Вы находитесь на странице: 1из 15
Vibration Analysis of Rotating Machinery This example shows how to analyze vibration signals from a gearbox using time-synchronaus averaging and envelope spectra. These functions are especially useful in the predictive ‘maintenance of gearboxes, which contain multiple rotating components: gears, shafts and bearings, ‘This example generates and analyzes vibration data for a gearbox whose shafts rotate ata fixed speed. Time-synchronous averaging is Used to isolate vibration components associated with a specific shaft or gear and average out all other components. Envelope spectra are especially useful in identifying localized bearing faults that cause high-frequency impacts. Gearbox housing Cross-section Consider an idealized gearbox that consists of @ 13tooth pinion meshing with a 36-tocth gear. The pinion is coupled to an input shaft Connected to a prime mover. The gear is connected to an output shaft. The shafts are supported by roller bearings on the gearbox housing. Two accelerometers, AA, and AA,, are placed on the bearing and gearbox housings, respectively. The accelerometers operate at @ sample rate of 20 kHz The pinion rotates at arate fgg = 22-5 Hz oF 1950 rpm. The rotating speed of the gear and output shaft is ‘Number of pinion teeth (N) = Ay Foon = sie "Seamer of gear teeth (N, The tooth-mesh frequency, also called gear-mesh frequency, is the rate at which gear and pinion teeth periodically engage: Jen = Sean XNp = Soar Np fs = 2063; % Sample Rate (Hz) Np = 235 % Number of teeth on pinion Ne = 35; % Nunber of teeth on gear #Pin = 22.55 % Pinion (input) shaft frequency (Hz) Gear = FPin*Np/Ng; % Gear (Output) shaft frequency (Hz) #esh PintNp; X Gear Mesh frequency (Hz) Generate vibration waveforms forthe pinion and the gear. Model the vibrations as sinusoids occuring at primary shaft gear mesh ‘frequencies. Analyze 20 seconds of vibration data, The gear-mesh waveform is responsible for transmitting load and thus possesses the highest vibration amplitude. A. records vibration Contributions from the two shafts and the gear-mesh, For this experiment, the contributions ofthe bearing rolling elements to the vibration sighals recorded by A are considered negligible. Visualize a section of noiseree vibration signal t = e:1/Fs:20-4/F5; véin = 0.4¢sin(2¢pi*fPintt); Pinion wavefore vfout = 8.2"sin(2tpirfcear*t); % Gear waveform wesh in(2*pitMesh**); —-% Gear-nesh waveform plot(t, vétn + vfout + vmesh) xdin([@ @.25]) xdabel (‘Tine (s)') ylabel (‘Acceleration’) 15 Acceleration S 0 0s on 15. 02 025 Time (s) Gonorate High-Frequency Impacts Caused by a Local Fault on a Gear Tooth ‘of the gear is suffering from a local fault such as a spall. This results in a high-frequency impact occurring ‘Assume that one of the ‘once per rotation ofthe gear. Affected gear tooth (Exaggerated) Pinion The local faut causes an impact that has a duration shorter than the duration of tooth mesh. A dent on the toath surface ofthe gear generates high-requency oscilations over the duration ofthe impact. The frequency of impact is dependent on gearbox component Properties and its natural frequencies. In this example, i s arbitrarily assumed that the impact causes a 2 kHz vibration signal and occurs ‘over a duration of about 8% of 1/Mesh, oF 0.25 milliseconds. The Impact repeats once per rotation ofthe gear. Apt = Gear; finpact = 2000; tinpact = @:1/F5:2.5e-4-1/#5; ximpact = sin(2*pi*#inpact*timpact) /3; Make the impact periadie by convolving it with a comb function, xcosb eros(size(t)); Ind = (0.25°F5/#esh) :(F5/ipf) :Length(t) Ind = round(nd); xCosb(Ind) = 45 xPer = 2¥conv(xComb,xImpact, "sane"; ‘Add the fault signal xPer fo the shaft signal. Add white Gaussian noise tothe output signals for both the faul-ee and the faulty gear to ‘model the output from A WNoFault = vfln + vfout + vMesh; veault = vNoFault + xPers WNoFaultNolsy = vNoFauls vFaultNoisy © + randn(size(t))/5; vFault + randn(size(t))/5;, Visualize a segment ofthe time history. The impact locations are indicated on the plot forthe fauily gear by the inverted red triangles. They ‘are almost indstinguishable,

Вам также может понравиться