Case – Balancing as a compromise

A company in the waste processing industry noticed an urgent problem on a fan (3000 rpm). This machine is crucial for the unit. It had recently been overhauled. The people performing the overhaul had remarked on site that there was excessive axial play in the bearing on the clutch side. This could not be fixed immediately. The entire launch of the unit would have to be postponed.The company was not satisfied with the action of the overhaul company and contacted us. When we arrived on site, there were very high vibrations on the continuous measurement (100% - 22 mm/s). There were also sparks on the level of the axle seal on the side of the fan.

Observation during measurement:

Initially the fan was started to get an idea of what the problem was. When measuring, the machine always runs against a closed press, as the unit is out of use.


We notice very high vibrations. The following graph shows the spectrum of the horizontal measurement on measuring points 3 and 4.


A runout-measurement was conducted with the following result:

  • The runout-measurement shows that the vibrations diminish very rapidly. This isn’t the default runout of an imbalance. We notice a phase of limited alternation.
  • In the vibration measurement we also notice a big difference in both horizontal and vertical direction.

Both phenomena seem to indicate an amplification range. Balancing was started, but we warned that the chance of success was doubtful (due to the amplification range and the play phenomena). Presumably the fan makes a waltzing movement because of the increased play. A higher balancing class can limit this phenomenon.


In the reference measurement we do not notice a 90° phase difference between the 

sensors. This also indicates an amplification range in the horizontal direction.


These vibration values can hardly be improved by means of balancing on site close to an amplification range.


The measurement shows a strong component on the blade frequency, because the machine runs against a closed press valve.

The amplitude of the blade frequency is almost as high as the 1x vibration. After this measurement we declared the machine ready for use again. There was a measurement during the launch of the machine as well. At a valve opening of 50% there was another measurement.


Measurement point 4 on the fan still shows the same increased play behaviour:


Since the vibration values are limited, we can consider this acceptable.

Conclusion of the measurement:

  • Balancing has limited all phenomena (play, amplification).
  • After the action the machine runs almost like before, according to the continuous measurement.

Advice of the measurement:

During the next shutdown increased attention should be paid to an increased internal play and ways to prevent this.