Marcelo Martins Stopa

Incipient Detection of Cavitation Phenomenon in Centrifugal Pumps

The aim of this paper is to present a computer-based tool to aid in the process of detecting cavitation in centrifugal pumps. The method proposed in this paper, namely, the Load Torque Signature Analysis, employs electrical signals from the motor to estimate the torque developed by the pump and determines the occurrence and intensity of the phenomenon through its frequency spectrum information.

Load Torque Signature Analysis: An alternative to MCSA to detect faults in motor driven loads

In this paper, an alternative to the MCSA (Motor Current Signature Analysis) method to detect faults in motor driven loads is presented. The proposed method named Load Torque Signature Analysis - LTSA - processes an estimative of the load torque to obtain a signature of the fault on the load. A comparison of the LTSA and MCSA method is done under the viewpoint of frequency bandwidth, sensitivity and implementation drawbacks. The influence of electrical and mechanical parameters on LTSA performance is investigated. Mathematical relationships among these parameters which allow establishing the nature of the frequency response are obtained. It is shown that, LTSA can be an interesting alternative to be applied in several situations where MCSA performance is poor. Simulation and Experimental results validating the studies are presented.

Charge Behavior Analysis in Ball Mill by Using Estimated Torque

The aim of this paper is to present a method to analyze the dynamic of the charge in a ball mill. The method is based on the mill torque estimated from angular position and motor electrical signals. The estimated torque signal obtained from a Luenberguer velocity observer is bandpass filtered to isolate the mill-filling-related component. The root-mean-square value of this component, used as an indication variable, is correlated to vibration signal measured through an accelerometer in order to check the validity of the proposed method. The proposed method does not need any sensor be installed directly on the mill and requires only sensors usually available for motor control purposes in frequency inverter drives.

Charge behavior analysis in ball mill by using estimated torque

The aim of this paper is to present a method to analyze the dynamic of the charge in a ball mill. The method is based on the mill torque estimated from angular position and motor electrical signals. The estimated torque signal obtained from a Luenberguer velocity observer is bandpass filtered to isolate the mill-filling-related component. The root-mean-square value of this component, used as an indication variable, is correlated to vibration signal measured through an accelerometer in order to check the validity of the proposed method. The proposed method does not need any sensor be installed directly on the mill and requires only sensors usually available for motor control purposes in frequency inverter drives.

A simple torque estimator for in-service efficiency determination of induction motors

Monitoring energy consumption and condition of electric motors is important for economical purposes and safe operation of drives and processes in industry. Those activities are especially valid in processes requiring high availability such as those founded in mining and metals industry. Methods able to implement these functions without process interruption and/or requiring minimum changes on the drive configuration are extremely welcome. The air-gap torque estimator is at the core of the main efficiency estimation methods used today and therefore it is crucial to its successful implementation. Most of the torque estimation algorithms require two or three voltage and current measurement to be implemented. Unfortunately, this number of sensors is not always available for all the motors used in industrial processes. Therefore, a solution employing a reduced number of sensors is necessary. In order to meet this requirement, this paper presents a new and simple technique to estimate torque that is able to be used in algorithms of efficiency estimation of in-service induction motors. The technique requires data acquisition of current and voltage of only one phase for its implementation. Sensors for measuring these variables are usually available in the Motor Control Center — MCC — cabinet for monitoring and protection purposes. From one phase current and voltage samples, instantaneous values of torque, angular velocity, power and efficiency are calculated. Simulation and experimental results are used to check the validity and performance of the proposed technique.

An alternative five level NPC converter for medium voltage AC drives and technical issues

This paper presents a study of an alternative five level inverter for medium voltage ac drives based on neutral point clamped concepts. The main advantage of this topology is its higher voltage operation compared to the commercial three level structures but with less mounting complexity when it is compared to a traditional five level neutral-point-clamped converters. Besides the converter is suitable for various low frequency modulation techniques which are able to reduce switching losses and thus increase the converter efficiency. In this line a detailed explanation of the topology is presented along with a cost comparison among the traditional multilevel topologies based on topological analysis. A description of a five level modulation process for this topology is also presented and in this line a discussion of modulation issues in very low switching frequency is also addressed. The differences and similarities of modulation technique in this study from a classical phase-disposition pulse width modulator are discussed. An important feature of this modulator is its ability to improve the converter harmonic spectrum. All this work is supported by results which are aimed to confirm these ideas.

An evaluation of the MCSA method when applied to detect faults in motor driven loads

In this paper, an evaluation of the MCSA method when applied to detect faults in motor driven loads is presented. The evaluation is based on transfer functions obtained from small signal model of the induction motor. The inherent behavior of the induction motor as a torque transducer is explored to explain some of the characteristic features of its application to monitoring the driven load connected to its shaft. The influence of the electrical and mechanical parameters on its performance as a load fault detector is investigated. Mathematical relationships among these parameters which allow establish the nature of the frequency response are obtained. It is shown that, although the MCSA (Motor Current Signature Analysis) is a simple and effective method to detect faults on the driven load, care must be taken on its usage. This is due to the non-ideal transducer behavior of the induction motor. Simulation and Experimental results validating the studies are presented.

On-Line Torque and Drawing Force Estimation in Wire Drawing Process from Electric Motor Variables

In this paper, an alternative on-line drawing force and torque estimation method for use in wire drawing processes is presented. The method considered here is based on a simple Luenberguer observer and uses only measured variables available from a motor drive system such as terminal voltages and currents and rotor velocity. The proposed algorithm permits to implement functions such as condition monitoring of die wear, work piece fracture and lubricant analysis. The performance and some practical implementation features of the load torque estimator are analyzed. Theoretical development and simulation results are presented to support the use of such method on load torque and drawing force estimation

Study and implementation of a high gain bidirectional dc-dc converter for photovoltaic on-grid systems

The aim of this paper is to present a bidirectional dc-dc converter with high-stepping ratio for photovoltaic systems operating in parallel arrangement operating with energy storage systems. The converter utilizes a resonant LC circuit to continue provide an increasing voltage at the output terminals by rotating the polarity of the capacitor using a thyristor based H-bridge. Compared with the existing converters, the proposed circuit reach higher voltage gains with low switching frequency using simple circuitry. It works efficiently with minimal switching loss even dealing with high currents. The prototype has shown the usefulness of the converter for low input voltage applications, mainly when a string of photovoltaic panels is connected in parallel, with the possibility to feed in a commercial inverter and battery banks comprising a dc bus bar.

Bidirectional power converter with adaptive controller applied in direct-current microgrid voltage regulation

Recently, many researches have been done related with the integration of distributed generation in the distribution system. However, in large scale applications, there are several challenges in terms of control design. A simple example is the possibility to operate in island mode or connected to the main power. The operation in island mode requires that power sources should be inserted inside the microgrid structure. The use of dc microgrid is increasing because of advantages in the integration of dc charges, dc loads and storage systems through electronic converters. Between theses advantages, the elimination of frequency are reactive power control are highlighted. However, this simplification does not exempt the microgrid from stability problems, in order to achieve different voltage levels for several power devices. These converters introduce destabilizing effects on the system, leading the microgrid to show significant oscillations in the dc bus voltage. The converter control is the primary concern in operating a microgrid. This work presents a dc microgrid operating in island mode. An adaptive controller is designed to control the balancing of power through a dc-dc bidirectional converter. This converter is connected to a storage device, controlling the dc bus voltage. Simulation results allow us to validate the proposed controller under variations in the availability of energy sources and during load changes.