Hansjörg Kapeller

A Detailed Heat and Fluid Flow Analysis of an Internal Permanent Magnet Synchronous Machine by Means of Computational Fluid Dynamics

This paper presents a comprehensive computational fluid dynamics (CFD) model of a radial flux permanent magnet synchronous machine with interior magnets. In the CFD model, the water jacket cooling and a simplified model of the topology of the distributed stator winding are considered. The heat sources of the CFD model are determined from a finite-element analysis of the machine. The numerically determined temperature distributions of the machine are compared with measurement results from sensors located both in the stator and rotor. The particular focus of this paper is the analysis of the temperatures and the heat flow in the air gap and from the stator winding heads and the rotor to the inner air. Different operating conditions and two particular rotor designs with different inner air flow configurations are investigated. The potential of improving the thermal utilization of a rotor design with fan blades attached to the mounting plates of the rotor is shown.

Comparison of a CFD Analysis and a Thermal Equivalent Circuit Model of a TEFC Induction Machine With Measurements

For a totally enclosed fan-cooled induction machine, two methods of numerical analysis are compared with measurements. The first numerical method is based on computational fluid dynamics (CFDs) and the second one uses a thermal equivalent circuit (TEC). For the analysis based on CFD, a 3-D induction machine including housing is modeled. The numeric solution of the flow equations is determined for stationary temperature distributions. For the TEC, a discretized one-and-a-half-dimensional model of the induction machine is considered. With the TEC model, stationary and transient operating conditions can be simulated. Measurement results are determined by iron-copper-nickel sensors embedded in the stator winding and the housing, as well as by an IR sensor for measuring the rotor temperature. With these measurement signals, stationary and transient operating conditions can be analyzed. For stationary operating conditions, additionally, the housing temperatures are determined by an IR camera. The investigated simulation and measurement methods reveal different local and global temperatures, and thus, only certain aspects and characteristics of the obtained temperatures can be compared. Nevertheless, certain conclusions can be drawn from comparing these aspects considering the actual restrictions of each of the applied methods.

Robust Rotor Fault Detection by Means of the Vienna Monitoring Method and a Parameter Tracking Technique

The Vienna monitoring method (VMM) is a model-based rotor fault detection method that utilizes the voltage and current models for the computation of a fault indicator. So far, the VMM was investigated with fixed rotor parameters only. In this paper, the parameters of the current model are provided by a parameter tracking technique. For this advanced rotor fault detection method, measurement results are presented for steady-state and varying load torque operations.

Phenomenon Rotor Fault-Multiple Electrical Rotor Asymmetries in Induction Machines

In the literature the effects caused by a single or several adjacently broken rotor bars, or a broken end ring are thoroughly investigated. The phenomenon of various non-adjacently broken rotor bars has not been studied, so far. Since non-adjacently broken rotor bars may give rise to fault signatures which are not directly related with the fault extent, it is important to understand the nature of multiple electrical rotor asymmetries in induction machines. The purpose of this paper is thus to investigate several combinations of electrical rotor asymmetries to systematically elaborate the phenomena related with broken bars and end rings. In this paper a sophisticated simulation model and measurement results are used to analyze the phenomenon rotor fault.

Detection of rotor faults under transient operating conditions by means of the Vienna Monitoring Method

This contribution investigates the Vienna monitoring method during transient operation. The Vienna monitoring method is a rotor fault detection technique based on two space phasor machine models. These models derive quantities such as flux and torque. An induction machine with rotor asymmetries gives rise to double slip frequency oscillations of shaft torque. These oscillations are also reflected in the modelled torques. Accordingly, the fault indicator of the Vienna Monitoring Method is derived from the difference of the computed torques to compensate load effects. This paper presents measurement results regarding the Vienna Monitoring Method during stationary and transient operating conditions. The transient conditions include varying voltage and frequency as well as varying torque. The two transient cases are a challenge for any rotor fault detection technique, since the magnitudes and frequencies of fault-specific signatures are highly load and speed dependent.

Heat and fluid flow analysis of an internal permanent magnet synchronous machine by means of computational fluid dynamics

This paper presents a comprehensive computational fluid (CFD) model of a radial flux permanent magnet synchronous machine with interior magnets. In the CFD model the water jacket cooling and a simplified model of the topology of the distributed stator winding are considered. The heat sources of the CFD model are determined from a finite element analysis of the machine. The numerically determined temperature distributions of the machine are compared with measurement results from sensors located both in the stator and rotor. The particular focus of this paper is the analysis of the temperatures and the heat flow in the air gap and from the stator winding head and the rotor to the inner air.

Detection of mechanical imbalances during transient torque operating conditions

Mechanical rotor imbalances and rotor eccentricities give rise to specific mechanical vibration, and certain electric, electromagnetic and mechanical harmonics. The extent and location of these harmonics depends on the actual severity of the imbalance, machine supply and load torque. Many fault detection techniques evaluate the spectral components of any of the mentioned electric or mechanical quantities. Such techniques are therefore usually only applicable for steady state operating conditions. To realize an imbalance detection technique working under transient conditions, load dependencies have to be taken into account. In this contribution a technique is used that employs a phase locked loop to track the imbalance-specific harmonics of the electrical power. This technique, therefore, overcomes the slip and load dependency of the imbalance-specific harmonics. The proposed imbalance detection technique is applied to measured data. Results for steady state and transient operating conditions are then compared.

Modeling and Simulation of a Large Chipper Drive

This paper presents a simulation model for a large chipper drive used in a paper mill. If the chipper drive is a slip ring induction motor, several advantages arise from using a rheostat in the rotor circuit. This paper will investigate the impact of a rotor circuit rheostat with respect to starting behavior and heavy duty load impulses. The chipper drive is modeled in Modelica, an object oriented multi domain simulation language. In addition to the theory, a practical case is presented to demonstrate the applicability of the derived knowledge.

Enhanced Thermal Model of a Totally Enclosed Fan Cooled Squirrel Cage Induction Machine

Thermal simulations are a very important tool for optimizing the design of electric machines. In this paper, two thermal simulation models for totally enclosed fan cooled (TEFC) induction machines are compared with thermal measurement results. For both one dimensional models of the induction machine a thermal equivalent circuit of the machine is presented. The enhanced model was built with the possibility to split the middle part into equal sized parts to model a more detailed axial temperature distribution in the machine. For modelling the circuits, MODELICA language was used in combination with the simulation tool DYMOLA. Simulation results are obtained for both thermal models at the thermal equilibrium. For a comparison between simulated and measured results two measurements on an 18.5 kW four pole induction machine have been accomplished. With iron-copper-nickel sensors embedded in the winding, the housing and between the ribs of the machine temperatures have been determined.

Stator winding turn fault detection for induction machines

Stator winding failures are a serious problem of induction machines. In the case of bearing failures or rotor faults, fault-specific signatures increase slowly. An induction machine with such failures still can be operated for a certain time without taking strict precautions. Stator faults may occur much faster due to insulation breakdown or overload and are therefore more critical with respect to time. This paper presents measurement results of an accidentally arising stator winding fault, which was recorded during the investigation of stator asymmetries. Furthermore, an on-line fault detection technique as well as some measurement data applied to this technique, are presented. The used technique is based on extracting a so called injected negative sequence current.