The Impact of Wafters on the Thermal Properties and Performance of In-Wheel Motor

Abstract

Electric vehicle (EV) proliferation is accelerating, characterized by the rising quantity of electric automobiles on global roadways. The electric machine is a crucial component of an EV, and the heat generated within the motor requires consideration as it impacts performance and longevity. A prevalent form of machine in EV is the in-wheel motor (IWM), which is notable for its compact size. However, it presents more significant cooling challenges. This research offers a new cooling method to cool the IWM. The system consists of wafters mounted on the housing of the IWM. Testing was conducted to determine the effect of wafters on the thermal properties and performance of IWMs. The machine used in this research is a brushless direct current (BLDC) motor featuring an outer rotor configuration and a peak power output of 1.5 kW. Testing was carried out experimentally and by simulation, and the simulation used Ansys Motor-CAD software. The research results show that applying wafers to IWM reduces the temperature of IWM components by up to 13.1%. IWM with wafters results in a torque increase of 0.14%, a power increase of 0.64%, and an efficiency improvement of 0.6% compared to IWM without wafters.

1. Introduction

Electric vehicle (EV) sales rise annually. Global electric car sales in 2021 surged to 6.6 million, marking a substantial growth compared to the 120,000 sales recorded in 2012. In 2021, there were approximately 16.5 million electric automobiles on the world’s roads, three times the number in 2018. In 2022, sales of electric vehicles (EVs) were expected to expand significantly, with 2 million electric vehicles sold in the first quarter [1]. The rising sales of electric vehicles can be attributed to various reasons, including government legislation. The rules may involve minimal fees on electric vehicles, such as the 0% retail tax for electric vehicles in Indonesia or providing incentives for electric vehicle consumers in Malaysia [2]. Reducing taxes and providing incentives lowers the price of electric vehicles so that they are more affordable for the public, which causes sales levels to increase. The public’s awareness of the dangers of greenhouse gas emissions to the Earth’s atmosphere also influences their switching to electric vehicles [3]. The annual increase in cars equipped with internal combustion engines (ICE) leads to a corresponding rise in carbon dioxide (CO₂) emissions discharged into the environment [4]. CO₂ emissions can be reduced using more electric vehicles [5].

An electric vehicle comprises various components, with the electric motor being vital. The rise in temperature in electric motors must be prioritized as it directly affects the motor’s performance and longevity [6]. The primary heat sources in electric machines are the losses in the copper and iron components [7]. High temperatures in electric motors can lead to demagnetization of the magnets [8], insulating material damage [9], decreased efficiency [10], shortened motor lifespan [11], and potential motor burnout [12]. An increase in the temperature of a motor leads to an escalation in the resistivity of copper. Higher resistivity reduces the electrical voltage flowing through the copper, thereby reducing the machine’s efficiency [13].

The cooling of electric machines has been the subject of many studies. The end windings on electric motors are immersed in oil to cool and effectively reduce heat in the components [14]. An oil spray cooling system performs better than a conventional oil cooling system. A conventional oil cooling system is one in which the oil used for cooling remains stationary at the bottom of the motor and does not circulate out of the motor [15]. Integrating a radiator into the oil spray cooling system improves its cooling effectiveness. At base-speed operational conditions (4000 rpm), the coil temperature when using a radiator is 20 °C lower than without a radiator. At maximum-speed operational conditions (11,000 rpm), the coil temperature is 138 °C, while without a radiator, the coil temperature exceeds 150 °C [16].

Besides the use of oil as a cooling medium, water can also cool electric motors by creating spiral-shaped water cooling channels. The implementation of spiral-shaped water cooling ducts resulted in a 25–32% temperature reduction for the windings, stator, rotor, and permanent magnets, compared to the absence of a cooling system [17]. The implementation of spiral-shaped cooling channels, utilizing water as the cooling medium, results in a decrease in the average temperature of the machine from 90.85 °C to 83.85 °C [18]. An axial water jacket produces better heat reduction than a circumferential one [19]. Refrigerant is an alternative to water to cool electric motors [20].

Another technique for cooling electric motors involves utilizing airflow by incorporating grooves into the motor enclosure. Grooves augment the motor cover’s surface area, resulting in a larger contact area with the surrounding air and facilitating heat dissipation from the motor, reducing its temperature [21]. The presence of fins on the housing of an electric motor has an impact on its cooling. Fins that are placed axially provide superior cooling performance compared to fins installed radially [22]. Permanent magnet synchronous motors (PMSM), equipped with phase change material (PCM) of the paraffin type, effectively reduce the motor’s temperature. Paraffin is injected into a hollow electric motor cover that contains specifically designed cavities, eliminating the need for a complex cooling system to cool the motor [23]. Silicone gelatin in the end space can cool the PMSM when working in high operational conditions [24]. Using flat heat pipes (FHP) on electric motors also reduces heat in the winding and eliminates copper loss by up to 80% [25].

An intriguing study approach for cooling electric motors involves the utilization of fan blades (wafters) to lower their temperature. Kang et al. [26] researched the use of fan blades to cool electric motors with internal rotor types. The fan blades are mounted on the rotor so that they rotate with the rotor, thereby increasing airflow in the machine. The application of wafters in this research succeeded in reducing heat in the magnet, and air resistance loss was reduced by 52%. Satrustegui et al. [27] used wafters on an induction-motor-type electric motor with an internal rotor type. According to their study, the temperature at the end of windings decreased by 16.2 °C due to using fan blades (wafters). Staton et al. [28] compared the results of experimental research with lumped thermal analysis networks on internal-rotor-type induction motors, in which wafters are installed on the induction motor. The findings demonstrate the lumped thermal analysis network’s reliability for investigating electric motors’ thermal properties. All the research on using wafters to lower the temperature of electric motors was conducted exclusively on electric motors with an internal rotor configuration.

This study presents an innovative cooling system that utilizes fan blades (wafters) on electric motors with an external rotor configuration. External rotor electric motors, commonly called in-wheel motors (IWM), are extensively utilized in EVs. Using IWM in EVs eliminates powertrain components such as clutch, transmission, differential, and axle shaft, making the chassis structure simple [29]. The use of an in-wheel motor also increases vehicle stability [30]. However, the downside is that in-wheel motors are more difficult to cool [31]. Fan blades (wafters) have never been applied to IWM with BLDC type. In addition to testing the effect of wafters on the machine’s thermal properties, this study also tested the effect of wafters on the torque and power produced by the BLDC in-wheel motor.

2. Materials and Methods

2.1. Experimental Test Setup

The experimental configuration is depicted in Figure 1. The electric motor in this study is powered by 48 volts of direct current (DC) electrical energy. Testing was conducted without load and at a rotational speed of 485 rpm. The tachometer is used to determine the speed of the test machine. The environmental temperature when the experiment was carried out was 28 °C. A thermographic camera is used to know the electric motor’s thermal characteristics. The thermographic camera brand utilized is Fluke Ti401 Pro, boasting a precision of ±2 °C. Figure 2 shows the outcomes obtained from a thermographic camera on an in-wheel motor.

The test motor used in this study is an external rotor BLDC with a power of 1500 watts. The test motor has 48 permanent magnets mounted on the rotor and 54 slots on the stator. One permanent magnet is 30 mm long and 2 mm thick. The stator has an exterior diameter of 0.258 m, while the rotor has an outer diameter of 0.271 m. The distance between the permanent magnet and the stator is 1 mm.

Table 1. Specifications of the test motor.

Specification	Dimension (m)	Specification	Dimension (m)
Air gap	0.001	Outer diameter of stator	0.258
Length of a magnet	0.03	Stator lamination length	0.03
Bearing diameter	0.045	Thickness of the magnet	0.002
Length of the rotor	0.05	Shaft diameter	0.028
Rotor outer diameter	0.271	Slot width	0.009
Slot aperture	0.0017	Slot depth	0.017

provides more comprehensive parameters for the test motor, while Figure 3 and Figure 4 depict the construction of the test motor.

2.2. Proposed Technique for Cooling In-Wheel Motors

The novel cooling system for the IWM consists of wafters installed on the end cap IWM. The wafters are composed of polylactic acid (PLA) polymer. Using PLA material on the wafters causes an insignificant increase in the in-wheel motor’s weight—the total weight of the wafters is only 24 g. The use of wafters has no impact on the electromagnetic efficiency of the electric machine [28]. Figure 5 and Figure 6 show the proposed technique for cooling in-wheel motors.

This study aims to employ wafters to enhance airflow in the end cavity of the IWM. The wafters cause the air in the end cavity to be high-speed and turbulent [26]. This airflow is expected to absorb the heat from the rotor, stator, winding, and magnet by convection. Convection is the transfer of heat to or from a solid surface via fluid motion over that surface, with the rate of heat transfer influenced by both the fluid’s thermal properties and the fluid flow features [32]. The rate of heat convection $\left(q_s\right)$ is calculated using Newton’s equation of cooling, which relates the surface temperature $\left(T_s\right)$ to the air temperature at a considerable distance from the surface $\left(T_f\right)$:

$$q_s=hA\left(T_s-T_f\right)$$

(1)

where A represents the surface area, while h denotes the heat transfer coefficient. The fluid’s properties and flow characteristics over the surface influence the heat transfer coefficient.

In fluid mechanics, a fundamental concept states that fluid molecules in contact with a solid surface cling to it and do not exhibit slide. During motion, a fluid close to the wall glides over another, increasing velocity from zero at the wall to the mainline speed. A fluid adjacent to the surface, termed the hydrodynamic boundary layer, exhibits a particular velocity profile and moves at a reduced speed. The magnitude of shear stress is contingent, depending on the fluid’s viscosity and the velocity gradient at the boundary, as per the following equation:

$$\tau =\mu {\displaystyle \frac{du}{dy}}$$

(2)

Without fluid movement at the surface, heat transfer into the fluid occurs solely through conduction from the adjacent wall region. As heat is transmitted, the fluid layers adjacent to the wall experience a temperature rise. The heat conduction rate is formulated as follows:

$$q_s^{”}=-k{\left({\displaystyle \frac{\partial T}{\partial y}}\right)}_{y=0}$$

(3)

The heat transfer coefficient can be expressed by integrating Equations (1) and (3), as follows:

$$h=-{\displaystyle \frac{k}{\left(T_s-T_f\right)}}{\left({\displaystyle \frac{\partial T}{\partial y}}\right)}_{y=0}$$

(4)

The fluid exhibits laminar flow at low velocities. At low speeds, heat transmission across the boundary layer occurs solely by conduction through the fluid layers in contact with the surface. The fluid’s thermal conductivity influences the heat transmission rate. As fluid velocities increase, the flow transitions to turbulence, forming vortices within the boundary layer. These vortices enhance heat transmission between the wall and the flow. The turbulence-induced mixing in most boundary layers reduces velocity and temperature gradients. As a result, the velocity and temperature gradients are elevated near the wall, leading to turbulent flow that produces increased heat transfer coefficients. A dimensionless parameter called the Reynolds number (${Re}_x$) dictates flow characteristics:

$${Re}_x={\displaystyle \frac{\rho ux}{\mu }}$$

(5)

The Reynolds number denotes the inertial-to-viscous forces ratio in fluid dynamics. For Reynolds values below 35,000, the flow is laminar. For Reynolds numbers beyond 4,000,000, the flow is completely turbulent.

Convective cooling at the endcaps of an electric machine is complex because it is difficult to accurately predict fluid flow and the heat transfer coefficient across the diverse surfaces in the end space. A general formulation for calculating the heat transfer coefficient has been developed as a result of the analysis of the heat transfer phenomena in the end cap region of entirely enclosed air-cooled motors by numerous authors, as follows:

$$h=k_1\left[1+k_2{V}^{k_3}\right]$$

(6)

The local air velocity is denoted by V. Empirical parameters are employed, specifically the constants $k_1$, $k_2$, and $k_3$. The correlation models for end-space cooling include radiation, forced convection, and natural convection. When the reference velocity is null, $k_1$ denotes natural convection and radiation. The additional forced convection from rotation is represented by $k_1{k}_2{V}^{k_3}$.

2.3. Simulation Test

Simulation studies were performed to investigate the thermal properties and performance of the IWM utilizing wafters within IWM housing. This study conducted experimental tests on IWM exclusively under no-load conditions. Thus, the impact of utilizing wafters is understood solely through the thermal characteristics of the IWM. In contrast, the effect of wafters on motor performance, such as torque, power, and efficiency, is unknown. In addition, simulation tests can undertake challenging and expensive work if they have to be performed experimentally, such as analyzing duty cycles on electric motors. The duty cycle in an electric motor is the percentage of time in which the motor can operate continuously at a specific load without experiencing damage due to overheating.

This research used the lumped capacity method for thermal simulation, with Ansys Motor-CAD 15.1 as the analytical program. The lumped capacity technique presumes a uniform temperature throughout an object or a designated area of an object, seeing any temperature gradients as insignificant compared to the temperature disparity between the object and its surroundings [32]. The heat exchange with the environment is examined, and a formula is developed to describe the temporal fluctuation of the object’s temperature.

The rate of heat transfer is expressed as follows:

$$q=hA\left(T-T_{\infty }\right)$$

(7)

The cooling rate of the object, as heat is transferred away, is described by the following equation:

$$q=-\rho Vc{\displaystyle \frac{\partial T}{\partial t}}$$

(8)

Combining Equations (7) and (8) gives

$$hA\left(T-T_{\infty }\right)=-\rho Vc{\displaystyle \frac{\partial T}{\partial t}}$$

(9)

$$T-T_{\infty }=\left(T-T_{\infty }\right)e^{-\left({\displaystyle \frac{hA}{\rho Vc}}\right)t}$$

(10)

For an object of arbitrary shape, the internal thermal resistance $\left(R_{thint}\right)$ is defined as

$$R_{thint}={\displaystyle \frac{s}{kA}}$$

(11)

External thermal resistance $\left(R_{thext}\right)$ can be defined as

$$R_{thext}={\displaystyle \frac{1}{hA}}$$

(12)

Dividing (11) by (12) yields a non-dimensional relationship called the Biot number (Bi).

$$Bi={\displaystyle \frac{hs}{k}}$$

(13)

The validity of lumped capacity analysis is contingent upon a sufficiently small Biot number. The lumped capacity method is acknowledged to produce sufficiently accurate results when $Bi$ < 0.1. When the Biot number exceeds 0.1, the lumped capacity method becomes inappropriate; therefore, it is essential to consider temperature gradients within the object and to solve the heat conduction equation. The general heat conduction equation can be reduced for one-dimensional heat conduction in the absence of internal heat generation to

$${\displaystyle \frac{{\partial }^{2}T}{{\partial x}^2}}={\displaystyle \frac{1}{\alpha }}{\displaystyle \frac{\partial T}{\partial t}}$$

(14)

Figure 7 illustrates the IWM image generated from the simulation. Components of the IWM that do not substantially affect the simulation outcomes, such as bolts and nuts, are excluded. The proportions of the IWM are identical to those of the genuine IWM.

Table 2. The IWM’s winding specifications.

Specification		Specification
Coiled configuration	Lap	Wire slot fill	61%
Path type	Central	Copper area	60.18 mm2
Phases	3	Winding area	123.1 mm2
Turns	4	Wire diameter	0.79 mm
Throw	1	Number of strands in hand	19
Parallel paths	1	Copper slot fill	56%
Winding layers	2	Slot area	134.3 mm2
Conductor/slot	154	Winding depth	15.747 mm

delineates the characteristics of the winding for the test motor.

Grasping the thermal properties of materials utilized in electric motors is essential for examining their thermal characteristics.

Table 3. Thermal characteristics of the test motor material [33].

Component	Material	Thermal Conductivity (W/m·°C)	Specific Heat (J/kg·°C)	Density (kg/m3)
Housing	Aluminum	168	833	2790
Rotor	Silicon Steel	30	460	7650
Stator	Silicon Steel	30	460	7650
Winding	Copper	401	385	8933
Magnet	Neodymium	7.6	460	7500
Axle	Stainless Steel	15.1	480	8055

shows the materials used in the test machine and their specifications. Thermal conductivity is one of the properties of a material that is very important in heat transfer. It indicates how well the material can conduct heat. The higher the thermal conductivity value of a material, the easier it is for heat to travel through it.

The impact of using wafters on the heat and performance of IWM through simulation was carried out by inputting the values of $k_1$, $k_2$, and $k_3$ in Equation (6) obtained from previous research, as shown in

Table 4. Convection coefficient data for end space of electric motor [32].

	${\mathit{k}}_1$	${\mathit{k}}_2$	${\mathit{k}}_3$
Mellor	15.5	0.39	1
Schubert	20	0.425	0.7
Stokum	33.2	0.0445	1
Di Gerlando	40	0.1	1
Hamdi	10	0.3	1

. All data in Table 4 are simulated, and the results are compared with the experimental test results. The simulation results closest to the experimental test results are used as a reference for subsequent IWM simulations using fan blades or wafters.

2.4. Data Analysis

A two-sample t-test was employed to evaluate the test outcomes of IWM utilizing wafters against those of IWM without wafters and to compare the results of simulation testing with experimental tests. The two-sample t-test is a parametric statistical test that compares the means of two independent data sets. The primary purpose of this test is to determine whether there is a statistically significant difference between the two means. The first step taken to conduct a two-sample t-test in this study is to formulate a hypothesis:

$H_0:{\mu }_1-{\mu }_2=0$ (there is no difference between an IWM with and an IWM without wafters).

$H_1:{\mu }_1-{\mu }_2\ne 0$ (there is a difference between IWMs with wafters and those without wafters).

The next stage is to determine the significance level $\left(\alpha \right)$, which in this study is 0.05. The two-sample t-test uses the following equation:

$$t={\displaystyle \frac{{\overline{x}}_1-{\overline{x}}_2}{s_p\sqrt{\frac{1}{n_1}+\frac{1}{n_2}}}}$$

(15)

$$s_p=\sqrt{{\displaystyle \frac{\left(n_1-1\right)S_1^2+\left(n_2-1\right)S_2^2}{n_1+n_2-2}}}$$

(16)

where $s_p$ is the pooled standard deviation, $\overline{x}$ is the data mean, n is the data size, and $s^2$ is the data variance. If the test for equality of variances results in the conclusion that the two populations have different variances, then the two-sample t-test statistic is as follows:

$$t={\displaystyle \frac{{\overline{x}}_1-{\overline{x}}_2}{\sqrt{\frac{S_1^2}{n_1}+\frac{S_2^2}{n_2}}}}$$

(17)

The p-value is used, which is compared with the level of significance $\left(\alpha \right)$, where if the p-value < $\alpha$, $H_0$ is rejected.

3. Results and Discussion

Figure 8 depicts the heat in the IWM housing with and without wafters based on a real test. During testing, the motor operates without load. The IWM operates at a velocity of 485 revolutions per minute throughout the testing period. The test duration is 11,400 s, or 190 min if converted to minutes. The housing temperature appears to increase as the electrical motor operating time increases. After the electric motor started operating, it was seen that the housing temperature on the IWM without wafters showed a higher temperature increase compared to the housing temperature on the IWM with wafters. At 11,400 s, the housing temperature on the IWM with wafters is 35.4 °C, while the housing temperature on the IWM without wafters is 37.6 °C, so it can be concluded that wafters can reduce the housing temperature by 5.85 °C if the IWM operates under no-load conditions for 190 min. However, if tested statistically, the heat decrease of 5.85 °C from the use of wafters under no-load operating conditions is not significant because the p-value obtained (0.082) is higher than the significance level (0.05), so $H_0$ is accepted, and $H_1$ is rejected. Additional testing is required to evaluate the impact of the wafter’s application on the IWM under loaded conditions.

Figure 9 compares the heat rise of the IWM endcap without wafters between the simulation and experimental tests under no-load conditions. The p-value from the two-sample t-test is 0.656, indicating that $H_0$ is accepted because the p-value is more significant than α. $H_0$ is accepted, which means that the temperature in the IWM housing without wafters is the same in both the simulated and the experimental tests. The difference between the experimental test results and the simulation on the IWM without wafers is 1.3%.

Figure 10 shows the heat rise of the IWM endcap with wafters between the experimental testing and simulations under no-load conditions. Of the five simulation tests, visually, the simulation (Hamdi) showed results that were closest to the experimental test results. A two-sample t-test statistical test was conducted between the experimental and simulation test results. The p-value of the experimental test with simulation by Hamdi [34] is 0.527, the p-value of the experimental test with simulation by Mellor [35] is 0.105, the p-value of the experimental test with simulation by Schubert [36] is 0.057, and the p-value of the experimental test with simulation by Stokum [37] is 0.047. The p-value of the experimental test with simulation by Di Gerlando [38] is 0.007. Based on the statistical test, the simulations by Hamdi [34], Mellor [35], and Schubert [36] have a p-value that is greater than the alpha value, namely 0.05, so $H_0$ is accepted, which means there is no difference between the simulation results and the experimental test results. The one that has a more remarkable similarity to the experimental test results is the simulation by Hamdi because it has a more significant p-value. The p-values of the simulations by Stokum and Di Gerlando are smaller than the alpha value, so $H_0$ is rejected, and $H_1$ is accepted, so there is a difference between the results of the experimental test and the simulation results for the IWM equipped with wafters. From the results of the visual observations and statistical tests, it can be concluded that the convection coefficient from the research conducted by Hamdi is similar to the convection coefficient from the use of wafers in this study, so for the following analysis, the end-space cooling correlations data from Hamdi will be used, as shown in Table 4. The error between the experimental test and the simulation (Hamdi) was only 1.56%, an acceptable error value.

Figure 11 illustrates the thermal properties of crucial components of the IWM, specifically the windings, stator, rotor, magnets, and housing, utilizing wafters at an output power of 1.5 kW and a torque of 15 Nm. The data were acquired from a simulated test. To achieve a power output of 1.5 kW, a peak current of 24.94 A (amperes) and a DC bus voltage of 70 V (volts) are required. The RMS current is 17.64 V, the RMS current density is 2.344 V, and the winding connection is a star connection. The graphic indicates that the winding of the IWM component reaches the highest temperature. The stator temperature exceeds that of the rotor, whereas the rotor and permanent magnet temperatures are nearly identical, with the housing exhibiting the lowest temperature. The temperature of the IWM components utilizing wafters is lower. The temperature of the in-wheel motor components with wafters attains a steady state more rapidly than that of the IWM components without wafters. The IWM with wafters attains a steady temperature in 53,010 s (883.1 min), but the IWM without wafters stabilizes at 62,700 s (1045 min). The stable temperature of the IWM components with wafters is achieved 162 min faster than the IWM without wafters.

Table 5. Temperature drop in IWM components due to wafters application.

Component	Without Wafters	Wafters	Temperature Drop
Housing	50 °C	49 °C	2%
Magnet	65 °C	60 °C	7.7%
Stator	128 °C	111 °C	13.3%
Rotor	65 °C	61 °C	6.2%
Winding	132 °C	116 °C	12.1%

shows the temperature drop in the IWM components due to wafters application.

Comparing the proposed cooling system to an engine without a cooling system provides a baseline for evaluating its effectiveness. However, to honestly assess its potential and identify areas for improvement, it is crucial to compare it to other established cooling systems. Chen et al. [17] applied a spiral-shaped channel cooling fluid system, in which the outer rotor in-wheel motor studied had a power of 10 kW. The results show that the winding temperature is reduced by 32%, the rotor temperature is reduced by 30%, the permanent magnet temperature is reduced by 26%, and the rotor temperature is reduced by 25%. In another study, conducted by Dajun et al. [39], the type of cooling system used was a water cooling system with a spiral cooling channel. The in-wheel motor power tested was 75 kW. The results showed that the stator temperature was reduced by 27%, the winding temperature was reduced by 24%, the rotor temperature was reduced by 10%, and the permanent magnet temperature was reduced by 10.3%. In this study, the rotor temperature only decreased by 6.2%, the stator temperature decreased by 13.3%, the winding temperature decreased by 12.1%, and the permanent magnet temperature decreased by 7.7%. The proposed cooling system is less able to cool the in-wheel motor than a water cooling system with a spiral channel. However, the main advantages of this proposed cooling system are cheaper costs and more straightforward installation compared to the water cooling system. In the water cooling system, additional components such as fluid pumps and radiators are needed, and additional power is required to operate these systems. However, in the future, further research will be needed to optimize the potential of this wafter cooling system so that the cooling performance of the system becomes better.

Figure 12 illustrates torque versus speed for an IWM both with and without wafters, whereas Figure 13 depicts power versus speed for an IWM under the same conditions. Applying wafers to the IWM marginally enhances the torque generated by the IWM, while the increase is not substantial. The mean torque generated by the IWM without wafters is 14.23 Nm, whereas the mean torque generated by the IWM with wafters is 14.25 Nm. Applying wafters to the IWM enhances the torque generated by the IWM by 0.14%. The average power generated by the IWM without wafers is 775.3 watts, whereas the average power generated by the IWM with wafters is 780.3 watts. Applying wafters to the IWM enhances its power output by 0.64%.

Figure 14 plots the efficiency of an IWM with and without wafters against the rotating speed. The IWM’s efficiency with wafters differs from that without wafters, but the difference is minor. The average efficiency of an IWM with wafters is 92.6%, whereas that of an IWM without wafters is 92.1%. Applying wafters to the IWM improves the machine’s efficiency by 0.6%.

4. Conclusions

This research introduces a novel cooling mechanism for electric machines in electric vehicles, especially for IWMs. The cooling system involves installing wafters on the IWM endcap. Testing was conducted on the BLDC IWM with the rotor positioned externally to assess the impact of wafers on the thermal properties and the performance of the electric motor. The electric motor comprises 48 poles and 52 slots. The maximum achievable power output is 1.5 kW. Experimental methods and simulations were performed in this research. An experimental investigation was performed under no-load conditions at an IWM rotation speed of 485 rpm, utilizing a thermal camera to ascertain the operating temperature of the electric motor. The simulation test was conducted utilizing Ansys Motor-CAD software. The research findings indicate that applying wafters on IWMs effectively lowers the operational temperature of the components. The performance metrics of the IWM, including torque, power, and efficiency, improved after applying the wafters. In no-load situations, the housing temperature of the IWM with wafers is 1.5 °C lower than that of the IWM without wafters. The advantages of wafters in cooling IWM components become particularly obvious when the IWM functions under load conditions, generating 1500 W of output power. The wafters on the IWM effectively lower the winding temperature from 132 °C to 116 °C. Applying wafters resulted in a torque enhancement of 0.14%, a power gain of 0.64%, and an efficiency improvement of 0.6% for the IWM.