Projects:2020s2-7542 Thermal Modelling of Electric Machines under Overload Conditions
There are some aspects and limitations that prevent electric machines from getting more power. First, the stator winding insulation can get too hot that could shorten life or even damage the insulation coating on the surface of the copper wire. Another limitation is that in machines that use permanent magnets, such as most of the synchronous machines, as the magnet gets hotter, it approaches a point where it starts to lose its magnetism. Generally, the more power applied to an electric machine, the hotter it gets which means more losses.
Contents
Introduction
Induction motors are used in different commercial and industrial applications. A motor is used to transfer an electric energy to a mechanical energy, where in an ideal motor operation, all of the electrical energy would be transformed to a mechanical energy in the form of rotation. As the current passes through different components, losses occur. First, the stator has copper losses in the form of heat which is proportional to the I2 flow through the stator winding, whereas the iron loss is proportional to the V2 across the stator core. On the other hand, the rotor has copper losses only due to the insignificant iron loss that occur in the rotor. In normal operation, the rotor would have a low frequency due to the low value of the slip, the relationship between the slip and the rotor frequency is described in the following equation:
𝑓𝑅=𝑠𝑓
During each stage, heat is conducted until it is dissipated to the ambient where fins are utilized to maximize the amount of heat out of the systems. This paper will investigate temperature limitations, and hence losses. For example, winding losses are produced in the form of heat in the copper wire of the winding of an induction machine which is caused by the resistance of the copper wire. The project goals are:
1. To investigate the power relationship to the temperature of induction machines;
2. investigate the effectiveness of some cooling methods on induction machines;
3. use a different kind of thermal modelling of machines; and
4. compare measured quantities to the simulated and calculated quantities.
Project team
Project students
- Majed Alsahli
- Mohammed Almusallam
- Shaoqing Liu
Supervisors
- Dr. Wen L. Soong
- Dr. Dmitry Chaschin
Motivation
Motors and generators are an important part of today’s modern civilization, from something as big as power generators to a small fan in a room. Investigations on such revolutionary devices are continuing, as in recent years, electric cars that use traction motors have been in production, however, thermal management is one of the challenges that electric traction motors face. Especially, high-performance electric cars demand high power from their motors and generally with high power comes a high amount of losses in the form of heat. Therefore, electric cars are only allowed to draw a high amount of power for a short period of time with limited repetitions.
Background
There are different ways of doing thermal analysis, there is lumped thermal circuit analysis, which is an analytical equivalent circuit-based approach. The basic electric circuit analysis still applies, such as Kirchhoff’s circuit laws. However, in the thermal circuit, the temperature is treated as the potential difference between two nodes that imitate the electric potential in an electric circuit. The heat flow is the flow of heat from high-temperature surfaces (high potential) to the colder surface (lower potential) which imitates the flow of electron current in electric circuits and is measured in Watts (W) [10]. Furthermore, thermal resistance imitates the electric resistance in an electric circuit. Therefore, ohm’s law of electricity (𝑉=𝐼𝑅) translates to Δ𝑇=𝑞𝑅Θ in thermal circuits.
Another way to analyze a thermal problem is by using the finite element model that models any thermal problem as a 2D device to show the heat distribution across a hot object, by setting up the thermal conductivity and heat capacity of different materials in different parts of the object. The induction machine is modelled in 2D finite element model that is simplified, as a comprehensive geometry of the motor is usually complex to include the 3D heat flow in the motor (axial, radial, and tangential). However, it is not impossible to create a thorough model of the motor, although, large computing resources are needed [11]. Furthermore, heat sources are defined by calculating volumetric heat generation which, by definition, is equal to the heat flow (power flow) divided by the volume of the material that generates heat. The model is beneficial to calculate the temperature difference between two nodes. As a result, it allows us to calculate the thermal resistance between the temperatures at the two nodes and to calculate the average temperature rise.
Through the finite element model, we can see later how the heat is conducted from the winding to the stator core. Conduction is the transfer of energy from the more energetic particles to the less energetic of a substance, which means the heat is transferred from the hot substance to the colder one where, the hot is the more energetic. Therefore, heat travels towards the direction of decreasing temperature. The finite element model allows us to simulate the heat generated in the winding based on the power input to a machine or a transformer. Therefore, experiments can be compared to the results obtained from the simulation and calculated values can be verified. The induction machine’s stator and rotor core consist of lamination stacked on top of each other. One piece of the stator and rotor lamination is shown in Figure 4, where each one is coated with an insulator to prevent eddy current from being conducted between laminations in the axial direction. The stator winding and the power rating of the machine must be set so that the current density should not be high which woul result in high hot-spot temperature [12]. The goal of testing the thermal behaviour of induction machines is to determine the thermal limits that constraint the machine and decreases the life of the machine as these limits are crossed.
Temperature data helps to predict the insulation resistance behaviour as well which is expected to decrease as the temperature increases. The insulation resistance limit is set by the manufacturer, it is expected from the user to not cross the limit at which the high temperature will damage the insulation, in result, the motor could fail. An induction machine radial cross-section is displayed in Figure 4 to show a typical stator and rotor lamination slots shape which can vary between machines.
Different lamination shapes would control how efficient a machine is, in result, of different thermal management changes. Motor design goes through an optimization process to achieve a homogenous thermal distribution through the motor radial cross-section.
Finite element analysis is done on the motor radial cross-section to simulate the motor thermal behaviour. An initial model of the machine stator and motor is shown in Figure 5. Due to the symmetry of the motor, only a quarter of the motor stator and rotor lamination is needed for thermal model which is then used to determine the peak and average temperature across the stator winding slot.
Method
Since the project investigates the temperature of electric machines, a 1:16 turn ratio transformer was used as a proxy to electric machines stator winding and lamination which can replicate the thermal behaviour of machines. Figure 2 shows the experiment set up where two different thermometers were used to measure the temperature of the core and the outside surface of the winding as well as the temperature of the ambient. A DC supply is used which supplies a voltage-dependent voltage source, which means that the voltage is adjustable while the current is varying depending on the value of the voltage. The copper winding then is set to heat up due to its resistance. The copper conductor resistance increases with temperature which will cause the current to decrease. However, the temperature and current eventually reach a steady-state condition. In general, the higher the power dissipation in the winding, the higher the steady-state value is. As the thermometer of the winding takes measurements at the surface of the winding, the temperature is much higher towards the center of the copper wires. This is due to the fact that the heat in the winding is described by a parabola as shown in Figure 3. According to expectation, the core won't have any losses hence there won’t be any heat generated from the core, and all the heat that is measured off the core is conducted from the heat source, which is the winding. The reason that the core will not have any iron losses is that the supply is a direct current (DC) which will not have an alternating current (AC), which in result will produce an alternating flux in the iron core. The losses in the winding, hence the heat, are proportional to the voltage squared where the small change in the supply voltage will drastically affect the temperature rise of the winding. Therefore, different voltages were tested, to compare the effect of voltage on the temperature.
The temperature profile of the steady-state test is expected to have the same as the steady-state test done on the transformer (Figure 6). Where the temperature will be rising until it reaches a steady-state condition where the generated heat is equal to the heat dissipated to the air or conducted to other components in the machine.
Results
The finite element model shown in Figure 1 is built based on the materials used in the induction machine which are iron for the stator and rotor cores and copper for the stator winding. Therefore, the materials' thermal properties are set, such as the heat capacity and thermal conductivity. Other physical properties are also set, such as the thickness of each element.