Localizing and eliminating natural vibrations

The water tank had to be drained to enable impact tests in the wet area.
Fig. 1: The water tank had to be drained to enable impact tests in the wet area.
The frequency spectrum of the vibration velocity is shown logarithmically.
Fig. 2: The frequency spectrum of the vibration velocity is shown logarithmically.

In the course of modernization efforts, a run-of-the-river hydroelectric unit was equipped with a new, more powerful impeller. Upon startup, the unit exhibited additional noise and vibration at around 70% load that had not been present with the old impeller. The acceptance of the power plant was endangered and PRUFTECHNIK was contracted to search for the causes.

On the day of measurement, technicians used a VIBXPERT to measure the vibration behavior of the unit at various loads and to analyze the results on-site. The analysis showed that disturbances in the critical load range occurred with a slight delay and then appeared as broadband excitations in the frequency range of around 190 Hz, both acoustically and in the housing vibrations – only to disappear again as the load increased further. Thus, cavitation and disturbing vibrations, which had been assumed to come from the newly wound generator, could be excluded. The only possible remaining source of vibration was natural vibrations arising from the hydroelectric unit itself. The question therefore was: Which components of this complex structure were responding to the natural frequencies? To find out, impact tests were made. First, possible vibration exciters were looked for in the dry area. Because none of the components were sensitive to vibration in the frequency range in question, the wet area of the hydroelectric unit had to be drained to be able to enter the unit. In the course of the search for the 190-Hz excitation, various impact tests were performed on the guide vane apparatus, guide equipment, blades, etc.

The front edge of each impeller blade was ground down.
Fig. 3: The front edge of each impeller blade was ground down.
The front edge of each impeller blade was ground down.
Fig. 4: The front edge of each impeller blade was ground down.

VIBXPERT found what we were looking for on two structures: the walls in the impeller and a large guide plate.But which of these was the cause? The search was continued by excluding one of the candidates. To do this, the guide plate was temporarily tensioned until the impact test exhibited a frequency shift – away from 190 Hz. After sometime when the disturbing vibration reappeared at partial load condition, it was clear that they had to come from the impeller. Yet how could the massive impeller be modified?


The idea for the solution came from PRUFTECHNIK: Chamfering the impeller by grinding down the front edge of each blade would be sufficient to eliminate
the disturbance.

Two days later in the evening, the happy manufacturer personally returned the borrowed VIBXPERT to PRUFTECHNIK.

Do you like this application?

Enquire now!

 

 

Contact us