Abstract

Testing and simulation of fractionary electromechanical rotative drives

D. P. Burgoa^I; C. A. Martin^II

^I david@emc.ufsc.br

^IIUniversidade Federal de Santa Catarina Departamento de Mecânica, Laboraratório de Hardware – GRUCON, C. P. 476, 88040-900 Florianópolis, SC. Brazil, cam@grucon.ufsc.br

ABSTRACT

This paper concerns the development of drives that use electromechanical rotative motor systems. It is proposed an experimental drive test structure integrated to simulation softwares. The objective of this work is to show that an affordable model validation procedure can be obtained by combining a precision data acquisition with well tuned state-of-the-art simulation packages. This is required for fitting, in the best way, a drive to its load or, inversely, to adapt loads to given drive characteristics.

Keywords: Test bench, simulation, fractional electrical drives, automatic test and measurement, system dynamics analysis

Introduction

Fractionary Electromechanical Rotative Drives and its Applications

Fractionary electromechanical rotative drives, FER Drives, are already applied in truly uncommon quantities and even though, the tendency of new applications remains strongly growing. It is also observable that nowadays FER Drives are available in a great variety of conceptions and designs.

Main examples of applications are the vehicles industry (automotive, aerospace, etc.), appliances (especially audio, video, home), computers (especially in peripherals as printers, disc and tape drives, image plotters and scanners, fans, etc.), in industrial automation (measurement and registration instrumentation), machines and manipulators from small to medium size, etc. (Bahniok, 1989) (Pimentel, D., 1996)

Similarly it is observed that the users are increasingly demanding in terms of the best dynamic performance of drive systems (high acceleration; high rotation speed; smooth motion even at very low rotation), low friction, light weight, efficiency, etc. and also high reliability and durability, and, of course, it all must become available at low costs, each day more pressed by an extremely competitive global supplying market. (ELECTRO-CRAFT, 1989).

The market forces an accelerated development of new and specific drive systems. Their knowledge in depth is also required in application projects in order to achieve integrated solutions, matching the drives and loads.

Analysis of Fractionary Electromechanical Rotative Drive Systems

The behavior of FER Drives can be described by parameters and characteristic curves determined by manufacturers. Projects of good quality applying these drives are then achieved, combining such data with product register files and computer aided evaluation and also with model simulation softwares.

It is possible to proceed a drive system development directly based on a specific application. Besides being restricted to a particular application, such approach presupposes the availability of all the drive data.

However, the information is not always conveniently set in the products data sheets, thus creating insurmountable difficulties, especially in the case of matching components of distinct sources and that may be supported by different local standards. Besides this, many tests carried out show that there is also the problem that some products are simply not able to reach (and hold) the specifications presented in their respective data sheets.

Therefore, this work proposes the development of specific solutions in two steps. a) At first, a test bench is used for experimentally simulate all the conditions under which the drive system is going to be used. This can provide a good matching between drive characteristics and the load, so that further operation during the application will be perfect. b) A drive model simulation system aiming the validation of the model used in applications such as drive controller development.

Test Systems for Fractionary Electro-Mechanical Rotative Drives

With the purpose of realizing tests to obtain reliable data for the development of motion systems with FER Drives in the power level bands corresponding to torques of 5 Ncm to 2 Nm and also for didactic presentations and quality control in the production of drive components and systems, it is indispensable to have a test system, that, by the means of loading and measuring/instrumentation devices, allows the determination of the dynamic behavior of the drives under test, in the form of its frequency response and its time response at acceleration/deceleration and at coupling / uncoupling of mechanical loads and also by obtaining static parameters and characteristic curves, all depending on the kind of loading applied, e.g. inertial loads without friction, viscous friction loads with negligible inertia and/or any of these combinations. This test system is broadly presented in (Pimentel, D., 1996). It follows only highlights of it.

State-of-the-art drive systems for fine mechanics and precision mechanics, e.g. drives using transmission by synchronizing belt or direct drives, powered by DC or AC servomotors or also by stepping motors, can be tested reliably by reproducing the real conditions of the applications operation. Modes of operation in open loop or closed loop control systems, e.g. to control position, velocity and/or torque, can be outlined, while the data acquisition operates in an analogic way or totally digital.

In the conception of the mechanical structure of the corresponding test bench performed for the system (it itself classifies as equipment of fine mechanics) it was tried to guarantee the maximum operational flexibility and simplicity at mounting each of the many variants of test configurations that result from the combination of bench elements such as the drive under test, the mechanical load elements and the transducers for the various mechanical, electrical and thermal variables and parameters. For this purpose those elements are mounted on devices especially designed for frequent test configuration changes, while otherwise the coupling of the respective axles is fast and safe. A strong mechanical base structure construction guarantees the effective damping of the vibrations induced by drives of the expected power range. Moreover, it was guaranteed especially that, for example, parasitic moments of inertia, friction and elasticity, specially non-linear effects, was made negligible, in order to achieve an uncomplicated modeling of the system and to get a very good conformity in the further comparison between the measurement results and the results of the mathematical simulation of the model of the drive system under test.

Instrumentation of the Test Bench for the Various Measurements

On a test system for FER Drives like that shown in Fig. 1, the necessary instrumentation must be available to obtain: The static rotation x torque curves, which requires the utilization of a torque sensor and a tachometer; the time response of position, velocity, acceleration parameters that must be acquired fast and precisely; this is done by the use of a computer interface for the communication with an encoder, that delivers the position directly by up-down counting the number of its output pulses; through the correspondingly pulse rate, the angular velocity can be derived; from the variation of pulse width the acceleration can be get, and from its two quadrature pulse channels also the rotation sense can be detected.

The measurement of the characteristic constants of the drives such as torque constant, voltage constant, electrical and mechanical time constants, requires additionally an ammeter and a voltmeter; the power for the drive system under test is supplied by a regulated power supply; to obtain the frequency response of the drive and its components a Control Systems Analyzer - CSA (HP Application Note, 1990) is interfaced with the host computer.

Management of the Test Bench

In order to make the tests totally automatized, it was elaborated a managing software in a WindowsÓ environment, formerly with the programming language "Borland C"Ó and now, applying "Delphi 3"Ó , that offers a better and easy graphics programming interface.

This program consists basically of the following modules:

"Main", which presents the basic options of the manager, the most important begin the test option, which allows to set the commands individually and visualize the acquisitions, Fig. 2.

"Parameters", which allows to specify the operating parameters of the drive system under test and the various signal conditioners of the bench (screen "hardware", Fig. 3) and another screen "torquemeter" for setting its electronics.

"Tests", which allows to set the parameters for the various programmable tests with the test system (e.g. for the torque vs. speed characteristic, Fig. 4); after performing the test, the results are presented graphically.

Furthermore, the software also offers the option of choosing the user interface language, that at this time are, English, German, Portuguese and Spanish. It also offers the option of keeping in a motion products data base, the data of different motors already tested, aiming later comparisons with the results of new test. Another option allows the setting of data tolerance ranges in order to enable the classification of the tested drives..

Numerical Simulation of Fractionary Electromechanical Rotative Drives

In the teaching and development work on motion drives, apart of the testing itself, it is very important, for various reasons, to perform mathematical simulations of the behavior of the drive system. The simulation allows the validation of the acquired data and the model by comparing the measurement results with those of the simulation. The simulation also can show how the drive behaves when parameters of the system change, here including the boundary cases that lye outside the physically safe operating range of the equipment. In such cases, of course, the mathematical simulation model should have previously been fully validated by comparing its results with the test results. The model must be obtained or developed before. This may be a time domain or frequency domain model, since data obtained with the test system supports both options. In order to start the simulation, a very simple frequency domain model – transfer function – of the system is presented. Its parameters are exchanged with the measured data. Then these equations are implemented to a simulation software package, like Vissim©, Simnon©, etc. For the present work the powerful Matlab/Simulink© package was applied. The results are presented as graphics outputs of the frequency behavior and time response. After the first validation is completed by having a good matching between measured data and simulation output, the simulation model is ready for the execution of runs with model parameters freely changed by the user, with the aim of observing the response of the drive when applied in a system subject to alterations. Obviously, a lot of parameters can be altered also physically in the test bench for matching the corresponding alterations introduced in the model, so that it is now possible to once more compare the measured results with those of the simulation when both, the physical system and its model, are simultaneously altered.

Results

The presented results cover only few of the possible tests types performed at the bench with fractionary high dynamics motors. All the tests herein presented refer to the dynamic behavior of the drives. More about tests, mainly static ones is present in (Pimentel, D., 1996).

Test for obtaining the mechanical time constant

The mechanical time constant, t _m, is a dynamic mechanical characteristic of the drive system. Applied a step function input voltage (Fig. 5), the mechanical time constant is defined as the time delay for reaching 63.3 % of the final speed value (ELECTRO-CRAFT, 1989). The value and time delay of the step function input applied on the motor is generated by the HP-CSA.

The mechanical time constant is a parameter through which one can classify whether or not a drive system is of high dynamic (The current value of t_m for truly high dynamic fractionary (smal) drives is today under 1 ms, (Bahniok, 1989))

Test for getting the Frequency Response

To obtain this plot, it is applied to the drive input an harmonic signal of variable frequency, and the amplitude ("dB") and the "Phase" responses are obtained (HP Application Note, 1990), Fig. 6. Also this plot allows the evaluation of the dynamic response of the drive under test. In this plot the drive resonance frequencies can be observed, what is useful when designing the drive system.

Figure 7 presents the test system structure for torque-frequency response measurement. Figure 8-left shows the output data of the obtained torque-frequency response (amplitude (dB) and PHASE) in the form of Bode plots together with the fitted amplitude response curve (CURVE FIT), and the corresponding numerical data in the pole, zero and gain table (Fig. 8-right, S Curve Fit).

Eq.1 shows the transfer function of the drive system under test, achieved by introducing the measured numerical data in the polynomial form.

Figure 9 finally shows the resulting Matlab© amplitude and phase response output of the Transfer Function evaluated for the data obtained in the previous measurement.

Comments and Conclusion

A comparison between the Bode plots generated by the CSA and those of the Matlab© simulation shows the excellent matching already achieved.

The current development will allow a full linking between the bench instrumentation, its management system and the simulation softwares.

The results obtained for the tested motor, are presented similarly to the manufacturer's catalogs (HONEYWELL, 1985). However, it was noticed that some of the values determined for the given motor sample at the test system and shown in Table 1 do not fully match that of the factory motor sheet.

Presented at COBEM 99 – 15th Brazilian Congress of Mechanical Engineering, 22-26 November 1999, São Paulo, SP. Brazil.

Technical Editor: José Roberto F. Arruda.

Publication Dates