the sound level meter cannot display the noise directly radiated by the measured product due to the influence of other sound sources and the introduction of reflected sound. With the development of science and technology, people naturally think of the sound intensity method. However, the test instrument of the sound intensity method is currently expensive, and the test is more complicated, and it is still in the research stage. Therefore, people have carried out research on the vibration method for the test of sound waves. It is hoped that the amount of noise radiated by the machine can be determined by measuring the amount of vibration on the surface of the machine, which is usually called the vibration test method of air noise. The results of many years of theoretical analysis and application research have shown that this is a very simple and effective method. Under very harsh environmental conditions, it is almost unaffected by environmental noise and reflected sound. A special weighted vibration meter can be used to determine its noise radiation value by measuring the amount of vibration on the surface of the machine. At present, this method has been successfully used in actual production.

The use of vibration measurement method to test the noise of products at the production site was proposed when other methods could not solve the problem of on-site noise detection of products in a simple, rapid, economical and accurate manner. West Germany, the United States and other countries have been conducting research on this technology for many years. The German BBC company spent more than ten marks on the research of vibration method and successfully applied this technology to on-site noise detection of contactors. After years of research, the United States has stipulated in the Navy MIL standard that the vibration method should be used to measure the noise of micromotors. The International ISO Organization for Standardization has published standard technical documents on vibration measurement method.

China began to explore vibration measurement method in the late 1970s. After more than ten years of experimental research, it is clear that in order to obtain the practical application of the vibration method, the following six technical problems must be solved, namely:
(1) The actual radiation efficiency index curve of each electromechanical product must be obtained;
(2) The weighted network curve of the instrument must be made according to the radiation efficiency index curve that changes according to the sound source size;
(3) The calibration of the instrument and the reference value of the decibel quantity must be solved;
(4) The key measuring points of the surface vibration of each machine must be determined;
(5) The problem of aerodynamic noise superposition and correction must be solved;
(6) For the detection on the "assembly line", the problem of simplifying the measuring points must also be solved.

Through experimental research on motors, electrical appliances, and refrigerators, the above problems have been solved and corresponding instruments have been developed.

The VIB-4 computer vibration and noise measuring instrument belongs to this type of instrument.

2. Basic principles

Sound is the result of mechanical vibration. When an object has mechanical vibration within the sound frequency range, it will cause the surrounding medium to vibrate accordingly, thereby radiating sound outward in the form of sound waves. The radiation of sound waves is essentially the process of energy transfer of mechanical vibration waves. Therefore, the premise of studying vibration sound radiation is first the problem of wave form.

When an infinitely large plane plate vibrates, the sound waves radiated should be plane waves. In an ideal plane wave sound field, the sound intensity is uniform. The relationship between the effective value of the sound intensity at any point and the effective value of the sound pressure and the effective value of the vibration velocity at that point is a relatively simple linear relationship, namely:

Where I -- sound intensity W/m2
　　 V -- vibration velocity mm/s
　　 P -- sound pressure Pa
　　 PC -- acoustic impedance Rayleigh (sound ohm)

If the adjacent flat plate medium vibrates with the flat plate, the vibration velocity of the plane can be regarded as the sound wave vibration velocity of the adjacent medium, from which the near-field sound intensity or sound intensity level can be directly obtained, and the sound power level can be calculated.

However, the vibration of actual machines is rarely flat and wavy. From the perspective of the far field, many small-sized electromechanical products can be approximately regarded as radiating noise outward in the form of spherical waves. Therefore, the theoretical analysis of spherical waves is closer to reality.

Obviously, the sound intensity of the sound waves radiated by a pulsating ball is related to the vibration intensity (V), the wave number (a function of frequency and sound velocity), the size of the sound source, and the vibration form m of the pulsating ball. In addition, it also reflects the radiation efficiency of the vibration sound. The vibration of the actual machine is relatively complicated, but as long as the radiation efficiency curve can be found. According to the formula, the sound intensity level (LI) can be obtained, and then the sound power level Lw can be calculated according to the area of ​​the surface where the sound intensity level is measured.

3. Determination of actual radiation efficiency

The noise radiated by any electromechanical product after operation comes from multiple vibration sources and has a wide frequency characteristic. Most electromechanical products are a combination of multiple vibration sources vibrating at various frequencies. The overall vibration mode is very complex, so it is difficult to carry out pure theoretical derivation. In addition, the noise of the near field and far field of the machine surface will be quite different, which makes it more difficult to derive the radiation efficiency index of the actual product in theory.

However, a large number of test results also reflect that the vibration and acoustic characteristics of electromechanical products do have a strong regularity. First, the regularity of the radiated sound field. Under certain far-field conditions (the distance specified by the test method standard), small-sized products all have the characteristics of spherical sound wave radiation. Even the test results of larger-sized products are consistent with the far-field test results. Therefore, their noise radiation laws can be attributed to the sound radiation type of the pulsating ball. The modal analysis results of the motor clearly show that under a single modal frequency, the vibration of the motor is very close to that of the pulsating ball, but most of them vibrate according to the second and third order vibration modes. [page]

A large number of experimental studies have been conducted on motors, electrical appliances, and refrigerators to obtain the actual radiation efficiency index of these products. The experimental research procedure is roughly as follows:
(1) Select a series of typical prototypes according to different size structures, such as motors with bases from 80 to 280, and refrigerators with several volume structures.
(2) Determine representative vibration areas and corresponding typical measurement points on the surface of the prototype (motor, refrigerator, electrical appliance), such as 6-7 points for motors, 4-5 points for electrical appliances, and 29 points for refrigerators. At each measurement point, measure the surface band vibration level and A-weighted vibration level according to 1/3 octave (33 frequency bands) and A-weighting. At the same time, measure the 1/3 band sound power level and A-weighted sound power level of each prototype according to the anechoic chamber precision method of international and Chinese standards. This is generally called the anechoic chamber sound pressure level measurement method.
(3) Statistically analyze the surface vibration measurement results and the results measured by the far-field sound pressure level method in the anechoic chamber (according to 1/3 octave band and A-weighting) to obtain the actual radiation efficiency curves of various products.

After statistical analysis of more than 100,000 data, we obtained a very regular 10lgσ curve, as shown in Figure 1. The radiation efficiency index curves (curves 1, 2, and 3) in the figure are highly correlated with a large number of measured values. The correlation coefficients of the three products are all above 0.9, 0.998 for motors, 0.989 for electrical appliances, and 0.903 for refrigerators, which fully demonstrates the regularity of the actual radiation efficiency of electromechanical products. It also proves the feasibility of determining the noise power level by measuring the surface vibration of electromechanical products.
In the actual radiation efficiency index curves of several electrical products: curves 0 and 1 (solid lines) are the radiation efficiency index curves of the pulsating ball at the 0th and 1st order vibrations, respectively;
　　　curve 2 (dotted line) is the radiation efficiency index curve of the motor.
　　　Curve 3 (double dashed line) is the radiation efficiency index curve of the electrical appliance.
　　　Curve 4 (dashed line) is the radiation efficiency index curve of the refrigerator .

Comparing several curves, it can be seen that the curve of the motor is closest to the pulsating ball, which is almost between the 0th and 1st order theoretical curves. The similarity can also be confirmed from the modal vibration analysis of the motor.

IV. Practical application results

According to the actual radiation efficiency index curve, a weighting network is made, added to the amplification circuit of the vibration meter, and then the A weighting network specified in the sound level meter is added. In this way, the vibration meter can directly measure the noise level of each product. Since the curve is obtained based on the relationship between the surface vibration level and the noise power level obtained in the far field, the correction for the near field test has been introduced into the curve. Therefore, since the average value of the surface vibration velocity measured by the vibration meter with the weighting network added, plus ten times the logarithm of the surface area, can be used as the far-field noise power level of the product, corresponding to sound sources of different sizes, the radiation efficiency index weighting network can be adjusted according to the size. In this way, the application range of the vibration noise detector is wider.

This instrument is used to verify and test several products:

the basis of the comparative test is based on the calculation results of the sound power level measurement using the national standard or international standard of the corresponding product in a precision semi-anechoic chamber. There are two states for measuring the sound power level of the machine by the vibration method. One is the comparative test under the same operating state in the anechoic chamber, and the other is the comparison of the results of the on-site vibration method measurement on the production line with the results of the sound pressure method measurement in the anechoic chamber.

The statistical analysis and calculation of the actual verification results show that the machine sound power level measured by the vibration method has a strong correlation and high accuracy with the measurement results of the sound pressure method in the anechoic chamber. The statistical analysis and calculation results of several product verification tests are listed in Table 1

Table 1 Statistical analysis and calculation table of the measurement results of the vibration method and the sound pressure method in the anechoic chamber

Product category Correlation coefficient of the two methods r Average difference between the two methods Standard deviation of the difference between the two methods σn-1
Motor 0.999 0.015dB 0.65dB
Electrical appliances 0.996 -0.53dB 0.9dB
Refrigerator 0.892 0.67dB 0.98dB

The actual verification results are sufficient to prove the vitality and extensibility of the vibration method. If other types of products can obtain the corresponding actual radiation efficiency index curve, the application scope of the vibration method will be wider. The

verification results of random batch sampling also prove that the test accuracy of the vibration method can be guaranteed and is quite high. Of course, it mainly depends on whether the regularity of obtaining the actual radiation efficiency curve is strong and accurate.

The advantage of the vibration method is that it can avoid the influence of background sound and reflected sound during air noise testing. Therefore, for on-site noise control, the noise of each machine can be measured and analyzed when several machines are running at the same time. Therefore, it is also suitable for the load noise measurement of the motor. In particular, the test method of load noise will be developed in the direction of difference superposition method in the future, making the application of vibration method more advantageous. I believe that further application will further prove the superiority of this method.