JOURNAL OF COMPUTER SCIENCE AND ENGINEERING, VOLUME 13, ISSUE 1, MAY 2012 29
Analysis and diagnosis of the different defects of the bearings and gearings by vibration analysis Abdelkader Benbouaza, Bachir El Kihel and Fabienne Delaunois Abstract— To detect the different shortcomings of the rotating machines, the conditional maintenance by vibratory analysis is the best tool of surveillances and diagnosis. The bearings and the gearing are among the mechanical organs the more used in most industrial domains, they are conceived to function on very long terms, it remains true that one there ready hardly of attention in a lot of cases. In conditions of ideal or bad use, the deterioration due to the fatigue of the rolling, a bad lubrication provokes several shortcomings. These shortcomings are the reason of a great deal of breakings of machines. In order to solve this problematic, a test bed is finalized in our laboratory, it allows us to do an analysis separate or combined on the different deficient bearings and gears. Index Terms — strain; bearing; gearing; vibration analysis; Conditional Maintenance.
—————————— u ——————————
BEARING 1
T
Introduction
ODAY, more and more manufacturers are turning to vibration analysis to obtain an increased availability of their machines: they control the actual state of their equipment (rotating machinery) to avoid stoppages due to unplanned outages. Bearings or gears element bearings are critical components of rotating machines located between the fixed and movable parts of the structure; they ensure the transmission of forces and the rotation of the shaft. But if they are the most common, these components are also the most fragile. It must be said that the different elements that constitute them (the balls, cage and rings) are constantly sought. The repeat pass balls on tracks causes high mechanical stresses that lead gradually degraded by metal fatigue, and eventually cause the appearance of point defects (cracking, spalling, ..). Another cause of damage, defects and assembly design. Indeed they cause overloads that lead, ultimately, premature wear of the bearing. Finally, like any system accuracy, turnover is sensitive to a dusty environment, the presence of cuttings or impurities in the body system,
lubricants polluted or damp conditions ... are external agents that can cause to extensive damage. Vibration analysis is a powerful tool for detecting these defects. It may be a lack of cage, ball, indoor track or outdoor track. These defects are characterized by different repetition frequencies depending on known parameters[1]. The correct positioning of sensors depending on the type of assembly and turnover is paramount to ensure a reliable and significant. Degradation of a bearing can occur in two stages [2]: -Localized defects, chipping vibration behaves such as trains of shocks, but the low vibratory energy is measured while the value of the peaks is important. - Generalized defects, wear, geometry defects, the vibration has significant peaks and vibratory energy also important.
2
Different types of bearing failures [3]: Ø
————————————————
• Abdelkader Benbouaza is with industrial engineering Department,National School of Applied Sciences University Mohammed 1, BP669, OUJDA 60000 MAROC • Bachir El Kihel is with industrial engineering Department, National School of Applied Sciences University Mohammed 1, BP669, Oujda 60000 Maroc • Fabienne Delaunois is with Metallurgy Department University of Mons Street of Epargne, 56 Mons 7000 Belgique Fbienne.
Ø Ø
© 2012 JCSE www.journalcse.co.uk
Foreign material: one of the most common sources of problems in the bearings to wear and punctures by foreign particles. Incorrect mounting: it must respect the rules of editing and adjustments for each type of bearing. Misalignment: a frequent source of problems causing overload.
30
Ø
Ø
Ø Ø Ø Ø
Electrical damage: when electric currents pass through a stage there is an arc and burns at points between races and rolling elements or current jumps the gap of air. Inadequate lubrication: absence, lack or poor choice of lubricant causes overheating or excessive wear in the bearing. Default sealing: it allows foreign particles and contaminants entering the bearing and causing
wear. Fatigue. Corrosion. High température.
3 Methode for detecting defects: There are many technical means of achieving a diagnosis of bearing condition. Among them: - Spectral analysis: has become the fundamental tool for processing vibration signals because it is much easier to work in the frequency domain than in the time domain. Thus was born the idea of the frequency analysis (spectral) where the signal amplitude / time is converted to signal amplitude / frequency using the Fourier transform of equation 1. This representation allows to know the spectral content of energy or power, present in the signal at frequency f, and thus to detect the presence of a defect shock generating periodic at a frequency of defects [4]. Equation 1
- The envelope detection [5] (or "amplitude demodulation"). Its principle is to filter band pass in the time signal, and then perform the Fourier transform of the envelope of the signal obtained. The method thus allows overcoming the low frequency interference signals issued by other defects of the machine (unbalance, misalignment, etc...), and to keep only the signals emitted by the defects of the bearings (high frequency). Then it only remains to find the characteristic frequencies of defects on the spectrum of the envelope of the filtered signal. "This method is useful for making a diagnosis at an early stage, and to highlight defects generally difficult to observe (when the signal strength is low and distributed over a wide frequency range). [6] It also helps to determine reliably and quickly repetition frequencies of shocks, even in the case of periodic shocks embedded in noise background. "However, it requires knowing the frequency domain of interest for filtering, and it is not suitable in the case of speed or variable loads. Finally, other techniques such as frequency analysis, it shows its limits in the low rotation speeds [7].
Fig1. principle of the envelope analysis
-The cepstral analysis. In this case, we observe the lines present on a particular spectrum while the cepstrum. Often called a "spectrum of the spectrum", the cepstrum is used to highlight the periodic phenomena behind a spectrum. It is therefore very efficient to find and follow the evolution of repetitive impacts, such as those created by spalling of the bearing. But the vibrations are not the only phenomena to betray the presence of defects on a
roll. Some manufacturers make direct impacts caused by the repeated passage of the balls on faults [8]. Fig2. principle of the cepstral analysis
4
Characteristic frequencie:
In most cases, degradation results in chipping a track or a rolling element bearing, producing a shock each pass [9]. These shocks cause a vibration type pulse that propagates on the ring, then the outer cage of the bearing [10]. These pulses become the source of amplitude modulation of the random signal of vibration. They are characterized by a very steep climb and a very short time and therefore there will be an increase of the RMS value of vibration signal and the crest factor. Defective bearings generate vibration frequency equal to
31
the speed of rotation of each part of the bearing. They correspond to the rotation including logs, rollers and the cage or the passage of the balls on the rings. For each bearing type and according to their production sizes can be considered characteristic frequencies formulas given below [11].
Fig5. Bearings controlled
7
Fig3. Frequency characteristics of the defects of elementsof a bearing
4.1 Note The headways are theoretical frequencies substantially complied with as long as there perfect rolling without slipping of the ball. As soon as the seize starts, the friction tends to slow the game deteriorated. Failure manifests itself at a lower frequency than the frequency calculated. Although this difference is difficult to understand a critical element to establish the urgency of surgery, the seizure being the ultimate stage of a defect that could cause bearing and by blocking the sharp deterioration of the rotor [12].
5
With the help of a vibration analyzer we performed various tests of spectral analysis and envelope analysis on three 6005-2Z bearings (bearing breast; bearing with default on the inner ring and the third; contains a defect on the outer ring) and we observed spectres for each test. The raised measurements are radial measurements and the bearing changed is RLT2. Characteristics of the 6005-2Z bearing. N: = number of balls 10. DR: pitch diameter = 36 mm. DB: ball diameter = 6mm. Default track external BPFO = 194 Hz. Default inner track BPFI = 274 Hz.
Description of the test
To make a study of the various defects that may have a bearing on (default cage, ball, indoor track or outdoor track) we conducted a test in our laboratory. The bench consists of: - Single phase electric motor 0.35 KW; 2800rpm. - bearing 6005-2Z. - Elastic coupling.
Fig4. Picture of the test bench realized
6
Realised tests
Bearing control
To create defects; on the two bearings 6005-2Z we conducted a groove on each ring bearings. The groove formed on the outer ring is greater than that on inner ring.
Fig6. layout of the test bench realized
7.1 Envelope Analysis 7.1.1 Bearing breast
32
7.1.2 Bearing with default on the inner ring 7.2.2 Bearing with default on the inner ring
7.1.3 Bearing with default on the outer ring 7.2.3 Bearing with default on the outer ring
7.1.4 Notes For breast-bearing, there is some appearance of peaks at frequencies which are not significant (18.8 Hz, 259.4 Hz ...) with very low amplitudes. -In default of the bearing with the inner ring, there is time to peak of the rotation frequency (46.9 Hz), the peak frequency of the defect of the inner ring 279.1 Hz, and harmonies surrounded the peak the default inner ring spaced from the rotation frequency. - For the bearing with fault on the outer ring, there is appearance of the peak frequency of the defect of the outer ring 193.8 Hz 387.6 Hz and multiple.
7.2 Spectrale Analysis 7.2.1 Bearing breast
7.2.4 Note -The amplitude of the signals of the outer ring fault is larger than the two others.
8
Conclusion:
The bearing is the very important element in the construction of machines and more particularly in rotating machines. Its rapid wear is very expensive for our companies. To make a diagnosis on the state of bearings, we must look first of all these characteristics and types of mounting. According to tests conducted we conclude the following: - Envelop aanalysis is better than spectral analysis for the detection of bearing defects. - The amplitude of the signals measured increases when the defect is large. When we have a rotating shaft assembly: - The inner race defects are subject to modulation of the speed of rotation of the shaft, where a large numbers of sidebands of the higher the frequency of the defect. - The frequency of outer race defect excited when a roll-
33
ing element passes over a defect. Therefore appear only BPFO.
Gearing 1
Introduction
A gear is composed of the set of two toothed wheels meshing with each other, for transmitting power between two shafts close with a constant gear ratio.
Fig7. Characteristic frequencie of the gearins
Fe= Fr1. Z1 = Fr2. Z2 Fe: Couple Frequency Fr1 and Fr2: frequencies of rotations of the wheels 1 and 2 Z1 and Z2: Number of teeth of the wheels 1 and 2.
4 2
Main gear damage [13]: Among the most gear defects encountered: Ø
Pitting: The pitting is a destruction of surface fatigue often not exceeding 0.3 to 0.5 mm in depth. The pitting is not dangerous for a gear. It does not warrant any reversal or replacement. However, it can be slowed not a choice most suitable lubricant. However, creep-related landslides, it will facilitate the deformation profiles. The quantification of the deformation profiles is essential
Ø
Scaling: Scaling is a dangerous damage which representation is the removal of metal plates. This is related to excessive pressure often related to poor load distribution.
Ø
Seizure: The seizure is a damage linked to a metal / metal contact during meshing of material causing a tear. Cracking or breakage of teeth: treatment of a crack is needed to stop its propagation. The urgency of the repair work depends on the ratio of conduct that only a specialist can calculate the gears.
Ø
3
Description of the test
To make a study of the various defects that can have a gear (deterioration of a tooth pitches enough ...) we conducted a test in our laboratory. The bench consists of a: - Single phase electric motor 0.25 KW, 1400 RPM. - Bearing 6005-2Z. - Elastic coupling. - 3 gables.
Fig8. Picture of the test bench realized
5
Gearing control
For our test we used two gears, in one breast and the second has tooth deteriorated.
Couple Frequency [14]:
Observe a gear comprising two gear wheels 1 and 2, Z1 and Z2 with teeth and rotating frequencies Fr1 and Fr2. Each time a tooth of the drive wheel engages in a driven wheel 2, there is a periodic charging socket at the rate of engagement of the teeth at a frequency Fe is equal to the meshing frequency of rotation of the wheel multiplied by its number of teeth.
Fig9. Bearings controlled
6
Realised tests:
We used the same analyzer of vibration in a breast gear and deteriorated gear; we observed spectre for each test. The measurements are radial measurements. Fe= Fr1. Z1=22,33x36= 840Hz.
6.1 Breast gear
34
- For a deteriorated gear there is appearance of the frequency couple and its multiple plus the frequency of rotation. - To a gear pitch insufficient; there appearance of the frequency couple and its multiple, the peak Fe2 is the greatest.
Acknowledgments If the tooth is correct, the spectrum is made up of components whose frequencies match the engagement frequency and its harmonics.
The authors would like to acknowledge the partial support of the laboratory GIMP ENSAO and the department metallurgy of Mons University.
6.2 Deteriorated tooth:
REFERENCES
If one wheel has a tooth deteriorated, there is a hard shock, each turn of the pinion. Shows the corresponding spectrum of a comb lines whose pitch corresponds to the frequency of rotation of the pinion deteriorated stretching to high frequencies.
6.3 Insufficient center-to-center:
If space between the shafts carrying the two wheels is insufficient, it occurs at the bottom of meshing teeth. The driving tooth forces to release the engagement and which causes wear on both sides of the teeth. In spectrum analysis this phenomenon results in the appearance of a line amplitude at twice the dominant meshing frequency Fe and the partial loss or total amplitude at this frequency.
7 Conclusion:
- Spectral analysis is the best tool for faults gear detection. - For a healthy gear there is appearance of the frequency couple and it’s multiple.
[1] G. COLLIENNE, “Comparison of a sensor sound and an acceleration sensor in the vibration part of the maintenance2004”. [2] Technique de l’ingénieur [3] Roland BIGRET Jacques-Louis FERON “diagnostic-Maintenance availability models of rotating machines - measures - vibration analysis” MASSON 1995. [4] Philippe “numerical and experimental methodological approach to aid in the detection and monitoring of vibration of chipping defects of ball bearings” University of Reims Champagne Ardenne Faculty of Natural Sciences16 décembre 2004 (Thesis ESTOCQ). [5] Marie-Line Zani “bearings, components to be monitored meadows” Article preventive maintenance Avril 2003. [6] Jacky DUMAS & Bruno BENNEVAULT “START MEASURING VIBRATION”01dB-STELL (Groupe MVI technologies). [7] Ing. X. MILISEN Dr Ir V. LENAERTS “Comparison of different methods for detecting faults in bearings”. ISILF n°20 2006 (Review of Scientific). [8] Pr. EL KIHEL bachir Pr. ELAMRANI yassine “During vibration control” National School of Applied Sciences Oujda. [9] [12] [14] Landolsi Foued “During technical surveillance of the major faults”. [10] Arquer P, “Predictive diagnosis of machine condition” Masson 1996. [11] Benllali. Yacine, Hadjadj. Aoul .Elias, Khalifa. Dalila.Thesis “projection conditional preventive maintenance bearings by vibration analysis”. [13] Catalogue société NOVEXA “vibration analysis of gear fault”. First A. Author Abdelkader Benbouaza engineer research in industrial engineering on the theme maintenance of the complex instalations industrial by vibration analysis. Second B. Author Bachir El Kihel Professor of Higher Education Grade: C, head of department math and mechanical. PhD status in Structural Design. PhDUniversity of Compiègne engineering school France in Vibration and cotrol non destructive. Third C. Author Prof. Fabienne DELAUNOIS Associate ProfessorHead of Service Metallurgy University of Mons.
