Understand Heat Load


Permanent Magnet Synchronous Motor (PMSM) has the advantages of high efficiency, compact structure and high-power density, and is widely used in various industrial fields. With the continuous improvement of the power density of PMSM, the cooling problem is becoming more and more serious because the common cooling scheme is becoming more and more difficult to meet the requirements. Seeking a better cooling scheme is one of the main problems restricting the development of PMSM to higher power density. The main cooling methods of permanent magnet synchronous motor are air cooling, oil cooling and water jacket cooling. Air-cooled motor promotes heat dissipation by increasing the internal and surface gas flow rate, with the advantages of simple structure, easy maintenance, stable and reliable operation. But the PMSM often needs to operation closely, so that the cooling gas is difficult to enter the motor interior.

 The oil-cooled motor sprays the cooling oil to the blades fixed on the rotor through, and the cooling oil will be stirred to the motor cavity by the blade. Using this method, the cooling performance of PMSM is remarkable and the heat dissipation of the end winding can be taken into account, but the cooling oil will also produce motion damping to the rotor, and to prevent the leakage of oil, the sealing requirements of the motor is high. Because the thermal capacity of the stator water jacket is strong and the waterway arrangement is flexible, at the same time, the influence of cooling system on the electromagnetic performance of the motor is very small, water jacket cooling has become the most common cooling method for PMSM, and a lot of literature has been studied. Owing to the water jacket is usually placed out of stator and integrated with housing, relatively far from the magnet and winding which are usually the main heat source, water jacket cooling method is not good enough for the internal cooling of PMSM, which limits the development of PMSM to higher power density.

In contrast, if the idea of internal water cooling in the large generator is applied to high power density PMSM, cooling pipes in stator slots can directly cool the inner heat source, being closer to winding and magnet compared with water jacket cooling method. A lot of literature and engineering examples relating the application of IWC scheme on large generator have been produced at home and abroad. However, the application of the internal water-cooling method on PMSM is rarely reported. Given that PMSM is usually relatively small in size and space, the material and tubes layout of IWC scheme and the influence on the performance of the motor is different from those of large generators, which need special research. The paper employs the formed coil winding and resin coating technology to form the cooling channel in stator slots, making full use of the slot space and depressing temperature rise of the winding to a large extent. However, the cooling channel stripping process is complex and expensive, which is not applicable to large-scale industrial applications.

In order to solve the problem mentioned above, it is more practical to use a copper tube with low cost and mature processing technology to form the IWC system in PMSM. This paper takes an outer rotor PMSM as an example, presenting a kind of IWC scheme which is convenient for engineering realization and has little influence on the electromagnetic performance of PMSM. The arrangement and connection of tubes have been designed aiming at reducing circulation current loss in the parallel waterways; From the angle of restraining eddy current loss in copper tubes, the position and thickness of tubes have been optimized. The designation and optimization presented in this paper are both verified by FEM simulation.

The losses in PM alternators are grouped into (a) stator loss, (b) rotor eddy current loss, and (c) windage loss.

  1. Stator losses –

The stator loss consists of copper loss and iron loss. Copper loss is the loss due to the current going through the armature windings. The copper loss consists of I2R loss and stray load loss.

where m1 is the number of phases, I is the armature current, and R is the dc armature resistance. The I2R loss can be significant when a large current flows through the conductor with large Ohmic resistance.

The stray loss comes from (a) the skin effect resulting from the same source conductors and (b) the proximity effect resulting from the field-induced from adjacent conductors sharing the same slot. The skin effect is caused by electromagnetic induction in the conducting material, which opposes the currents set up by the wave E-field.

The copper loss is temperature-dependent, so the copper loss is calculated at the expected copper temperature. The copper I2R loss increases when copper temperature increases due to increased winding resistance, while the copper stray load loss reduces with increased temperature. In addition to the analytical method described above, loss due to proximity and skin effects can also be simulated based on transient time-stepping FEA.  However, this method is very time consuming especially when multiple strands are used.

Iron loss produced in a magnetic material operating in an alternating magnetizing field is generally separated into two components: hysteresis loss and eddy current loss. Hysteresis loss is due to a form of intermolecular friction when a varying magnetic field is applied to the magnetic material. The loss per cycle is proportional to the area enclosed by the hysteresis loop on the B-H characteristics of the material. The hysteresis loss increases with the maximum magnetic field

The term “eddy current” refers to circulating electric currents that are induced in a sheet of a conducting material when it is subjected to an alternating magnetic field. These eddy currents pro- duce power that is dissipated as heat. The eddy current loss per unit volume, at frequencies which are low enough for the inductive effects to be neglected.

Soft magnetic materials form the magnetic circuit in an electric machine. An ideal material would have high permeability in order to reduce the reluctance of the magnetic circuit, high saturation flux density in order to minimize the volume and weight of the iron core, and low losses. However, it is impossible to optimize all of these properties in a single material. This is because there are a large number of factors that affect magnetic properties (chemical composition, mechanical treatment, and thermal treatment are the most important), and the result is often a compromise. For example, nickel steel has low iron loss but low saturation flux density, while cobalt steel has higher saturation flux density but also higher iron loss. The iron core of a machine is made up of thin laminations in order to reduce core loss. Laminations as thin as

0.127 mm are generally used in high-frequency applications in order to reduce iron loss.

  1. Rotor Losses-

The rotor loss generated by induced eddy current in the steel shaft and permanent magnets is not significant compared with the total machine loss. However, removing the heat from the rotor to ensure reasonable operating temperatures of its components is more difficult than removing the heat from the stator. Thus, an accurate prediction of rotor loss becomes important especially at high speed.

There are several methods to reduce rotor eddy current losses. Reducing the slot opening and increasing the magnetic gap between rotor and stator can reduce no-load rotor loss. Increasing the number of slots per pole and using fractional winding can reduce rotor loss caused by the space harmonics of the armature winding. Increasing the switching frequency and using external line inductors can reduce rotor loss caused by time harmonics of the phase currents. Since rotor loss caused by time harmonics is dominant in most applications, increasing the switching frequency and using external line inductance to reduce current THD is a very effective way to reduce rotor loss.

  1. Windage Losses-

Windage loss is heat generated in the fluid due to the relative motion (shearing) of the fluid that flows between the rotor and stator [13]. Windage loss, depending on various gases at various operating conditions, as used in high-speed machines can be very high, contributing to overall machine inefficiency. The windage loss generation is a function of shaft rotational speed and fluid properties such as temperature, pressure, density, and temperature gradients at stator and rotor walls.

The windage loss generated in the clearance between a rotating cylinder and a stationary cylinder with homogenous laminar flow (no axial flow) can be estimated.

Theoretical relations and experimental validation taking into account the combination of axial flow and rotational flow, in the case of cooling media passing through the gap, can be included to obtain a better estimate. Also, the surface roughness of the stator tooth and rotor surface affects windage loss and must be taken into account

Efficiency –

The overall efficiency of an electric motor can be calculated as a ratio of the input power and the output power. The power can be calculated either by multiplying the instantaneous current and voltage values of calculating it with the torque and rpm values.

Po = τ×2π×rpm/60

Pi = V x I

Efficiency (η) = (τ×2π×rpm/60) / (V x I)

The efficiency of the motor is in the range of 80-98% depending on the loading conditions and variation in the load.

When no data is provided from the manufacturer, the above formulas should be used.

Emrax provides with a tested efficiency map. For a given torque and rpm the Emrax 228 motor has a set efficiency which has been provided as an efficiency map by the manufacturer. This map shows the variation of the efficiency for a particular torque and rpm values.

From this map, we can calculate the instantaneous efficiency of the motor.  This will help us to calculate the heat load of the motor for a given circuit.

We assume that all the inefficiency is converted to heat and all the heat is dissipated through the cooling system and hence the water/coolant from the water jacket will carry the heat to the radiator and dissipate it to the air.

To use the values provided by the manufacturer we need to make a lookup table and use the data in an “If” condition loop.

Steps to be followed

  1. Increase the scale of the map. The map provided by the manufacturers has the least count of 50Nm and 500 rpm. To get better results we need to refine the least count. This can be done by making additional lines on the graph.
  2. Make a lookup table for the efficiency for a particular torque and rpm value
  3. Simulate the track on any lap time simulation software such as IPG car maker. From this simulation, you can get the values of the following parameters
    • Vehicle speed
    • Motor Torque
    • Motor rpm
    • Current
    • Voltage
    • Electric Power
  4. From the simulated values, for each instance, we can calculate the power output using the above-mentioned formula.
  5. Using the lookup table, we can find out the efficiency at each instant and hence the inefficiency of the system.
  6. Using the two calculated values we can find out the heat to be dissipated per instant and hence the average of these values will be the average heat load to be dissipated by the system.

An example of this will be provided in the simulations.

Motor Controller Heating

The motor controller used in electric vehicles can be liquid-cooled or air-cooled. Taking the example of the Bamocar D3 400/700 motor controller, this controller has been provided with a cooling jacket for a coolant to carry away the heat expelled by the controller.

As per the technical data sheet of the motor controller, the major heating elements in the controller are the diodes and the IGBTs. The manufacturers have provided a graph of Irms vs heat loss from each of these components.

To calculate the total heat loss from these components, we will have to refer to these graphs.

Steps to be followed –

  1. Refer to the datasheet to find out the heating behaviour of the motor controller. In the case of the Bamocar D3 controller, the manufacturers have provided the graphs.
  2. Increase the scale of the graph. The map provided by the manufacturers has the least count of 200W and 100A (Irms vs Power loss). To get better results we need to refine the least count. This can be done by making additional lines on the graph.
  3. Chose values from the graph and note down the values in an excel sheet to develop a similar curve as per the graph provided.
  4. Use these values to find out a best-fit curve for the values so as to find a trend line equation for the heating characteristics for a single diode.
  5. Repeat the process for a single IGBT
  6. To find the best fit curve for the graph, you can draw a spline and find the equation from Matlab as well
  7. Multiply the values of single diodes and IGBTs but the number of components assembled on the controller. In this case, there are 6 of each.
  8. From the simulated values, find the instantaneous values of Irms. Take a suitable factor of safety to curb the inaccuracy of the simulation software to simulate Irms Values.
  9. Apply the equation to find out the heat loss for each instant.
  10. The average of these values will provide the value of the average heat to be dissipated from the controller.


SEE ALL Add a note
Add your Comment

Course Enquiry Form

Sign Up

Course Reviews

Why Choose DIY Course?

   Course work & interactions are 100% online.

   Study at the time and place that suits you.

   24/7 access to course material.

   Learn from world-class experts in their field.

Guidebook-for-Mechanical Engineers | DIYguru

Check your Certificate Code

Copyright © DIYguru Education & Research Pvt. Ltd.