Field balancing

Vibration measurement for field balancing

Field balancing is a special force-up method which is performed directly on the rotor to be balanced without the balancing bench. This allows rotors on machines to be balanced when installed without dismantling. Field balancing uses vibration measurements and the setting of defined imbalance test weights in order to indirectly infer the existing imbalance .

Typical process for field balancing

Rotor prepared with reflective strips and laser light barrier for operational balancing

In the idealized case, field balancing takes place on a built-in rotor as follows. The frequent case is described here that a rigid rotor is balanced in two planes .

Attach vibration sensors (at least 2) to representative locations. Typically on the input and output side bearings.
Set the zero mark on the rotor (reflective strip) and align the light barrier on it so that you measure exactly one trigger pulse per revolution at a given angle.
Let the rotor run at nominal speed. For all sensors, measure and save the synchronous oscillation ("1st harmonic") as the original oscillation with amplitude and phase (against trigger).
Place a sample weight with a defined mass on a defined radius and angle on balancing plane 1.
Let the rotor run again and measure the amplitude and phase again with sample weight 1 and finally save.
Remove the sample on level 1 and place the sample weight on balance level 2 instead .
Measure the amplitude and phase again, now for sample weight 2. Then remove sample weight 2.
Calculate the so-called influence coefficients from the knowledge of the original vibration, the changes in the vibrations based on the known sample masses and the angle and mass of the sample masses set. This is usually done by a special program or the balancing device. The influence coefficients contain the system knowledge as the relationship between mass and vibration change.
Calculate the original imbalance from the original vibrations and the influence coefficient. The balance weights to achieve a balanced state are then exactly opposite the original unbalance.
Set the calculated balance weights on balancing levels 1 and 2 on the rotor
Finally run the rotor again at nominal speed to check the success of the balancing measure. The vibration values should have decreased significantly.

Differentiation from balancing banks

In contrast to field balancing, the rotor is “ hard ” mounted in balancing banks , i. H. the speed at which the first resonance mode ( natural frequency ) occurs is far above the balancing speed. Therefore, the rotor almost does not perform any oscillation with a relevant oscillation path. The balancing banks are force-measuring, so that the centrifugal force generated by the unbalance at a given speed per level can be measured directly. From the force measurement, if the geometry of the measuring planes and balancing planes is known, conclusions can be drawn directly about the existing imbalance.

Situation with field balancing instead of a balancing bench

Course of the vibration vectors of two measuring points during field balancing. One approaches more and more the 0-oscillation / origin.

In field balancing, the vibration-dynamic bearing condition of the rotor - with regard to the stiffness, mass and damping matrix - is initially unknown. At nominal speed , rotors can run below the first resonance speed (“hard”), above (“soft”) or even directly at or near a critical speed, ie “in resonance”. Since a force measurement only makes sense if the bearing is “hard”, vibration measurements are used instead in order to get a connection to the existing imbalance.

As explained in the section “Balancing benches”, there is a known relationship between unbalance and exciting force at a given speed and known geometry; In the second step, the relationship between the exciting force and the resulting vibration during field balancing is initially unknown. All that is known from vibration theory is that this is a multi-dimensional linear relationship in which every force excitation at a given point (unbalance level) leads to a proportional vibration response at a measuring point (measuring level). If you represent the sinusoidal oscillation of the force and the oscillation movement as complex numbers at a given speed, you get the relationship discussed - the so-called "influence coefficient" or influence matrix.

One makes use of this physical fact. As soon as one knows the influence coefficient matrix, one can deduce the existing exciting forces directly from a vibration measurement; and from there one concludes the existing imbalance (for each imbalance level); this is ultimately balanced out.

Problems and glitches

Due to the process, there are a number of possible challenges and malfunctions in field balancing:

Machines and systems that react non-linearly to sample unbalanced masses. A frequent and demanding situation to balance. See also the discussion here .
Machines that run on resonance at nominal speed and therefore react very sensitively to the smallest residual imbalance. Particularly problematic in connection with a non-linear system response.
Systems with " poorly conditioned " influence coefficient matrices due to unfavorably set sample dimensions, unskillfully chosen sensor positions or resonance that emphasizes a mode disproportionately. In such a case, the underlying mathematics / numerics tends to result in strongly incorrect compensation mass proposals, even with small measurement errors.
Frequency-controlled machines: The challenge here is not only to "balance out" the prevailing mode at one speed , but to minimize the entire imbalance in both levels so that no vibration mode reaches impermissibly high amplitudes over the entire speed range.

Suitable machines

Field balancing can be used on almost all machines that can be run with test weights and that are accessible for vibration measurements. Field balancing is often used for:

Fans , fans and blowers
Mulching and chopper shafts in agriculture
Propellers
Wind turbine rotors.

Individual evidence

^ Hatto Schneider: Balancing technology . 8th edition. Springer Vieweg, 2013, ISBN 978-3-642-24913-6 , p. 403 .
↑ ^a ^b Robert Gasch, Rainer Nordmann, Herbert Pfützner: Rotordynamik . 2nd Edition. Springer-Verlag, Berlin 2006, ISBN 3-540-41240-9 , "2.4 Balancing in three runs", p. 706 , p. 24 .
^ From Robert Gasch, Rainer Nordmann, Herbert Pfützner: Rotodynamik . 2nd Edition. Springer-Verlag, Berlin 2006, ISBN 3-540-41240-9 , "2.3 Forces measuring balancing in hard bearings".
↑ Fu, Zhi-Fang .: Modal analysis . Butterworth-Heinemann, Oxford 2001, ISBN 1-4294-9778-5 , pp. 304 .
↑ Successful balancing in non-linear systems - tips - conplatec . In: conplatec . November 8, 2018 ( conplatec.de [accessed November 8, 2018]).

[1] Hatto Schneider: Balancing technology . 8th edition. Springer Vieweg, 2013, ISBN 978-3-642-24913-6 , p. 403 .

[:0-2] Robert Gasch, Rainer Nordmann, Herbert Pfützner: Rotordynamik . 2nd Edition. Springer-Verlag, Berlin 2006, ISBN 3-540-41240-9 , "2.4 Balancing in three runs", p. 706 , p. 24 .

[3] From Robert Gasch, Rainer Nordmann, Herbert Pfützner: Rotodynamik . 2nd Edition. Springer-Verlag, Berlin 2006, ISBN 3-540-41240-9 , "2.3 Forces measuring balancing in hard bearings".

[4] Fu, Zhi-Fang .: Modal analysis . Butterworth-Heinemann, Oxford 2001, ISBN 1-4294-9778-5 , pp. 304 .

[5] Successful balancing in non-linear systems - tips - conplatec . In: conplatec . November 8, 2018 ( conplatec.de [accessed November 8, 2018]).