Fault Diagnosis for Dyring Cylinder in Paper Machine

Case Study

Fault Diagnosis for Dyring
Cylinder in Paper Machine

1. Abstract

Drying Cylinders in paper machine are challenging to monitor given that they rotate at a low rpm and significantly high temperature. Sudden expansion and contraction of the rolls can cause breakage and bring the entire machine to a halt. This paper demonstrates the use of Infinite Uptime’s low-speed bearing diagnostics capability to help detect faults at an early stage.

2. Introduction

The Dryer Section indicated in Figure 1 is part of the paper machine installed at paper making plant with an annual capacity of 300,000 tonnes. Sensor nodes enabling fault diagnosis for this are mounted at the load zone of the spherical roller bearings supporting the rolls at either ends. These sensor nodes communicate to a cloud dashboard and provided comprehensive vibration signal analysis and fault diagnosis. The necessity for installing the system was the unusually high failure-rate among bearings on the dryer section of the plants Valmet paper-machine. Between 2017 and 2022 these sudden failures cost the company in the region of 21 million USD in terms of lost production alone. Inclusion of the costs incurred in overtime payment to the maintenance personnel, ancillary damage caused to the other components by a sudden bearing failure and financial losses due to inability to deliver on schedule would considerably increase the estimate of the losses given above.

Figure 1: Drying Cylinders

3. Background

3.1 Drying Cylinder
Construction

Dryer cylinders on this paper machine are 5 feet in diameter, designed for a steam pressure of up to 100 PSI., 2-10 in number grouped together for drying the paper web. Cylinders are covered on 60 – 65% of its circumference by an endless felt arranged to travel with the rotating cylinder. They are driven by electric motors via reduction gear boxes directly connected with gears and pinions through indrive shafts and couplings. Drying cylinders are mounted on SKF 23048 CCK W33 spherical roller bearings at both ends, and the speed varies from 40 to 120 rpm. Servosystem 150 oil supply comes to the bearing through COL system with a pressure of 2.6kg/cm 2 . Normally the surface temperature of the bearing housing would be around 85°C and the surface temperature of the drying cylinder would be 110°C with an inlet steam pressure of 3.5 kg/cm 2 . Running the roll under with a worn out bearing under a fault condition could lead to surprise breakdowns

3.2 Sequence
of Events

Figure 2: shows the sequence of events from fault diagnosis to corrective action.

4. Solution

The plots presented in this section validate the presence of the fault as predicted by the system.

Advanced Fault Diagnosis in IDAP Platform

A. Increase in Axial Velocity from Jan 2 to Jan 23.

Figure 3: Increase in Axial Velocity

B. Bearing defects clearly indicated in the frequency plot for bearing 23048 CCK W33 C4

Figure 4: Spectrum plot showing bearing defects

C. Post correction action – vibration values significantly reduced 3 Feb onward.

Figure 5: Vibration levels before and after bearing replacement

D. The bearing defect frequencies for bearing 23048 CCK W33 C4 clearly visible before corrective action are not seen in the frequency plot captured after corrective action.

Figure 6: Frequency plot before and after corrective action

5. Site Inspection and Findings

When the bearing at the drive side was removed and inspected. It was found to be severely spalled on the inner race, outer race and rolling elements. See Figure 7. Thermal expansion of the shaft causing excessive axial force on the bearing resulted in accelerated fatigue, eventually damaging the bearing completely. Subsequently better steam and temperature control have been adopted with improved bearing lubrication mechanism.

Figure 7: Severe damage can be observed on the raceways and rolling elements

6. Conclusion

Once the bearing fault was detected, it was possible to schedule maintenance actions for the cylinder roll and carry out the bearing replacement. Every minute that the paper mill is down represents revenue loss. For the timely replacement, the machine was taken down for 6 hours. In an unscheduled breakdown, the time would have been 15 hours. A loss of 76.5 tons of paper was avoided.

Case Study