Introduction

This application note describes the use of MAARS Model 3000 on-line monitoring system for an Overhead Crane Monitoring application.  The application is implemented at an aluminum roll manufacturing facility on the Continuous Cold Mill Exit Crane.  Every pound of aluminum that is produced by the facility must be handled by this crane, so it is critical to the roll production process. A preventative maintenance program was established for the crane, but failures of the crane components were still of concern. The crane is inspected manually by the operator on a time basis, and yearly by a contract crane rebuild shop. Safety concerns preclude the use of a portable data collector to acquire vibration data from the crane while it is running. The MAARS Model 3000 Monitoring System was permanently affixed to the crane in April of 1999. Sensors are hardwired along the festoon from the crane components to the Model 3000. Wireless spread-spectrum data transmission is used to send the vibration data down from the crane to the plant Ethernet network, where the data can be accessed by any computer on the company network.

Several analysis parameters were defined to aid in diagnosis of internal problems in the hoist drive and gearbox.  Alarm limits were assigned to the parameters to provide predictive and diagnostic capability.  Because the Model 3000 is a simultaneous data acquisition system, all six sensors are monitored on every lift cycle of the crane. In order to provide trendable data, the sensors only write data to the database during the “valid RPM” part of the lift cycle, even though they are continually monitoring the crane at all times. After successful demonstration of the addition of vibration analysis to the Continuous Cold Mill Exit Crane, the manufacturer has requested that MAARS develop this application note to be published in the company-wide Reliability Newsletter.  

System Architecture

The MAARS Model 3000 Crane Monitoring Unit is a complete industrial PC in a NEMA 12 enclosure designed to be installed in a factory environment. Although not used during routine testing, a keyboard, monitor and mouse can be connected to connectors located on the side panel of the unit.  The 3000 unit has an Intel Pentium^TM II processor running at 166 MHz with a 4-gigabyte hard disk.  The unit runs on the Windows^TM 95 operating system, which greatly simplifies system maintenance.  All networking and communications is handled by the operating system.  Plant Information Technology personnel will be working with a familiar platform because of the Windows^TM 95 operating system.

PC-Based Server

The optional server for the Crane Monitoring system is a standard PC running Windows^TM 95 or Windows^TM NT.  It is connected to the plant Ethernet via a hub and standard cabling. As lift cycles are performed by the crane, the monitoring results (trends, waveforms and spectra) are downloaded to the server.  If the network connection is not available when a lift cycle is completed, the results are cached at the 3000 test unit until the network becomes available.  Thus, the crane monitoring unit can continue to operate in a stand-alone mode.  Using the MAARS PathFinder^TM  software, a variety of reports can be generated from the monitoring system’s data stored in the master database.

Crane Monitoring Process Description

The following section describes the steps that take place during a typical lift cycle on the crane.

1. The Model 3000 Crane Monitoring System is monitoring all of the sensors continuously and simultaneously.

2. An aluminum roll comes into the crane off the assembly line, transported on a carrier.  The crane operator moves the hoist into position to lift the roll.

3. Next, the operator starts the lift cycle by hoisting the roll.

4. The 3000 unit monitors the tachometer for a target RPM that is specified in the database.

5. Once the target RPM is reached, the Model 3000 begins storing data to the database for trending purposes.

6. After the data acquisition is complete, the 3000 unit analyzes the data and computes all the parameters specified in the database.

7. Next, the 3000 unit compares the results of the analysis parameters to the alarm values specified in the database.  If any of the alarm values are exceeded, the Model 3000 generates an alarm message over the network to the maintenance team’s PC.

8. All the data are then transferred over the network to the master database on the server.  If the network is not available, the data and results are cached on the local hard disk of the 3000 unit until the network is available.

9. If desired, the server can send a failure report to a printer on the network with pertinent crane fault information.

10.The Model 3000 Crane Monitoring System, still monitoring all channels and sensors simultaneously and continuously, repeats this process for the roll lowering part of the lift cycle.

Crane Monitoring Challenges

·         Crane Access – The Continuous Cold Mill Exit Crane is a critical crane to the aluminum roll production process, so access to the crane was very limited. This crane has a scheduled downtime of 4 hours per week for maintenance activities, and many weeks even this window of opportunity for crane access was closed due to production requirements. Once installed, the monitoring system must be able to be accessed entirely over the network. We installed a third-party software package called pcAnywhere^TM that allows complete control over the Model 3000 unit over the network as though we were physically connected to the Model 3000.

·         Power Requirements – The Continuous Cold Mill Exit Crane has only DC power available on the crane. The Model 3000 was designed to be powered from 110 volts AC. The solution was to power the Model 3000 with a DC to DC converter, using the DC power available on the crane.

·         Slow or Variable Speed Data Acquisition – Order tracking technology, along with “Valid RPM” machine states were utilized to overcome this problem. Because all channels are acquired simultaneously, data from all sensors are acquired for each lift cycle. In fact, several averages of the data can be acquired for each lift cycle. Shown here is a sample RPM gauge that has a green area that indicates the speed range at which the drive motor for the crane must be in for the system to acquire data. Multiple speed ranges may be defined as “Valid RPM” ranges, if required.

·         Environmental – The monitoring system must be able to withstand impacting from unloading the hoist load and running into the crane travel stops. As you can see from the graph of an unloading event shown here, up to 10 or 12 g’s of force are routinely experienced by the crane as the operator unloads the hoist. The Model 3000 was designed for a maximum instantaneous load of 30 g’s.

·         Communications – Once the data is acquired, it must be transmitted from the crane by some method that is wireless. Spread-Spectrum data transmission has high noise immunity, while maintaining a high throughput rate for the data. Early on, some concerns were raised that the wireless communications being used by the crane operator to control the crane might conflict with the frequencies used for the monitoring system. These fears were unfounded as both systems use different portions of the frequency spectrum for transmission. Wireless Ethernet modules are both inexpensive and have a data transmission rate of approximately 1.6 megaHz, much faster than modem communications. In addition, the transceivers may be placed up to 300 feet apart in the plant, and repeaters may be employed if more distance is required. In addition, only one receiver is required for up to eight cranes using the Model 3000, each with a transmitter attached.

·         Slow Speed RPM Detection – The Model 3000 uses a “universal tachometer” SmaartBrick^TM  tachometer signal conditioner that can acquire a signal from 30 mV to 120 V and output a clean TTL square wave for the Model 3000 A/D card to use for RPM data. A magnetic RPM sensor is used to sense the machine RPM down to very low speeds (about 1 RPM). This sensor has the advantage of not requiring any power to sense machine RPM, so no external power from the Model 3000 is required for any sensor or tachometer input.

·         Program Effectiveness - The new system is cost-effective and presents the crane operator with no new responsibilities. Because the data is transferred via the plant network, the maintenance technicians can monitor and analyze the crane from the maintenance office PC.  

Application of Vibration Technology

Vibration analysis of production cranes has proved to be very successful in reducing the number of unplanned crane outages at the aluminum production facility. The concept of using vibration analysis was begun at the facility using a portable data collector for general machinery, but the cranes could not be monitored due to personnel safety concerns.

Vibration analysis programs have been utilized in many plants for identifying degrading machinery as part of a predictive maintenance program. Typically, these programs trend data from a large population of machines, looking for imbalance, misalignment, looseness, shaft, bearing and gear defects, foundation problems and electrical defects in motors. These same problems may now be identified and corrected before catastrophic failure on overhead cranes using the Model 3000 Crane Monitoring System.

For example, gear defects in the crane hoist gearbox proved to be one of the most easily identifiable defects. One can see from the photograph and data shown below that this defect is easily identifiable through spectral analysis.

One critical area of Predictive Maintenance that is often overlooked is the reporting function. The Model 3000 Crane Monitoring System has the ability, through the PathFinder^TM Software package, of generating HTML formatted reports. These reports can be easily transferred via email over the company network to other personnel. In addition, copy and paste functions in PathFinder^TM  are completely compatible with the Microsoft Office^TM Software Suite, including Word^TM, Excel^TM and PowerPoint^TM. All of the data graphs presented in this paper were directly copied to the Windows^TM Clipboard and pasted into Word^TM ’97. An example HTML report is shown here.

Conclusion