ABB Millmate Rolling Force Measurement and Control System On site Troubleshooting and Maintenance Guide

ABB Millmate Rolling Force Measurement and Control System On site Troubleshooting and Maintenance Guide

In modern steel rolling production, accurate measurement of rolling force is the core link to ensure plate thickness tolerance, optimize roll gap setting, monitor the status of support roll bearings, and compensate for roll eccentricity. ABB's Millmate rolling force measurement system is based on the classic Presductor ® The principle of magneto elasticity, patented in 1954, has become the industry benchmark for cold and hot rolling mills worldwide due to its durability, high overload capacity, and excellent signal-to-noise ratio. The system mainly consists of Millmate Controller 400 (MC 400) controller, matching unit (PFVO 142/143 or QIPZ 127), and force sensors installed in the rolling mill.

However, even such a reliable system, after years of operation in harsh environments such as high temperature, high humidity, vibration, and dust, can still experience problems such as signal abnormalities, zero drift, and even sensor failure. When the original model is facing production stoppage or the spare parts procurement cycle is too long, engineers often need to complete fault diagnosis, model confirmation, and precise replacement in a very short time. This article is based on official technical information from ABB, combined with on-site maintenance experience, to provide a detailed breakdown of common fault phenomena, diagnostic steps, sensor selection and replacement process, and controller parameter debugging methods for the Millmate rolling force system. It provides a "hands-on" operation guide for maintenance personnel in steel rolling production lines.

Chapter 1 Core Technology Review: Presductor ® Magneto elastic principle

The core of Millmate rolling force sensor is Presductor ® Technology utilizes the magneto elastic effect of ferromagnetic materials - when mechanical stress is applied to the material, its magnetic permeability changes. There are four holes in the sensor body, which are respectively wound with primary coils (excitation) and secondary coils (measurement) that are perpendicular to each other. When there is no load, the magnetic flux of the two coils is orthogonal, and there is no induced voltage in the secondary; After loading, the magnetic permeability in the stress direction decreases, the magnetic permeability in the vertical direction increases, the magnetic field lines deflect, and the secondary coil induces a voltage proportional to the load.

Key performance advantages:

High overload capacity: The maximum allowable load can reach 700% of the rated value without mechanical damage; Performance data will not permanently change under 300% load. This ensures that even in the event of abnormal working conditions such as steel jamming or overlapping, the sensors of the rolling mill remain reliable.

High signal-to-noise ratio: The sensor output signal can reach up to 500mV/V, far higher than traditional strain gauges, and is extremely insensitive to insulation defects (the measurement accuracy can still be maintained even when the insulation resistance drops to 10k Ω).

No preload requirement: Unlike pressure magnetic sensors, Presductor can obtain reliable signals without applying preload, simplifying installation.

Redundant design: Each standard sensor contains 1500-2000 independent magneto elastic units, which can output the true average force value even if there is uneven local force.

Chapter 2 Identification and Selection Replacement of Force Sensor Models

There are four series of Millmate rolling force sensors, corresponding to different installation positions and load ranges. When facing sensor damage or discontinuation for replacement, it is necessary to accurately identify the prototype number and select a suitable substitute.

2.1 Circular Sensor PFVL 141C

Installation position: Under the mill screw or hydraulic cylinder, above the bearing seat of the upper support roller.

Features: Made from square steel billets, with stainless steel shrink rings on the outer side to protect the coil. The core diameter is based on a module of 30mm, with a total of 23 standard specifications covering 2.0 MN to 60 MN. It is suitable for rolling mill windows with ample space.

Size and load correspondence (excerpt):

Rated load (MN) Inner diameter ID (mm) Outer diameter OD (mm) 2 sensors Maximum cable length (m)

1.6 150 210 30

5.0 270 320 28

10 360 410 24

20 510 560 22

40 720 800 13

60 810 890 10

Replacement point: When selecting the rated load, take the next standard value that is not less than the original load. For example, if the original sensor is 18MN, the 20MN specification should be selected. Note that the outer/inner diameter must match the original installation flange, otherwise mechanical modification is required.

2.2 Rectangular sensor PFVL 141V

Installation position: Below the bearing seat of the lower support roll, there should be a sufficiently large flat surface for the lower crossbeam of the rolling mill.

Features: The length and width are measured in modules of 30mm, and when the length exceeds 900mm, it is measured in modules of 60mm. Covering 0.63 MN to 60 MN. There is no fixed size chart, and it needs to be calculated based on the required load and installation space.

Calculation formula:

L×W×zero point zero zero zero one=F nom (MN)

L×W×0.0001=F nom(MN)

Where L is the length (mm) and W is the width (mm). For example, if the rated load is 14MN and the given width is 370mm, the required length L=14/(0.0001 × 370) ≈ 378mm, rounded upwards to 390mm (multiples of 30mm).

Replacement reminder: Rectangular sensors are usually integrated into the Load Cell Package, which includes a lower pressure plate, sensor frame, and upper pressure plate. It is recommended to replace the pre tightening force pack as a whole during replacement to ensure even force distribution and easy installation.

2.3 Circular Sensor PFVL 141R

Installation position: between the rolling mill nut and the frame, suitable for situations where there is insufficient space under the grinder screw and bearing seat.

Features: Stainless steel laminated winding, outer stainless steel ring protection. The standard specifications cover 2.0 MN to 28 MN. The size parameters D2 (inner diameter) and D3 (outer diameter) are the key to selection.

Size and load correspondence (excerpt):

Rated load (MN) D2 (mm) D3 (mm) 2 sensors Maximum cable length (m)

2.0 100 200 30

6.3 285 420 25

10 385 525 24

16 480 660 22

28 620 850 17

2.4 Small circular sensor QGPR 104/102

Application: Suitable for situations where the space is extremely small or the load is less than 2MN, such as steel pipe rolling mills and small vertical rolling mills. Covering 0.1 MN to 2.5 MN. QGPR 104 is used for 0.1~0.63 MN, and QGPR 102 is used for 1.0~2.5 MN.

Replacement precautions: The matching device for the QGPR series is QIPZ 127, which is not compatible with the PFVO 142/143 used in the PFVL series. The controller model also needs to be matched: PFVA 401 series is used for PFVL, PFXA 401 series is used for QGPR.

Chapter 3 Common Fault Phenomena and On site Diagnosis

3.1 No rolling force signal or abnormally low signal

Phenomenon: The operator station displays a rolling force of 0 or much lower than the normal value, and it does not change with the rolling process.

Possible reasons:

Matching unit or sensor cable disconnection: Any disconnection in the circuit from the sensor to the matching unit and then to the controller.

Controller channel damage: Corresponding analog input channel failure.

Internal coil circuit of sensor: caused by overheating or mechanical shock.

Diagnostic steps:

Check the indicator lights on the matching unit. During normal operation, there should be a power indicator (usually a green LED).

Measure the resistance of the sensor coil (primary and secondary) with a multimeter. The resistance value of PFVL series is generally several tens of ohms. If it is infinite, it will break internally.

Check the 'Load Cell Test' function in the diagnostic menu of the controller. This function can individually activate each sensor and display the original signal value. If a channel displays 0 or close to 0, the sensor wiring can be exchanged to confirm whether it is a sensor or channel issue.

Check the controller's self diagnostics. If the message 'Transformer test failed' is reported, it indicates a sensor or matching unit malfunction.

3.2 Rolling force signal jumping or unstable

Phenomenon: The displayed rolling force value fluctuates up and down, and there is significant noise even when the rolling mill is unloaded.

Possible reasons:

Poor grounding or electromagnetic interference: The shielding layer is not grounded at one end or is parallel to high current cables.

Mechanical looseness: The sensor installation pressure plate is not tightly pressed, or the gap between the rolling mill window is too large, causing the sensor to bear lateral force.

The filtering time of the matching unit or controller is set too short.

Diagnostic steps:

Observe the analog output of the controller (0~± 10V or 4~20mA) with an oscilloscope, and check the noise frequency and amplitude. If it is high-frequency noise, focus on checking the shielding grounding.

Enter the controller parameter menu and increase the filter time (which can be increased from the default value to 2000ms) to observe the improvement in stability.

Check the mechanical clearance of the rolling mill. Connect the sensor signal to diagnostic mode, load it with a jack or pressure head when unloaded, and observe whether the signal is linear and whether there is any jumping. If there is a step like change in the signal during loading, it indicates poor contact between the internal coil or matching unit of the sensor.

3.3 Severe zero drift

Phenomenon: When there is no load, the displayed value is not zero and slowly changes with temperature or time.

Possible reasons:

Temperature compensation failure: The original sensor had temperature compensation components inside, which drifted after long-term aging.

The sensor is subjected to excessive lateral or torsional forces, resulting in residual stress.

Controller zero setting error or drift.

handle:

First, perform the "zero calibration" function of the controller (usually in the menu of Operator Unit 410). Perform zero point adjustment in an unloaded state.

If the zero point is still unstable, check if the sensor installation is tilted. Measure the upper and lower contact surfaces of the pressure plate and sensor with a spirit level, allowing for a tilt of ≤ 0.1mm/m.

For the PFVL series, the zero temperature drift index is ± 0.01%/℃. If the ambient temperature changes by more than 20 ℃, resulting in a zero point change of more than 0.2%, it is possible that the sensor temperature compensation component is damaged and needs to be replaced.

3.4 Overload or permanent damage

Phenomenon: The sensor output signal no longer changes linearly with the load, or there is a jamming phenomenon.

Reason: The actual rolling force exceeds 300% of the sensor's rated value (permanent data change) or 700% (mechanical damage). Commonly seen in severe steel jamming, overlapping rolling, or incorrect setting of rolling force.

Judgment: Use a known weight or hydraulic calibration device to calibrate the sensor. If the nonlinear error exceeds ± 1% (PFVL accuracy level is ± 0.5%) and the return difference is inconsistent after repeated loading/unloading, the sensor is damaged and must be replaced.

Chapter 4: Complete Process for Replacing Force Sensors

When it is confirmed that the sensor is damaged and the original model has been discontinued, please follow the steps below to replace it.

4.1 Record the key parameters of the original sensor

Search for existing equipment files or directly measure:

Rated load (MN)

Sensor type (circular/rectangular/circular)

Key dimensions:

Circle: outer diameter OD, inner diameter ID

Rectangle: Length L, Width W

Ring shape: inner diameter D2, outer diameter D3

Installation method: Does it come with a pre tensioning package

4.2 Select alternative models from the standard series

If the original load is 10MN, circular: PFVL 141C 10MN (OD=410mm, ID=360mm) can be directly selected. If the dimensions do not match, similar load specifications can be used, but it is necessary to ensure that the mechanical interface is compatible (transition flanges can be machined).

If the original load is 14MN, rectangle: Determine the size according to the formula L × W × 0.0001=14. If the original length is 420mm and the width is 330mm, then 420 × 330 × 0.0001=13.86, slightly less than 14. It is recommended to choose 440 × 330 or 420 × 340 (rounded).

For rings: it is necessary to ensure that D2 and D3 are completely consistent with the original mounting seat, otherwise the frame nuts or covers need to be reworked.

4.3 Matching Unit Replacement

Each PFVL 141 sensor requires a matching unit PFVO 142 (2 sensor scenarios) or PFVO 143 (4 sensor scenarios). The matching units are interchangeable, but it is recommended to use the same model when mixing old and new. The QGPR sensor requires the use of QIPZ 127 matching device.

Attention: The cable length of PFVO 142/143 is limited by the sensor load. For example, for a 20MN sensor, if 2 sensors are used, the maximum cable length is 22m; if 4 sensors are used, the maximum length is only 12m. When replacing, if the on-site cable length exceeds this value, the cable needs to be shortened or the number of matching units needs to be increased.

4.4 Cable Connection and Wiring

The cable between the sensor and the matching unit should be a specialized shielded cable provided by the manufacturer. It is recommended to use connector cables instead of fixed cables for easy replacement.

The shielding layer is grounded at one end on the matching unit side.

Avoid parallel wiring with power cables (>380V), with a minimum spacing of 300mm.

4.5 Controller Configuration

Enter the configuration menu of MC 400 through Operator Unit 410 or Ethernet VIP/TCP.

Select 'Measurement mode' as the standard mode (e.g. 'Roll force -2 load cells' or' 4 load cells').

Enter the rated load value (in MN) and type (PFVL 141C/V/R or QGPR) of each sensor.

Set filtering time. Recommend an initial value of 200ms and adjust according to actual stability.

Configure analog output: Select the signal type (0~± 10V or 4~20mA) and set the output range corresponding to the maximum force value (e.g. 0-10V corresponds to 0-20MN).

Perform zero calibration (press' Zero 'when unloaded).

Conduct simulation testing: Use Simulation mode to generate simulated force values within the controller and check if the display on the upper computer matches the simulated output. This helps verify whether the system integration is correct without the need for actual rolling.

4.6 Verification after replacement

Use a portable power source (such as a hydraulic jack+standard pressure sensor) to apply a known force to the rolling mill and compare it with the Millmate display value. The deviation should be within ± 0.5%.

Perform repetitive testing: Three loading unloading cycles, with a return error (lag) of less than 0.2%.

Record the new sensor serial number, replacement date, and configuration parameters, and store them in the equipment file.

Chapter 5 Maintenance and Self Diagnosis of Controller MC 400

The MC 400 controller is the brain of the system and supports multiple communication methods (analog, digital, Ethernet VIP, Profibus DP options). Common on-site malfunctions and their solutions:

5.1 Controller cannot start or restarts repeatedly

Check power supply: 85-264VAC, 100 (-15%)~240 (+10%). Low voltage can cause a restart.

Power consumption: 650VA in PFVL configuration, 140VA in QGPR configuration. Ensure sufficient capacity of UPS or power supply line.

The aging of internal batteries (used for storing configurations) may result in parameter loss. If replacing the controller, it is necessary to re download the configuration file using VIP or RS-232.

5.2 Communication interruption (Profibus or Ethernet)

Check if the terminal resistance of the Profibus connector matches (end to end ON, middle OFF). The highest baud rate is 12Mbit/s.

Ethernet communication: MC 400 serves as the server and adopts a speed of 10Mbit/s. Ensure that the switch port is set to 10M full duplex or auto negotiation.

View controller status LEDs: RUN, ERR, COM, etc. Refer to the code table in the manual.

5.3 Abnormal Analog Output

Measure the voltage/current of the analog output terminal. If the output is zero or full, check the output scaling settings in the controller.

For current output (0-20mA or 4-20mA), the load resistance should not exceed the specified value (usually ≤ 500 Ω).

If higher insulation withstand voltage is required, the insulation amplifier PXUB 201 (voltage or current output) can be used.

Chapter 6 Preventive Maintenance and Life Management

To maximize the fault free operation time of the Millmate system, it is recommended to implement the following maintenance plan:

Cycle content

Check the LED status of the matching unit and controller every week; Record the zero point value without load and monitor the trend

Clean the iron filings and cooling water around the sensor every month to ensure IP protection level (working temperature -10~+90 ℃, short-term 110 ℃)

Quarterly simulation output linearity test (three-point method: 0%, 50%, 100% rated load), deviation should be<0.5%

Perform full-scale calibration of the sensor annually using a standard power source; Check for corrosion on the cable shielding layer and joints

Consider replacing the electrolytic capacitors in the matching unit every 5 years (if any); If the zero drift trend of the sensor exceeds 0.1% per year, it is recommended to send it to the factory for maintenance

Suggestions for spare parts:

Reserve at least one matching unit of the same model (PFVO 142/143) and one set of sensor cables.

For old rolling mills, record the size and load of each sensor. In case of production stoppage, contact ABB for customization (non-standard sizes still need to be calculated as F=(π/4 × (D3 ² - D2 ²)) × 0.0001 MN).

Chapter 7: Case Study: Severe Fluctuations in Rolling Force Readings of Hot Rolling Mill

Background: The rolling force signal of the upper roll of the roughing mill R1 on the 1780 hot rolling mill in a certain factory fluctuates by ± 10% during the rolling process, making it impossible to perform automatic thickness control.

Troubleshooting process:

Observe that the displayed value of the controller fluctuates at a frequency of about 2-3Hz, which is independent of the main motor speed of the rolling mill (about 600rpm), and eliminate transmission interference.

Enter diagnostic mode and view the raw signals of four sensors (two operating sides, two transmission sides). It was found that the upstream sensor signal on the operating side was significantly jumping.

On site inspection of the matching unit of the sensor revealed that the shielding layer at the cable joint had corroded and broken, and the temperature of the matching unit housing was as high as 75 ℃ (close to the upper limit of 90 ℃).

Replace the matching unit and cable, and reconnect the shield. Zero point restoration of stability under no-load conditions.

Conduct simulation testing to ensure linearity of the signal. Test rolling on the machine, the fluctuation decreased to ± 1%, and returned to normal.

Conclusion: Damage to the shielding layer leads to electromagnetic interference coupling into the measurement signal. Regular inspection of cable connectors and shielding grounding is crucial.