Papers and Case Studies

Good Troubleshooting, Good Results

It’s often unlikely that you will fix a vibration issue on an asset with just one solution. Sometimes we get lucky, and the problem is simply unbalance. But more than likely you will have to implement at least two solutions to fully address the issue.

When Red Wolf reliability was contracted to troubleshoot and balance a belt driven fan that was consistently suffering from high vibration problems, this was entirely the case. Excessive vibration amplitudes had caused repeated damage to the structure for years. Initial readings on the fan bearings showed a dominant frequency at fan running speed, near 0.6 IPSpk. Figure 1 shows the amplitudes at fan speed that were initially measured at the inboard bearing. However, phase analysis showed a stationary force, so further troubleshooting was performed. Troubleshooting of the fan included bump testing around the structure and fan wheel, start-up and coast-down measurements, and a visual inspection of the entire structure. The testing showed a natural frequency on the fan wheel at approximately fan running speed which indicates a likely resonant condition. Figure 2 shows the response of the bump test. The visual inspection did not yield any new structural damage but gave good insight into how much known damage the resonance had caused. Consideration was also given to the motor, which was operating at nearly the same speed as the fan. Ultimately, it was decided to leave the motor unchanged until further testing proved it needed to be replaced with one operating away from the natural frequency of the fan wheel. The recommendation was made to increase the size of the fan sheave in order to decrease the fan running speed more than 30 percent away from the natural frequency, but no other sheave sizes were readily available. Despite the confirmation of a resonant condition, an attempt was made to balance the fan. Ultimately this proved unsuccessful as resonance is an aggravating condition that alters amplitudes and phase responses required for balancing.

Figure 2: Natural Frequency of the Fan Wheel at Fan Running Speed.

A sheave could not be obtained that would separate the fan speed from the natural frequency by 30 percent, so a sheave was chosen that would separate fan running speed by only 20 percent from the natural frequency. When the new sheave was installed and the current structural damages fixed, Red Wolf Reliability was called back to determine if the fan required balancing. Additionally, all the previous weight, including factory, was removed and the fan wheel cleaned. The readings on the fan bearings continued to show excessive vibration amplitudes, near 0.9 IPSpk at fan speed, but the new phase readings showed a rotating force since the resonant condition was mitigated. Since the phase data now showed a rotating force, the decision was made to attempt to balance the fan.

The initial balance grade quality was over G150, with displacement amplitudes over 6 MILSpk at the outboard bearing. After four balance runs, the amplitudes were reduced to 0.3 MILSpk at the inboard bearing, and 0.2 MILSpk at the outboard bearing in the radial directions. The final balance grade quality was G4.8, which is within the ISO 10816 balance grade of G6.3 that applied to this fan. Velocity spectra amplitudes at fan speed went from 0.89 IPSpk to 0.09 IPSpk. Figure 3 shows the final recorded amplitude at fan speed on the outboard bearing.