Both hardware and software requirements were established for parts used in conveyor belt core diagnostics as well as for signal and statistical analysis module.
a. Hardware requirements:
It was decided that the complete system should be housed in a Peli 1620 box. It should be connected to a magnetic head. Other elements, i.e. the magnetic head, the encoder and the power supply should be connected to one of the box’s walls. Wires of proper length should allow for proper assembly of the system. All system components should be installed so that their proper cooling is possible and should be appropriately protected, so that no damage occurs during transportation, assembly, disassembly or measurements. The upper part of the box should contain a mobile computer, so as to facilitate its removal and connection to peripheral devices. The system’s schematic diagram is shown below.
Schematic diagram of the system
b. Software requirements:
Before performing measurements, each of the 24 channels will be necessarily tested for proper operation. Detection of even a single nonoperational channel should be enough for the software to render measurements impossible. For each of the measurements, it should be possible to enter/generate data enabling its subsequent identification (conveyor number, measurement date, etc.). During inspection the user should be informed about belt speed, its length and measurement duration time, while the system should automatically detect the belt’s beginning and end point. For each of the channels, measurement data displayed in the form of color matrix should facilitate visualization and decision-making processes related to belt damage degree.
The software should generate a report containing:
- identification data,
- signal recording (for 24 channels),
- belt loop identification, selection of belt loops for analysis,
- signal analysis for the selected loop,
- splice identification,
- damage detection (without splices),
- statistical data (damage histograms for belt width and length, number of events per channel and number of events per running meter),
- suggestions on further steps to be taken about the belt.
Neodymium magnets screwed to belt edges in order to allow the software to detect the beginning and the end of the belt.
Color matrix with colors indicating belt damage levels
Gathering information about global solutions for both belt cut and belt cord damage detection, analysis of the results obtained, economical evaluation and selection of best solutions for potential clients. Also, proposal of own solution regarding rip detection and belt cord damage detection.
a. Information about global solutions
The analyzed companies may be grouped according to various technologies offered and solutions used. One of important distinctions to be made is whether the company is independent from the producers/suppliers/companies that offer belt repair services (Beltscan, CBM, CBT, CT) or has connections with such companies (rEscan – Fenner Dunlop i Veyance Technologies - Goodyear). Some companies took their belt inspection technology straight from techniques used for the inspection of steel ropes found in elevators, bridges and cable cars (Intron, TCK, prof. Kwaśniewski).
The vast majority of companies that offer such tools prefer their staff to provide scanning and belt condition monitoring services even on huge distances (they have no local representatives or even no will to have such representatives), and hence they offer remote scanning: CBM, CBT, CT, rEscan. Other companies are ready to sell the equipment and train local staff to use it, without forcing the client to use their services, which may result from lacking in remote control technologies or focusing on local market rather than international (Intron, TCK).
Comparison of the features that those systems offer leads to the conclusion that none of them is open enough to let the client freely use belt condition reports to facilitate rational belt management policy. Most of the systems are hermetic and oriented to own solutions. It can be observed that the most advanced systems and the oldest companies in the market attempt at offering increasingly wide scope of inspection and complex belt condition evaluation based on new systems, yet a lot of information currently stored in databases would be of no use in mines. For example, databases contain information on belt assembly date and repairs performed; they already offer the possibility to record the mass transported or the number of full cycles traveled around the conveyor. Such information may significantly facilitate evaluation or prediction accuracy. Prediction of belt replacement time is performed on the basis of the rate, at which the covers wear off, in relation to calendar time. If it was based on the mass/volume transported or on the cycles performed, it would be more accurate. Belt core’s structure is not the only repaired part. Most frequently repaired element are belt covers. Great number of performed repairs negatively affect the belt’s wear level, yet those repairs are not recorded by the magnetic system. They may be revealed in the machine vision system. If included in repair database, they might greatly increase understanding of current belt condition.
b. Information on the method proposed by Wroclaw University of Technology.
The method for belt cuts and edge damage detection proposed at Wroclaw University of Technology is based on detailed measurements of belt width, which may be carried out in three distinct ways:
- 1 - full machine vision system having a standard belt width measurement option, with belt measurement every 10 mm and simultaneous belt damage detection.
- 2- machine vision system for belt width measurements (only belt cuts and edge damage detection).
The 2D digital signal is measured by a line scan camera in such a way that the load-carrying surface of the conveyor belt is filmed (1) with two digital cameras situated at right angles to the surface.
Schematic diagram of machine vision system for
damage detection and belt width monitoring
After being recorded by the camera, the digital signal undergoes analysis and filtering processes (2), followed by area identification process (3) to identify the area, in which the investigated object (the conveyor belt) is located. Belt width measurement process (6) is carried out by identifying the left 4 and right 5 edges of the belt, regardless of the momentary position of the whole belt on the idler set. Belt width measurement is carried out by defining coordinates for boundary points. Input signal (7) (belt width) is the difference of the coordinates for complementary boundary points.
Change of belt width that exceeds predefined boundaries (alarm thresholds) is an indication of belt rip and is identified as such by the monitoring device. The above described method was disclosed in patent application No. P393273 and is protected by patent titled “Device for local tests of conveyor belts” (published as PRN/WI/451-90/10/13 of 2013.09.02).
- 3 - a system comprising mechatronic sensors installed on both sides of the conveyor belt, which will be described later in this report.
Wroclaw University of Technology proposal of a system for steel-cord belt monitoring and diagnostics.
Each of the presented systems has its advantages and disadvantages. However, it is possible to highlight some features that are important for the user and make the offer varied:
- color damage representation in the form of 2D (2Dk) map,
- high resolution in the belt’s transverse direction,
- low memory usage (LMU),
- access to source data for further processing,
- automatic extraction of data on splices and their condition,
- calculation of safety factor (CSF),
- automatic and integrated cover thickness monitoring and belt life time prediction,
- automatic rip detection integrated with belt condition monitoring system (ASP),
- automatic visual inspection of belt condition (AVI),
- integrated belt life time index (IBLI)
- optimization of belt replacement time based on IBLI(OBRT),
- integration of IBLI and OBRTwith belt management system in the mine (IBMS).