Noise analysis by means of a suitable excitation function

Choosing appropriate test functions for noise analysis of electric motors

In many cases, additional noise analyses are required for determining the noise behavior of an electric motor. If the test is to be performed without using an external load, then the choice of the correct test function is crucial.

The test function should be selected so that all the forces that produce noise can be analyzed by respective sensors. The most common sources of noise are: rolling bearings, commutators and electric forces.

Based on the testing of DC motors in the automotive industry, it is shown that when a dynamic load is introduced under test conditions, assembly errors that lead to unacceptable noise for the customer can be identified and qualified. To a great extent, the detection of these errors depends on the selected test function. Quasi-periodic and pseudo-random test functions are examined and given guidelines for their use. In addition, it can be shown that a sound analysis by means of a suitable excitation function can be performed simultaneously with parameter estimation, thus, reducing the test time.

Classification in acoustical-loss diagnostics

The acoustical performance characteristics of a machine or machine elements at the end of the production process are to be subjected to testing, as they are prone to variations due to stochastic fluctuations in manufacturing tolerances.

In regards to implementing acoustical-loss diagnostics, there are two basic problems:

  1. Environmental conditions
    The acoustic characteristics of an object of utility are generally defined using airborne sound limits. The measurement of airborne noise during production is possible only under immense effort (using airborne noise insulation), which is often out of proportion to the cost of the manufactured product.
  2. Functional relationship between airborne noise and structural noise
    The transmission of a product’s airborne noise limit values into corresponding structural noise limit values is functionally not possible. Thus, the structural noise limit values must be determined experimentally from the airborne noise limit values. The problem can be solved by means of a learning process. The learning algorithms use either a predetermined classification of the test object or a predetermined characteristic feature for a structure analysis of the test data and can be used for making a decision after the successful learning phase.

Regenerative testing

Electrical machines, which are energized and driven from the outside, induce a voltage which can be measured on the connection lines of the machine. The induced voltage is proportional to the speed and excitation. The course of the induced voltage gives information about the windings and the characteristics of the excitement around the circumference.

The measurement of the induced voltage provides a simple method to diagnosis the electromagnetic behavior of the motor. Regularities are derived from a moving conductor loop in a constant magnetic field.

Moment of inertia simulation for electric motors

For systems with multiple components, it is not only important for motor design to be interpreted with respect to its temperature behavior, rated speed and rated torque, but also to closely examine its system dynamics. When the exact overall system behavior is of interest—in the case of load jumps or starting and stopping of the engine—this becomes useful: e.g., in the area of control design and the selection of system components.

In the simplest case of the electric motor and the load, the inertia of the entire system of a drive train is composed of the sum of the inertia of the individual components of the system.

Moment of inertia
The angular momentum is calculated from the mass and the arrangement thereof in the body, and increases in proportion to the angular velocity. The angular momentum of a body will change, if it is replaced with another body. The angular momentum capacity, or moment of inertia, is a value for the amount of stored angular momentum in a body.

Cogging torque measurement

The electrically and magnetically excited harmonics occurring in electric rotating machines lead to torque ripples. The magnetic harmonic oscillations result from the constructed, non-homogeneous structure (grooving) of the engine.

Magnetic resistance is described in terms of reluctance. Reluctance torque arises in machines in the circumferential direction at different magnetic resistances. These are divided into different magnetic resistances in the rotor or in the stator. In the stator, they return back to the slot openings and hot cogging torque.

For permanent-magnet machines, the number of poles in the rotor multiplied by the number of strands in the stator equals the number of preferred stable positions in which the rotor moves. The amount of cogging torque is significantly influenced by the structural design. In cases with low cogging torque values, the current to break the rotor loose from a standstill are lower. The cogging torques overlap during rotation of the rotor with the produced air gap torque and do not contribute to torque generation.

Electric motor testing: characteristic curve recording

In the area of new drive system development and electric motors, as well as in the area of quality assurance, quick and efficient testing and archiving of data is essential for a successful product. For basic characterization of the test objects, characteristic curves or individual points of the curve (working or nominal points) are usually used.

As a rule, the characteristic curve is recorded quasi-statically, i.e., the test object is continuously slowed down or unloaded. Another approach is the static starting of individual load points. At these points, the transient oscillations are awaited and the values measured statically. The second method is more time-consuming and leads to a greater warming of the specimen, which inevitably leads to a false recording of the measurement values themselves.

S-operational modes according to DIN VDE 0530

To simplify the design of motors, there are different rated operation modes. Here, e.g., a continuous operation can be described as operation mode S1. The motor suppliers provide the respective designed characteristic curves for the different operational modes. Based on these characteristic curves, a motor selection is therefore possible.

Running temperature in regards to the thermal design of electric motors according to VDE 0530
During motor design, it is important to pay attention that the maximum permissible motor temperature is not exceeded. In practice, however, the application-specific temperature profile is not known, and makes it difficult to make the correct motor selection. The economical design of drive-lines under current energy-savings debates is of paramount importance. There is a large number of drive-line designs installed with the wrong motor dimensioning, when it would be possible to install one with more energetically favorable qualities.

It is understood, that the load on the engine under operation includes its chronological duration and order and, where appropriate, its starting, electric braking, idling and pausing.