A coal mill was observed to go into resonance at various coal feeder speed. Dynamic displacements of the three grinding rollers were measured using instrumentation techniques developed by Engineering Consultants Group Inc. (ECG). As a result of the resonant condition, shaft failures occurred.
A finite element model of the mill and grinding rollers was developed. The two major components of the FEA model were tied together through the coal bed stiffness. Results showed that the displacement measurements corresponded to the first and second vibration modes of the coal mill. A slip-stick motion between grinding rollers and coal bed in the bowl of the mill is believed to be the excitation source.
COAL MILL INTRODUCTION:
Coal mills grind pea-sized coal into a powder, which is blown into the furnace of a power plant. Normally, there are a number of mills that supply the ground coal to the furnace and the amount of coal going into the furnace is controlled by the amount of feed to each mill and by the number of mills on line . In the mill considered, the bowl of the mill turned at a measured speed of 58.5 rpm.
Three grinding rollers were forced against the coal bed by compression springs. There are stops on each roller to prevent loading on a bowl with no coal. Once the coal is fed into an operating mill the roller lifts off the stop. For 100% coal feeder loading, the coal bed on the bowl has a thickness of about one inch.
In the particular design that was considered, the operation of the mill was normal until the coal feeder reached about 65% of full loading. Then the three grinding rollers began to vibrate with amplitudes that increased by an order of magnitude at 75% of feeder loading than with the coal feeder loading below 60% of the full value. This vibration often led to premature mill shaft fatigue failures.
This mill grinding roller vibration was measured and analyzed and the entire mill modeled with finite elements with the objective of determining the reason for the vibration and to seek a means of eliminating the vibration and resulting shaft failures. Thus, the investigation was divided into two parts: an experimental study and finite element modeling.
COAL MILL EXPERIMENTAL INVESTIGATION:
Probes were mounted at the end of the compression spring shafts shown on Figure 2 by the bold arrow. These transducers recorded the dynamic motions of the three grinding rollers as well as the depth of the coal bed. These motions could be correlated with the rotation of the bowl. The coal bed depth increased with the coal feeder loading of the mill with a maximum depth of about 1 inch. As a result, the stiffness of the coal bed decreased as the loading to the mill increased. As will be described the analysis section, this change in stiffness changes the frequencies of the first two modes of vibration and brings the system into resonance. Measured coal depth with feeder rate for Roll 1 is plotted on Figure 3. Coal bed depths as a function of feeder speed of the other two rollers are similar.
Measured vibration amplitude of Roll 1 is plotted on Figure 5 for feeder speeds of 50%, 70% and 90%. The variation of the vibration magnitude can easily be observed. This amplitude is plotted on Figure 4. Resonance begins when the feeder loading exceeds 65% of capacity and ends at approximately 85% with a dominant frequency of 15.8 Hz. The resonant frequency was determined with a FFT of the measured motion as shown on Figure 5. This same resonant frequency was found in the dynamic measurements of all three rolls. However, this resonant value is not seen in the FFT of the roller motions below 60% feeder loading or above feeder loading of 90%. Two mode shapes of the three grinding rollers were measured at resonance. At times two rollers vibrated out-of-phase with the third roller almost motionless. In the second case, two grinding rollers move in-phase with the third roller out-of-phase. Moreover, the third vibrating roller had higher amplitudes then the two in-phase rollers.
COAL MILL FINITE ELEMENT STUDY:
The finite model of the coal mill was developed with the Noran Engineering, Inc version of NASTRAN and FEMAP. The actual modeling was developed with FEMAP, exported to NASTRAN where calculation was run and the results were imported back to FEMAP to view and interpret the vibration modes. The mill was modeled in two parts. First, a FEA model of the three grinding rollers in space was developed. The rollers were located in space such that the rollers would fit properly in the bowl. These rollers were basically rigid and pivoted about constraints that had to be located in the local coordinate systems of the rollers.
CONCLUSIONS: After the completion of this analytical work, a stronger blower was installed in the mill to carry the ground coal powder to the boiler. As a result of this change, the amplitude of vibration at resonance was reduced by a factor of approximately two. My reducing the amount of coal dust on the bowl of the mill, the friction between the bowl and roller is altered, probably reduced, giving support to the slip-stick theory.
FUTURE WORK: The effects of installing dampers in the mill will further reduce the resonant amplitude. Time will tell if the present reduction in amplitude will significantly prolong shaft life or eliminate shaft failures. It is planned to investigate the use of dampers attached to the ends of the roller shafts to further reduce vibration resonant amplitude in case that fix is required.
ACKNOWLEDGEMENT: The authors would like to acknowledge the assistance of the technical staff at Noran Engineering, Inc., Los Alamitos, CA during the development of the FEA model.
Archives
Recent Posts
Categories
Calendar
Meta
Experimental and Analytical Diagnostic Evaluation of Coal Mill Vibration
coal mill ABSTRACT:
A coal mill was observed to go into resonance at various coal feeder speed. Dynamic displacements of the three grinding rollers were measured using instrumentation techniques developed by Engineering Consultants Group Inc. (ECG). As a result of the resonant condition, shaft failures occurred.
A finite element model of the mill and grinding rollers was developed. The two major components of the FEA model were tied together through the coal bed stiffness. Results showed that the displacement measurements corresponded to the first and second vibration modes of the coal mill. A slip-stick motion between grinding rollers and coal bed in the bowl of the mill is believed to be the excitation source.
COAL MILL INTRODUCTION:
Coal mills grind pea-sized coal into a powder, which is blown into the furnace of a power plant. Normally, there are a number of mills that supply the ground coal to the furnace and the amount of coal going into the furnace is controlled by the amount of feed to each mill and by the number of mills on line . In the mill considered, the bowl of the mill turned at a measured speed of 58.5 rpm.
Three grinding rollers were forced against the coal bed by compression springs. There are stops on each roller to prevent loading on a bowl with no coal. Once the coal is fed into an operating mill the roller lifts off the stop. For 100% coal feeder loading, the coal bed on the bowl has a thickness of about one inch.
In the particular design that was considered, the operation of the mill was normal until the coal feeder reached about 65% of full loading. Then the three grinding rollers began to vibrate with amplitudes that increased by an order of magnitude at 75% of feeder loading than with the coal feeder loading below 60% of the full value. This vibration often led to premature mill shaft fatigue failures.
This mill grinding roller vibration was measured and analyzed and the entire mill modeled with finite elements with the objective of determining the reason for the vibration and to seek a means of eliminating the vibration and resulting shaft failures. Thus, the investigation was divided into two parts: an experimental study and finite element modeling.
COAL MILL EXPERIMENTAL INVESTIGATION:
Probes were mounted at the end of the compression spring shafts shown on Figure 2 by the bold arrow. These transducers recorded the dynamic motions of the three grinding rollers as well as the depth of the coal bed. These motions could be correlated with the rotation of the bowl. The coal bed depth increased with the coal feeder loading of the mill with a maximum depth of about 1 inch. As a result, the stiffness of the coal bed decreased as the loading to the mill increased. As will be described the analysis section, this change in stiffness changes the frequencies of the first two modes of vibration and brings the system into resonance. Measured coal depth with feeder rate for Roll 1 is plotted on Figure 3. Coal bed depths as a function of feeder speed of the other two rollers are similar.
Measured vibration amplitude of Roll 1 is plotted on Figure 5 for feeder speeds of 50%, 70% and 90%. The variation of the vibration magnitude can easily be observed. This amplitude is plotted on Figure 4. Resonance begins when the feeder loading exceeds 65% of capacity and ends at approximately 85% with a dominant frequency of 15.8 Hz. The resonant frequency was determined with a FFT of the measured motion as shown on Figure 5. This same resonant frequency was found in the dynamic measurements of all three rolls. However, this resonant value is not seen in the FFT of the roller motions below 60% feeder loading or above feeder loading of 90%. Two mode shapes of the three grinding rollers were measured at resonance. At times two rollers vibrated out-of-phase with the third roller almost motionless. In the second case, two grinding rollers move in-phase with the third roller out-of-phase. Moreover, the third vibrating roller had higher amplitudes then the two in-phase rollers.
COAL MILL FINITE ELEMENT STUDY:
The finite model of the coal mill was developed with the Noran Engineering, Inc version of NASTRAN and FEMAP. The actual modeling was developed with FEMAP, exported to NASTRAN where calculation was run and the results were imported back to FEMAP to view and interpret the vibration modes. The mill was modeled in two parts. First, a FEA model of the three grinding rollers in space was developed. The rollers were located in space such that the rollers would fit properly in the bowl. These rollers were basically rigid and pivoted about constraints that had to be located in the local coordinate systems of the rollers.
CONCLUSIONS: After the completion of this analytical work, a stronger blower was installed in the mill to carry the ground coal powder to the boiler. As a result of this change, the amplitude of vibration at resonance was reduced by a factor of approximately two. My reducing the amount of coal dust on the bowl of the mill, the friction between the bowl and roller is altered, probably reduced, giving support to the slip-stick theory.
FUTURE WORK: The effects of installing dampers in the mill will further reduce the resonant amplitude. Time will tell if the present reduction in amplitude will significantly prolong shaft life or eliminate shaft failures. It is planned to investigate the use of dampers attached to the ends of the roller shafts to further reduce vibration resonant amplitude in case that fix is required.
ACKNOWLEDGEMENT: The authors would like to acknowledge the assistance of the technical staff at Noran Engineering, Inc., Los Alamitos, CA during the development of the FEA model.