The axial displacement tracking system 

Illustration

Introduction.

The axial displacement monitoring system is one of the main ones in the general complex of measures for the protection of rotary machines (turbocompressors). Other disturbances in the operation of machines can also lead to catastrophic consequences, but deterioration of operation or a defect in a thrust bearing can occur with very weak signs of danger and in a very short period, leading to complete destruction of the machine. Therefore, first of all, technical techniques for measuring axial displacement are needed. At the same time, it is necessary to avoid mistakes when installing appropriate protection systems.

This article discusses the most important points when installing such systems:

- The concept of "cold" and "hot" surfacing zones;

- The relationship between the linear range of the system sensor and the range of potential change in the axial displacement of the rotor;

- The relationship between the monitor readings and the rotor position (engaged disk);

- a guide to setting threshold levels is provided.

Illustration
Illustration

The float zone is the normal allowable movement of the thrust disc in the thrust bearing clearance. The "cold" float zone (Fig. 1) is measured when the machine is stationary and at ambient temperature. However, this zone will increase when the machine is fully loaded at operating speeds. The change is due to the greater operating load on the thrust bearing.

Other contributing factors are thermal heating, insufficient thrust bearing assembly stiffness, uneven thrust bearing pad thickness, and oil film wringing (emphasis added).

Thus, for a fully loaded machine, there is a hot float zone that is usually significantly colder than the colder zone. The example in Figure 1 shows a machine with a cold zone of about 16 mils (0.4 mm), with a gap across the sensor of 48 to 58 mils, which corresponds to a proximeter output voltage of -8.4 and -11.6V, respectively. The hot zone is 24 mils (0.6 mm) with a gap of 38 to 62 mils (-7.6 and -12.4V). This 50% increase is not unusual.

Very often, inexperienced users of axial displacement monitoring systems do not take into account the difference between cold and hot float zones. They will set the threshold level to correspond to the moment when the driven disc touches the bearing, based on measurements of the cold zone of the machine when it was in a static position. If the zone blurs, the point of such a threshold level represents the position of the driven disc within the bearing gap, not the contact of the liner. Thus, a change in shaft position causes an incorrect signal in the monitoring system.

There are two ways to avoid false alarms. First, find out the difference between the cold and hot surfacing zones, and set the threshold level correctly. Second, set the threshold level signal so that it represents 5 to 10 milliseconds (125-250 microns) of babbitt wear. Then set the blocking level corresponding to 250-500 microns above the threshold level.

False threshold level activations can occur even after the hot surfacing zone allowance is made. This is possible if:

1) the threshold level is set too close to the bearing surface;

2) the hot zone allowance is insufficient;

3) a small error is contained in the installed sensor.

Ultimately, monitoring actual axial misalignment will not necessarily prevent bearing wear, but it will prevent serious axial seizure and potential failure of the machine.

In fact, some wear of a stubborn bearing is desirable from a monitoring perspective. If the monitor displays a threshold level of axial misalignment and the bearing is not found during inspection, operators and management may lose confidence in such systems. Most machines are equipped with stubborn bearings that can tolerate some loss of babbitt before a dangerous axial contact occurs. It is therefore appropriate to allow some loss of babbitt before a threshold level signal is generated.

In order to determine the ratio of cold to hot float zones, the machine manufacturers should be consulted. Their information can be compared with field experience to improve performance.

Relationship between sensor range and rotor run-up range.

The required range of axial displacement measurements for any machine must cover the maximum allowable change in rotor displacement in both directions in the thrust bearing. The range of rotor displacement must include not only the clearances in the thrust bearing (cold and hot zones of the thrust bearing), but also the allowable wear of the babbitt layer on both sides of the bearing (working and non-working pads).

The machine (Fig. 1) has a clearance in the thrust bearing (hot zone) of 24 mils (0.6 mm). An additional 17 mils (0.4 mm) of allowable bearing wear on each side moves the point of blocking value. Thus, the range of rotor location (total allowable rotor displacement) is 58 mils (1.4 mm). The figure also shows the linear range of operation of the sensor, which is greater than the allowable range of axial displacement of the rotor. This placement of accents is recommended for all monitoring systems. That is, the more the range of correct sensor operation exceeds the axial displacement of the rotor, the more correctly the system can be adjusted.

If the linear range of the sensor is slightly larger than the axial runout of the rotor, then setting the sensor to adjust the correct clearance is difficult, if not impossible.

For example, if the linear range of the sensor is 60 mils (1.5 mm), then patience is required to set the center of the sensor range to the center of the axial runout of the rotor (cold zone). In this example, the sensor should be adjusted as close as possible to 58 mils (1.4 mm) or -11.8 V so that the thrust disk abuts the working side of the thrust bearing.
The float zone is the normal allowable movement of the thrust disc in the thrust bearing clearance. The "cold" float zone (Fig. 1) is measured when the machine is stationary and at ambient temperature. However, this zone will increase when the machine is fully loaded at operating speeds. The change is due to the greater operating load on the thrust bearing.

Other contributing factors are thermal heating, insufficient thrust bearing assembly stiffness, uneven thrust bearing pad thickness, and oil film wringing (emphasis added).

Thus, for a fully loaded machine, there is a hot float zone that is usually significantly colder than the colder zone. The example in Figure 1 shows a machine with a cold zone of about 16 mils (0.4 mm), with a gap across the sensor of 48 to 58 mils, which corresponds to a proximeter output voltage of -8.4 and -11.6V, respectively. The hot zone is 24 mils (0.6 mm) with a gap of 38 to 62 mils (-7.6 and -12.4V). This 50% increase is not unusual.

Very often, inexperienced users of axial displacement monitoring systems do not take into account the difference between cold and hot float zones. They will set the threshold level to correspond to the moment when the driven disc touches the bearing, based on measurements of the cold zone of the machine when it was in a static position. If the zone blurs, the point of such a threshold level represents the position of the driven disc within the bearing gap, not the contact of the liner. Thus, a change in shaft position causes an incorrect signal in the monitoring system.

There are two ways to avoid false alarms. First, find out the difference between the cold and hot surfacing zones, and set the threshold level correctly. Second, set the threshold signal level so that it represents 5 to 10 milliseconds (125-250 microns) of babbitt wear. Then set the blocking level corresponding to 250-500 microns above the threshold level.

False threshold level activations can occur even after the hot surfacing zone allowance is made. This is possible if:

1) the threshold level is set too close to the bearing surface;

2) the hot zone allowance is insufficient;

3) a small error is contained in the installed sensor.

Ultimately, monitoring actual axial misalignment will not necessarily prevent bearing wear, but it will prevent serious axial seizure and potential failure of the machine.

In fact, some wear of a stubborn bearing is desirable from a monitoring perspective. If the monitor displays a threshold level of axial misalignment and the bearing is not found during inspection, operators and management may lose confidence in such systems. Most machines are equipped with stubborn bearings that can tolerate some loss of babbitt before a dangerous axial contact occurs. It is therefore appropriate to allow some loss of babbitt before a threshold level signal is generated.

In order to determine the ratio of cold to hot float zones, the machine manufacturers should be consulted. Their information can be compared with field experience to improve performance.

Relationship between sensor range and rotor run-up range.

The required range of axial displacement measurements for any machine must cover the maximum allowable change in rotor displacement in both directions in the thrust bearing. The range of rotor displacement must include not only the clearances in the thrust bearing (cold and hot zones of the thrust bearing), but also the allowable wear of the babbitt layer on both sides of the bearing (working and non-working pads).

The machine (Fig. 1) has a clearance in the thrust bearing (hot zone) of 24 mils (0.6 mm). An additional 17 mils (0.4 mm) of allowable bearing wear on each side moves the point of blocking value. Thus, the range of rotor location (total allowable rotor displacement) is 58 mils (1.4 mm). The figure also shows the linear range of operation of the sensor, which is greater than the allowable range of axial displacement of the rotor. This placement of accents is recommended for all monitoring systems. That is, the more the range of correct sensor operation exceeds the axial displacement of the rotor, the more correctly the system can be adjusted.

If the linear range of the sensor is slightly larger than the axial runout of the rotor, then setting the sensor to adjust the correct clearance is difficult, if not impossible.

For example, if the linear range of the sensor is 60 mils (1.5 mm), then patience is required to set the center of the sensor range to the center of the axial runout of the rotor (cold zone). In this example, the sensor should be adjusted as close as possible to 58 mils (1.4 mm) or -11.8 V so that the thrust disk abuts the working side of the thrust bearing.

Illustration

Fig. 2 The zero point corresponds to the location of the driven disk in the center of the overflow zone.
 There are two common techniques used to set the axial displacement monitor to the start of the reading. Both methods are correct.

Method 1: The zero reading is the center of the slip zone. In this case, the monitor is set so that the “0” on the monitor corresponds to the location of the thrust disc in the center of the bearing (Fig. 2).

However, because the rotors rarely operate in the center of the clearance of the thrust bearing, the meter reading (when the machine is operating under normal operating conditions) will not be zero. Usually the reading will be different from zero (more often the displacement will be in the direction of the working pads), which corresponds to the middle of the hot slip zone.

Using this method as an example, a typical reading will be 12 mils (0.3mm) in the positive direction on the meter (or slightly less depending on the quality of the oil film between the thrust bearing and the pads). The 12 mil reading refers to a gap voltage of -12.4V. A reading of 12 mils (0.3mm) in the non-working direction should represent the movement of the thrust plate onto the non-working pads (and a sensor voltage of -8.6V).

The advantage of this method is the logical connection between the linear range of the sensor and the range of the counter. This method makes it easier to work with the monitor, since they all use the same assumption (the start of the countdown is in the center of the riser zone). The disadvantage of this method is that different machines have different hot riser zones, so the monitor counter readings will differ. This makes it a little more difficult for the operator to control the units.

Method 2: The start of the countdown coincides with the working side of the riser zone.

Illustration

Fig. 3 The zero point corresponds to the location of the thrust disk on the working side of the bearing
The second method aims to overcome the shortcomings of the previous one. The main objective of this method is to bring the readings of all the axial displacement meters of all the machines to the same value, zero or close to it. This method makes the operator's work easier; attention is drawn only to readings that are very different from zero. However, the disadvantage of this method is the difficulty in setting up the measuring instrument, while the axial displacement monitors of each machine are set up differently.

The procedure consists in setting the size of the hot rise zone and setting the meter so that its zero reading coincides with the fit of the thrust disk to the working side of the bearing (it is desirable to find this location for the hot rise zone) (see Fig. 3). The problem with this procedure is the difficulty in simulating the conditions for the hot rise zone on a non-operating (cooled) machine. Therefore, the meter must be reset to an auxiliary value that can be simulated on a switched off machine.

Using the previous example again, we see that if the hot float is 24 mils (0.6mm) and the cold float is 16 mils (0.4mm), this gives a difference of 8 mils or 4 on each side of the bearing clearance. Thus, with the machine stationary and the thrust washer touching the working liners (cold zone), set the meter reading to 4 mils (0.1mm) on the non-working side. This corresponds to a voltage reading of -11.6V.

When the machine reaches the working condition (with the thrust washer near the working side of the bearing-hot float), the meter reading will be zero or -12.4V of the sensor clearance. Note that the meter will only read exactly "0" if the difference between the hot and cold floats is known exactly. Otherwise, discrepancies will appear.

As we can see, the two methods differ only in the meter readings under normal operating conditions. For normal conditions, the first method results in a meter reading that is different from zero, while the second method tends to zero. It is worth noting the similarity of the two methods. Using either method, the relationship between the axial displacement and the linear range of the sensor remains the same. In both cases, the sensor is rebuilt so that the center of the linear range coincides with the center of the bearing clearance.

A final warning: if a rigid relationship is established between the axial displacement, the sensor clearance, and the meter reading, do not change these relationships, especially after starting the machine. For example, suppose that method 2 was used to set the meter reading to zero for normal operating conditions of the machine. Similarly, suppose that after starting the reading does not coincide exactly with 0, since the size of the hot zone was determined with a small error. In this case, do not recalibrate the meter and sensor for sensor adjustment. If you recalibrate the meter after startup, all initial coupling coefficients will be lost.

A data set for auxiliary quantities can be helpful in the future, especially if the monitoring system malfunctions. For example, if the meter readings change all the time and there is no confidence in the true readings, then it is necessary to verify it. With an initial set of uncorrupted data, any meter reading can be directly translated into the sensor gap voltage and then into the rotor axial displacement. If the meter and gap were recalibrated after startup, then there is no confidence in the accuracy of the meter readings and the exact rotor position.

Adjusting monitor threshold levels.

When setting the threshold levels, it is necessary to realize that when this parameter is reached, it is not necessarily necessary to save the stubborn bearing from damage. The main goal is to prevent serious axial seizure and destruction of the machine. As a result, some wear of the stubborn bearing is permissible under operating conditions. The stubborn bearings usually have a sufficient layer of babbitt, which allows maintaining operability until the onset of axial seizure of the rotor. Thus, it is advisable to allow partial abrasion of the babbitt at the threshold alarm level.

Having realized our considerations and having accepted the concept of the transition of the cold zone of surfacing to the hot zone, we can proceed directly to the purpose of our task - setting the threshold levels. For most monitoring systems, there are 4 warning settings: the first and second alarm levels in each direction of rotor movement (in the active and inactive direction). Set the first alarm level (Alert) above the boundary of the hot zone of surfacing in each direction so that some part of the babbitt is lost before the alarm appears. Set the second alarm level (Danger) so that even the largest part of the babbitt is worn away, but does not allow the rotor to touch the stator part.

For the example shown earlier, the threshold can be set after the loss of 6 mils (0.15mm) of babbitt. 10 mils (0.25mm). levels -15.6 and -4.4V.

Calculating the correct gap of the proximity meters.

Always make sure that the gap in the proximity meter has been set so that it falls in the center of the linear range of the transducer system. and radial vibration.

The center of the linear range is easily found by studying the characteristic curve of the sensor. It can be done using a spindle micrometer type TK3-2 from Bentley Nevada. In the case of axial displacement, incorrect setting of the gap reduces the linear range of the transducer system (Fig. 4).

Illustration

Let’s assume that the sensor has been calibrated from -2.0V. This voltage is sufficient for the monitor to register the transducer as “OK” and measure some vibration. A vibration signal level of 1 mil peak-to-peak, which represents 2V p/p (peak-to-peak), appears as “NOT OK”. This condition does not allow the alarm to be activated, so with this gap setting, “OK” will not provide proper protection for the machine.

In addition, during operation, it is necessary to pay attention to the degree of cleanliness of the axial displacement sensor, this indicator can affect the reliability of the signal coming to the proximeter and monitor.

Protection of a steam turbine from breakdowns using thrust bearings.

Thrust bearings separate the rotating parts of a steam turbine from the stator part (Fig. 5).

Installation diagrams for axial displacement control sensors.
Their touch is very expensive - from $ 500 thousand and more, depending on the size of the turbine. 80% of this amount will be lost production. The best protection is an instrumentation system designed to detect axial displacement with two eddy current sensors mounted in pairs in the inlet section of the turbine. Typically, the sensors are mounted on the shaft end and the driven disk. These two sensors, calibrated for the gap voltage, display the distance from the sensor end to the end of the shaft. The shaft end, depending on where it moves, can move away from the sensor up to 300 mm or less.

Driven bearings fail quickly (less than 30 seconds) because the supporting hydrodynamic oil film loses its ability to hold axial loads. These axial loads reach 30 kg/cm2 on a babbitt casting. The metal temperature in the load zone of about 75% of the central area of ​​the liners reaches 125-140°C.

Commentary on the article "Monitoring axial displacement".

This article has been a revelation for us, dynamic equipment diagnostics employees who use vibration diagnostics methods. It has opened our eyes to many problems that we have to face in the process of work. I believe that it can become a desktop instruction for employees who adjust axial displacement monitoring systems. Especially if components from Bently Nevada are used.

Using vibration diagnostic equipment as a tool for assessing the condition of rotary machines, it is very difficult to confirm or refute the correctness of the rotor axial displacement meter readings. We must either abandon this trick and trust the control system, or go the way of a deeper study of the issues of technology and design features of the machine being inspected. It is changes (violations) of established technological processes and structural defects of the flow part and rotor that can lead to an “unexpected” change in the axial displacement of the compressor or turbine rotor. This assumes that the KVP employees who set up the system have performed their duties correctly.

Over the years of operation of the technical diagnostics center, based on more than 15 years of experience, we have developed an approximate algorithm for solving this issue. This experience is interesting because most often we use a scheme with one axial displacement control sensor, especially not of the highest manufacturing quality.

First of all, I would like to note that this algorithm applies to already working (tested) equipment that has passed the stage of adjustment and experiments.

For newly introduced or experimental equipment, it is necessary to involve the forces of scientific and technical laboratories specializing in these problems.

What list of measures must be performed for vibration diagnostics before making a verdict on the axial displacement problem?

1. Find out the readings of the axial displacement counter.
 The first method of setting the axial displacement reference system described in the article is most often used. In this case, the movement of the rotor in the positive zone means moving it to the working blocks of the thrust bearing and, accordingly, towards the suction side of the compressor or the exhaust side of the turbine (this scheme is most often used). This movement is natural for a properly functioning and centered machine connection. If in the negative direction, then suspicion immediately arises:
- The centering of the machines has changed, which could be associated with a change in temperatures at the suction side of the compressors. In rare cases, when using the end of the half-coupling as a thrust ridge for the sensor, this indicates a malfunction of the coupling assembly (beating, loose fit on the shaft);
- The suction flow rate on the pre-accident machine has changed, which contributed to the deterioration of the operation of the balancing device;
— if the shift change occurs quickly from the positive zone to the negative and back, this clearly indicates incorrect operation of the axial shift sensor. It must be unlocked at your own risk to continue the production program. Although it is necessary to consider the possibility of stopping the machine and replacing the sensor.
2. Check the change in the gap voltage.
If the voltage readings on the meter (Bently Nevada monitors) have changed according to the adjustment schedule adequately to the movement in mm, then this is one (but not the main) reason to require an immediate stop of the machine when the blocking readings of the axial shift are exceeded. Although in our practice there have been stories with oxidation, contamination of contacts on the proximity meter, partial damage to the cable that carries signals from the sensor. After eliminating these reasons, the readings returned to normal.
If the gap readings are -24V or 0, then this clearly indicates a cable break or damage to the sensor (sample).
3. Take vibration characteristics in place and from the monitor (if the device allows) for their comparison.
If you have a computer with vibration trends taken earlier at standard points, you need to compare them and identify changes. At the same time, if signs of rubbing or problems with lubricating oil appear in the vibration spectra of the machine you are interested in, you need to draw appropriate conclusions.
4. Check the records of the machine's technological process control system for changes in the main technological parameters.
Since the technological process provided by the machine under examination occurs within rigidly specified ranges of parameter changes, the deviation of one of them can affect the stable operation of the machine.
4.1. First of all, it is necessary to check the change in the temperature of the lubricant on the inspected thrust bearing. Although very often a thermocouple is installed on the drain from the supporting and thrust parts of the bearing, and the change in the parameter is very difficult to track. At the slightest signs of temperature increase, it is necessary to talk about possible damage to the babbitt of the stubborn pads.
4.2. Check the temperature and steam flow at the inlet and outlet for the turbine. Very often, salting of the turbine blades leads to a decrease in its performance, and thereby an increase in the temperature and flow at the inlet. As a result, this threatens the turbine with an increase in axial displacement towards the exhaust.
4.3. Check the change in flow and temperature at the compressor suction with adjacent housings. Changing these parameters affects the centering and operation of the blower unit. And this, in turn, leads to a change in axial displacement.
5. Check and require the technological personnel to analyze the condition of the lubricant for the content of gases (ammonia) aggressive to the sensor, moisture.
Very often in our practice, the increased content of ammonia and moisture in the oil tank led to a disruption in the operation of the relative vibration and axial displacement sensors. At the same time, we visually assessed the increased moisture content by the presence of condensate on the inner surface of the glass of the lanterns on the oil drains from the bearings. The fact of "soaking" the oil can often be checked in a rather primitive way by smelling the oil on the drainage from the oil tank or oil pump. In this case, the sensor's performance was restored when the oil temperature increased and ammonia was blown out more by nitrogen from the oil tank, and oil regeneration.
Based on the results of the surveys, it is possible to speak with a sufficient degree of certainty about the truth of the rotor axial displacement. In my 6-year practice, the method did not give failures associated with damage to the rotor of the pre-accident machine. Although there were rare cases of machine stops with exceeding all "reasonable" levels of axial displacement, and when disassembling with whole thrust bearings. But this is rather a claim to the stacks that set the gaps without taking into account the hot zone of the floating of the blocks of the stubborn bearing, the thickness of the babbitt on the blocks, the beating of the stubborn disk, when using the method of measuring the axial runout of the rotor using scrap and a digital indicator.

Menu:

Address:

1 Generała Grochowskiego Street,Office 48, Warsaw, Poland.

Phone number:

+38 066 1555747+48 660 871645

Created by: smelov

Privacy Policy

Cookie Policy

Disclaimer

Menu:

Address:

1 Generała Grochowskiego Street,Office 48, Warsaw, Poland.

Phone number:

+38 066 1555747+48 660 871645