RICERCA ITALIANA     

Thermal Aspects in Design and Control of Innovative Electrical Drives

Politecnico di Milano

Abstract

The research is aimed to analyze the thermal and the electromagnetic design aspects of the electrical machines and drives of various sectors, with the following purposes: improvement of the thermal modeling of components, development of design tools, forecast of the material thermal stresses, optimization of cooling, operation, and systems control.
Directly exploitable results will come in the following sectors: industrial processes; civil and commercial sectors; aerospace applications; propulsion systems.
Various types and ratings will be considered, in the manufacturing, consumer, transportation and services areas. Considering the increasing competition needs, and the concomitant increasing materials exploitation, the availability of suitable tools of knowledge and control of the thermal aspects is strategic: in fact, the project will be adequately sustained, also economically, from the industrial world.
This organic university-industry interaction will allow to join the theoretical approach with the experimental and manufacture aspects, with a strong orientation towards products and shared objectives. This fact will answer to the Italian industry problems, challenged to the product quality progress, at once containing the costs, in a global market where innovation is an absolute need.
The research program will be aimed to the development of thermal analysis methodologies for design and control purposes of motors, converters and drives, with simulation and test validation activities. The wide range of ratings and component types requires specific competences, with a choice of the Units based on equilibrium and complementarity criteria. One Unit will spend its activities on TEFC motors for industrial and aerospace use; a second Unit will study the rotating and linear permanent magnet (PM) machines with concentrated windings; a third Unit will analyze the propulsion and generation electrical drives for hybrid road vehicles; a fourth Unit will consider the electromagnetic and thermal aspects of machines with non conventional magnetic materials.
A coordinated and efficient use of the human and financial resources will be guaranteed, avoiding overlappings. As regards the expected final results, including development and software implementation of thermal models, in the following a list of analyzed problems and applications is given.

Self-cooled industrial induction motors
Thermal modeling and critical elements of the thermal networks.
Study of the endwindings-frame thermal exchange and analysis of radiation.
Thermal models for self-cooled industrial motors and for actuators in aerospace applications.
Thermal tests on motors of ratings up to 55 kW and comparison with the simulations.

Rotating and linear PM with armature concentrated windings
Thermal implications of the peculiarities of the machines with concentrated windings.
Calculation of main losses and of additional losses caused by the harmonic fields.
Study of the cooling and development of thermal networks for the steady-state and transient analysis.
Realization of prototypes and activity of experimental validation.

Electrical drives for hybrid propulsion road vehicles
Steady-state and dynamical thermal networks of PM electrical machines with closed-circuit liquid cooling.
Analysis of cooling solutions for static converters with modular structure.
Control of propulsion electrical drives, with predictive thermal analysis.
Prototype production and test activities.

Electrical machines using non conventional magnetic materials
Electrical and thermal characterization of non conventional magnetic materials
Setting up of Finite Element simulation models of machines with novel magnetic materials
Analysis of thermal flows and convective flows in different machines portions
Development of analytical models and setting up of design criteria
Realization of prototypes and experimental activities <<<

Principal Investigator

Antonino DI GERLANDO Politecnico di MILANO

Research Objectives

Foreword
The sensibility to the thermal problems of electric machines, static converters and electric drives has progressively grown recently, for the following, main reasons:
- the tendency to improve components and systems performances (compactness and low cost, motor torque/mass or torque/inertia ratios, efficiency, response promptness) has made important the thermal behaviour knowledge and the adoption of measures suited to limit the operating temperatures [1];
-the evolution in the material technologies and in the design methodologies stimulates to adopt a suitable approach also on the thermal aspects of the design, for an effective pursuit of optimal results [2];
-the development of electromechanical devices and of conversion apparatus with non conventional structural configurations and/or employment conditions asks to carefully examine the thermal exchange and storage situations, with the development of test validation activities and of suitable models [3];
-the increasing availability of Finite Element (FEM) codes for thermal and/or fluid-dynamics analyses certainly represents a key design tool, but it does not eliminate the need to develop suitable models based on a more synthetic circuit approach, more suited to quick parametric analyses [4];
-the development of accurate equivalent thermal networks of the devices necessarily requires to be accompanied by experimental tests on adequately instrumented prototypes, in order to reduce the uncertainty in the thermal parameters knowledge [5];
-the increasing demands of more quick and reliable thermal models of components and systems for control and diagnostic purposes has pushed to select simplified thermal networks, whose parameters can be identified by tests; this approach allows the temperatures monitoring with "thermal image" methods [6];
- the availability of non traditional magnetic and insulating materials has evidenced the need to revise the traditional design procedures, on the basis of their different thermal properties [7].

Bibliography References
[1] F. Caricchi, F. Crescimbini, A. Di Napoli, M. Marcheggiani, "Prototype of Electric Vehicle Drive with Twin Water-Cooled Wheel Direct Drive Motors", 27th Annual IEEE PESC Conf. Proc., Baveno (Italy), 23-27 June 1996, Vol. 2, pp. 1926-1932;
[2] D. Staton, A. Boglietti, A. Cavagnino, "Solving the More Difficult Aspects of Electric Motor Thermal Analysis", IEMDC'03IEEE Intern. Electric Machines and Drives Conf. 2003, 1-4 June 2003, Madison, Wisconsin, USA;
[3] B. Alvarenga, I. Chabu, J. R. Cardoso: "Thermal Characterization of Long Electrical Devices — Application to a Tubular Linear Induction Motor"; 2003 IEEE International Electric Machines and Drives Conference, pp.942-938;
[4] A. Di Gerlando, I. Vistoli: "Thermal Networks of Induction Motors for Steady State and Transient Operation Analysis"; Conf.ICEM'94, Paris, France, 5-8 Sept. 1994, pp.452-457 ;
[5] G. Henneberger, K. Ben Yahia, M. Schmitz: "Calculation and Identification of a Thermal Equivalent Circuit of a Water Cooled Induction Motor for Electric Vehicle Applications"; Electrical Machines & Drives Conf., 11-13 Sept.1995, pp.6-10
[6] A. L. Shenkman, M. Chertkov: "Experimental Method for Synthesis of Generalized Thermal Circuit of Polyphase Induction Motors"; IEEE Transactions on Energy Conversion, Vol. 15, No. 3, September 2000, pp.264-268
[7] YouGuang Guo; Jian Guo Zhu; Jin Jiang Zhong; Wei Wu; "Core losses in claw pole permanent magnet machines with soft magnetic composite stators", IEEE Transactions on Magnetics, Volume: 39 , Issue: 5 , Sept. 2003, pp.3199 – 3201

Targets
The main objective of the research is the development and the aggregation of theoretical and application know-how of the academic world, for the enhancement, the identification and the implementation of thermal models: the aims are design and control of electrical machines and drives in applications of various sectors (industrial, civil, aerospace and propulsion).
The project is characterized by a strategic collaboration with industries interested to the improvement of the components performances. Besides the scientific value of the obtainable results, a remarkable importance is assumed by the applicative consequences of these results for the electrical machines and drives manufacturers.

Even if the principal objective can be distinguished in partial goals, the following common points should be evidenced:
1. analysis of the thermal exchange conditions in the different structural configurations and operating situations of electrical machines and of power electronics converters;
2. thermal characterization of the materials: dielectric and thermal effects of insulating material aging; thermal effects of demagnetization in the permanent magnets; thermal problems of power semiconductor components;
3. accurate determination of entity and location of the losses, for the development of reliable thermal models;
4. study of the fluid-dynamical aspects and their implementation in the thermal models, with particular reference to design, operation analysis and/or control of the processes in which the components operate;
5. development and improvement of thermal networks of the components, based on the knowledge of the constructive and/or design data, suited to consider the variation of the operating conditions (for instance: cooling variation in adjustable speed self-cooled machines; loading conditions with severe operation cycles);
6. development and improvement of thermal simplified networks of the components, with test identified parameters;
7. improvement of the effectiveness of the cooling systems;
8. coupling among thermal and electromagnetic aspects in defining the design algorithm of the components, with particular reference to special purpose calculation tools;
9. in collaboration with the manufacturers, realization of adequately instrumented prototypes for thermal measurements and characterization of local phenomena and/or validation of the corresponding complex models (thermal exchanges in correspondence to the endwindings, distribution of temperature in the slots, thermal exchange in the air-gap, accurate determination of radiation phenomena,…);
10. using as reference the values extracted by the experimental data, comparison among the results achievable from a thermal circuit analysis and those of selected numerical simulations, employing FEM commercial codes for the analysis of the thermal field.
It is important to underline that the research is aimed to simultaneously deepen the thermal problems of the electrical machines and of the static converters for industrial, civil, aerospace and propulsion applications. This research will allow to analyse components and systems ranging from some hundred watts (fractional power) up to some hundred kW, largely covering many sectors of low voltage motorizations and drives. The obtainable results will be therefore of particular interest, because of general validity for a wide range of applications and systems.
The research Units fully hold the competences required for the achievement of the previously described technical goals; they have developed various collaborations at national and international level and they are well integrated in the industrial environment, in such a way to obtain really useful results. It is worth to underline that various firms previously contacted by some research Units have answered positively to the proposal of an active and profitable cooperation within this research, also from the financial point of view.
To confirm the interest on the proposed research it is important to underline that foreign researchers will be involved in the project. <<<

Timescale

24 months

National and international background

The use of electromechanical components and of power electronics systems is progressively spreading and specializing in various areas, both of the civil society and in the military sector ("more electric drive" solutions). Consequently, there are several examples related to the fields of the industrial processes, of the automation systems, of the applications in civil sectors (both "consumer" and commercial), of the transportation systems (not only terrestrial - electrical vehicles -, but also naval - "all electric ship" solutions – space and aeronautical - "more electric aircraft" -).
The increasing demand of performances optimization are more and more linked to the pursuit of other important objectives, among which: "custom" component realization (with dimensional and service constraints specifically oriented to the application); containment of the manufacturing costs (through the deeper exploitation of the active materials and the adoption of configurations more suitable for the automated production); operation reliability and robustness (also under severe environmental conditions and in greatly variable operating cycles).
The development of effectively competitive products asks for an integrated approach, that combines the followings aspects:
-in case of development of devices with innovative configurations, the adoption of tools and methodologies well suited for the component design and verification is essential, characterized by adaptability to different specifications and typologies, accuracy in the performances forecast and rapidity of elaboration (especially when coupled with optimization procedures);
-even in case of adoption of already available commercial components, their correct selection and effective management requires a suitable modeling of their behaviour, in order to reliably appraise its operating limits;
-in all the cases, it is necessary to operate a careful study of the operating conditions of the considered process and an optimization of its control, examining the mutual interactions among the components, with a system viewpoint.
For a long time, in analyzing these problems, wide part of the efforts has been devoted to the following aspects:
-refinement of the methodologies for the electrical machines electromagnetic design (analytical methods, numerical approach with FEM codes), frequently uncorrelated with size, thermal and cost limits;
-studies for the definition of static conversion solutions with suitable functional features (in terms of waveforms, regulation capabilities, layout compactness);
-deepening for the development of high performance control schemes (position and speed control precision, response promptness, robustness and immunity to the disturbances).
Lately, the researchers and the manufacture industries producing machines and drives became progressively interested in the cooling problems of devices and apparatus, in consideration of the increasing importance connected to a suitable knowledge of the operating thermal conditions, not only in the sizing stage, but also during the process control.
It is well known that the subject of the component thermal verification has already represented for a long time a classical chapter of many design treatises: thus, it is spontaneous to question ourselves about the elements of interest and novelty included the present research project. Among the principal reasons, the following ones can be quoted:
-the classical thermal analyses are largely related to the basic component typologies, mainly with reference to the steady-state thermal operation;
-currently, the operating conditions are more and more frequently far from the traditional ones of the continuous service, both as load diagrams, and as environmental conditions (however, a lot of Standards tend again to refer to the continuous service as an equivalent condition: see IEC Std. 60034-1: Rotating electrical machines. Part 1: Rating and performance);
-the electromagnetic design must be more and more tightly coupled with the thermal analysis, also in consideration of the progressive use of new types of electric motors and solutions with non conventional configurations, with application oriented custom design, for which the choice of the utilizations can hardly exploit the data of previous experiences;
-a typical critical element for a reliable formulation of a thermal analysis is constituted by the uncertain knowledge of the thermal exchange parameters: the vast specialized literature suggests numerous "per unit" expressions of such quantities, that nevertheless not always can be correctly referred to the actual geometric, radiating and fluid-dynamical conditions;
- often, the thermal analysis, traditionally performed on the electrical machines, neglects relevant aspects of fluid-dynamical nature: in fact, the cooling fluids are affected by turbulent flows that require to be taken into account by suited models;
-the availability of manufacturers inclined to construct thermally monitored prototypes, suited to the execution of tests for thermal validation and characterization, is a conclusive supporting element for developing reliable and efficient tools of thermal analysis and it is a peculiar feature of the project.

The actual international situation regarding the different aspects of the thermal analysis is characterized by the following approach methods:

[A] Analysis of the thermal problems of the electrical machines by means of FEM programs [A.1-5].
The common characteristic of these analyses is the adoption of simulation programs potentially able to guarantee elevated precision in the results, frequently through a coupling among electromagnetic and thermal modules; nevertheless, not always such programs are suited to model the structural complexity of the components and to adequately implement the phenomena of convective and radiating exchange; finally, these programs present a remarkable computational heaviness, and they are difficult to be integrated in optimization procedures. On the other hand, this approach remains of meaningful usefulness for comparative evaluations with other calculation and/or experimental methods.

[B] Study of the thermal-fluid-dynamical aspects of the electrical machines, by using codes based on CFD techniques (Computational Fluid Dynamics) [B.1-5].
These methods allow very sophisticated analyses, by using packages devoted to the study of the cooling fluid flow profiles and to the related thermal exchange conditions with the internal and external surfaces of the cooled components; on the other hand, even if capable of powerful and detailed analyses, these codes present a tremendous simulation heaviness; thus, they are generally used just for selected analyses, because not well suited for coupled magneto-thermal studies or for extended analyses of parametric sensitivity.

[C] Analysis of the distribution of temperature in the electric machines, with lumped parameter thermal networks [C.1-5].
This approach, for a long time traditionally at the base of the thermal verification of the electric machines, has progressively improved methodologies and modeling accuracy, to the point that acceptably reliable commercial packages are currently available for the operation analysis of some types of standard machines, at constant rotational speed; the advantages in the use of the thermal networks are remarkable: they allow an easier computer implementation; they are suited to be well integrated in an optimization iterative process; they reduce the needed calculation times and resources; they allow a direct approach to the study of the model parametric sensitivity. The improvements achieved during the last years are the result of patient, progressive activities of validation and identification conducted through simulations by FEM and/or CFD codes and, above all, through experimental tests; nevertheless, a lot of job remains still to do, for instance in the following directions:
-to get accurate thermal analysis of self-cooled electrical machines operating at variable speed;
-to obtain a reliable thermal modeling of electrical machines whose employment requires a totally enclosed structure and extremely high overloading capabilities;
-to perform studies of thermal conditions in severe environments or in absence of atmosphere;
-to effectively join the thermal analysis with the electromagnetic design;
-to complete the tools of thermal analysis by enclosing the whole drive in the model.

[D] Thermal analysis of static components and of power electronics devices [D.1-6].
It is a sector as important as that of the electromechanical components, also for the particular vulnerability to the overloads, typical of the static components and of the semiconductor devices, because of their modest thermal time constants; thus, it is essential to adopt suitable integrated tools for the thermal verification, capable to suggest a correct system sizing and to lead to an effective thermal verification of the elements that are more stressed by the service conditions.

[E] Simplified thermal models, for monitoring and diagnostic purposes, with parameters identified by test [E.1-5].
This approach to the thermal problems is substantially different from the designer viewpoint. In fact, the goal is not a thermal network synthesis based on the constructive data and on the knowledge of the cooling conditions of the component, aimed to optimize its sizing. On the contrary, the components are already available on the market, but their constructive data are unknown: the required action is represented by the development of suitable thermal models, with the purpose to correctly forecast the component behaviour in the process and/or for thermal state monitoring purposes (for example, by using sensorless techniques or "thermal image" models). Also in this sector, the activity to be developed results considerable, both for the definition of circuit models acceptably representative of the thermal evolution of the most critical portions of the components, and to select meaningful test procedures and measurement arrangements suited for the identification of the parameters of such networks.

By examining the industrial activity segment, with particular reference to the Italian national electromechanical industry of the sector of converters and drives, a situation of remarkable fragmentation emerges, mainly characterized by a high number of small and medium firms, generally with fair levels of adaptability and dynamism towards the market, but often with insufficient economic and human resources to directly carry out activities of development and innovation on their own. If not adequately supported in the competition with the foreign industries, often of greater dimensions, this kind of firms can result the most vulnerable from the point of view of an unstoppable regression.
From here the usefulness emerges of a contribution from the academic world to the applied research, aimed to the component improvement, in which the close examination of the thermal problems presents a strategic role. The products of the proposed research can have positive outcomes for the Italian industries and indeed this condition should be one of the key elements as regards the criteria for the global project estimation and the evaluation of the single steps of the proposed activity. On this subject, it is opportune to underline that this research has stimulated remarkable interest in known manufacturing Italian firms of electrical motors and drives, that, for such reason, have subscribed letters of intent concerning the financing of the research in object.
The research Units involved in the project have a consolidated experience of studies and tests in the field of the electrical machines and drives and on these subjects they have notoriously developed and consolidated remarkable collaborations with firms of the sector.
All the Units have the required competences in terms of design and realization of electromechanical, power electronics and control components and systems. As specifically detailed in the respective forms B, in the past, all the Units have participated in National and/or European research programs. <<<