Electrical and Current Signature Analysis Application Part 8 Gearbox Analysis with Information

In parts 7a and 7b we explored a gearbox on a VFD with limited information, that was also having some interesting operating issues. In this case, we are going to take a look at a system with a gearbox, including gears and bearings, and some of the tools I like to use. In this case, the machine is 1785 full load RPM, the gearbox we are looking at has a few more sets of gears and bearings than what we are going to evaluate in this article. In this case, it has teeth starting from the drive end of 19 pinion, 74 wheel (stage 1) and 21 pinion and 72 wheel (stage 2). The bearings we are going to evaluate are 6330 on the DE (Drive End) and ODE (Opposite DE) for the machine and both bearings on stage 1 are NU2324 (we will not evaluate the thrust bearings) and NU2334 pinion and NU2336 wheel on stage 2.

Now for some of the tools I like to use:

The above is an iPhone and iPad app for calculating gears in a gearbox. In this case it also allows you to create a custom gearbox in the app, which we will do here, to provide a visual with resulting outputs. We will also talk about how we do simple calculations to accomplish the same thing.

The above is also a good app for bearing multipliers. However, it does not have all of the bearings we are interested in. The software I use for ESA does have a good database of bearings and you can also go to the bearing manufacturers sites to obtain the 1X multipliers. These become important whether you are performing ESA or vibration analysis.

The above is the assembled calculator for stage 1 and 2 of the system. The speeds of each stage are important for analysis purposes. The pinion on the top left is operating at the machine operating speed of 1790.4 RPM, the next is (pinion/wheel) * RPM, or (19teeth/74teeth)*1790.4 = 459.7 RPM, the second stage is (21teeth/72teeth)*459.7 RPM = 134.08 RPM. If I divide each by 60Hz, that would give me 29.84 Hz, 7.662 Hz and 2.235 Hz. Peaks related to these frequencies in the gearbox would be indicators of meshing and other issues related to the gears, such as cracked, work or broken teeth. These multipliers are also used to evaluate bearings for what is attached to each stage.

Above are the frequencies for the bearings: BPFI is inner race; BPFO is outer race; BSF is related to the ball; and, FTF is the cage frequency. Yes, you can identify the portion of the bearing that is in distress. You will also note that it shows 1X through 4X as sometimes the problems with the bearings will show up only in harmonics (multiples) of the fundamental (1X) frequency. In the above picture I also used Hz instead of RPM (CPM). For the other bearings, including the machine bearings, I have gone ahead and calculated the 1X Hz values as below:

The top table is the multipliers for each bearing and the second table are the frequencies based upon the speeds for each stage. These we will use to identify problems with the bearings.

While some of the bearing issues will also be found in the higher frequencies we evaluate, the above image will help us start to review the conditions of the machine and gearbox. In a repeat of a previous posting, it is important to note that sometimes a peak may show without a pair or one may appear higher than another. This has to do with the filtering used by the various technologies. In the above, we are using a Hanning window.

Some of the peaks also appear to have wide bases and the ones that show only one side can be seen here in the waterfall spectra. This shows that various peaks fade in and out as the machine and gearbox operate.

We can also see from the torque spectra, above, that there is a regular pattern at about 13 Hz (time is across 1 second and the number of peaks or valleys are counted). This can be seen in the current spectra with sidebands around those peaks. It is also noted that peaks close to operating speed (~30hz and ~90Hz) are not present. This eliminates alignment, balance or bent shaft as a possible problem. The 13 Hz peak is found as sidebands around 60Hz as 13.761Hz, which is related to the input to output speed ratio of 13.35 Hz. The wide base shows that this varies a little and includes the 13.35Hz, resulting in the identification that something is binding in the gearbox at the second stage.

The ratios are calculated from the high speed, or machine speed. So, the first stage has a ratio of 3.89, which is also found to the right of the line frequency at 64.209Hz with a slight base. The sidebands in close to the line frequency show at similar frequencies to the sidebands about the 13.761 Hz peaks. An evaluation of the PPF indicates that these do not relate to the rotor, but are related to the gearbox, indicating possible severe wear or even a cracked/fractured tooth.

With the bearings we are evaluating, we are also seeing peaks at the sidebands around line frequency related to the second stage bearings (NU2334 and NU2336) inner race. (I will note that the other peaks are indicative of other bearings not being evaluated in this article) This may indicate that there are orbiting issues with the second stage of the gearbox. The peaks are high with sidebands around the 13 Hz signature, indicating that the problem is relatively severe.

The recommendation was to have the gearbox evaluated for second stage issues of which damaged inner bearing race(s) of the bearings and gear tooth wear/fractures were possible. Both issues were confirmed.

In Part 9 we will take a look at evaluating your system with the power quality aspect of ESA.

MotorDoc LLC is a service company that provides PdM, quality assurance and forensics of machinery failures. For more information, contact us at [email protected]. You can also find additional information at https://motordoc.com.

MotorDoc LLC will also be scheduling SMRP Approved Provider Electrical Signature Analysis training in 2018. Sign up for the newsletter at motordoc.com to be informed.