Modification of the In-Wheel Motor Housing and Its Effect on Temperature Reduction

Abstract

This research proposes a novel cooling system to minimize the external rotor type of electric motor temperature by installing fan blades (wafters) on the inner housing of the electric motor. Fan blades (wafters) are made by printing using 3D printer technology and using polylactic acid (PLA) as the material. Wafters are then installed on an in-wheel motor with a power of 1500 W, having 48 poles and 52 slots. The study included thermal simulation and experimental techniques to ascertain how fan blades (wafters) affected the electric motor’s thermal properties. The motor rotated at 500 rpm during the experimental test with no load condition. The temperature of the electric motor is known using an infrared thermal imager. Using Ansys Motor-CAD 15.1 software, thermal modeling employs the lumped circuit model approach. Thermal simulation results show almost the same results as the experimental test results. Applying wafters on the in-wheel motor housing significantly reduces the winding temperature by 3.047 °C or experiences a temperature reduction of 4.34%. Using wafters in the in-wheel motor housing also speeds up the stable state temperature of the electric motor by 9 min compared to in-wheel motors without wafters.

1. Introduction

Electric vehicles (EVs) are a technology that has great potential to replace fossil fuel vehicles. Electric vehicles will disrupt road transportation, and leading international manufacturers now consider this their priority [1]. The shift in the automotive power system from internal combustion engines (ICE) that use fossil fuels to electric motors that use electrical energy is because production and reserves of fossil fuels, especially petroleum, continue to decrease. For example, several decades ago, Indonesia was a top oil exporter that believed it had a vast fossil fuel reserve. Around 2004, Indonesia had to become an oil net importer due to lower oil production than demand [2]. Using electric vehicles is expected to slow down the depletion of fossil fuels because the efficiency of electric vehicles is higher than vehicles that use internal combustion engines [3]. Using electric vehicles also reduces CO₂ emissions due to using internal combustion engines [4]. In addition, the electric vehicles now offered on the market, such as the Renault Twizy, have proven themselves dependable. After 4.5 years and covering a distance of over 30,000 km, there were no faults in the electric drive system, and the battery remained in excellent condition [5].

The government’s role in increasing the use of electric vehicles is also huge at this time. This role can take the form of providing many supporting facilities for electric vehicles, such as electric vehicle charging infrastructure, as well as creating various policies, such as reducing taxes for electric vehicles, increasing taxes for vehicles that use fossil fuels, and providing incentives for buyers of electric vehicles [6]. Current discoveries in battery technology for electric vehicles (battery storage systems) are also extraordinary, so an electric vehicle can cover a distance of more than 500 km and have a speedy charging time [7]. Therefore, people’s desire to use electric vehicles, both plug-in hybrid electric vehicles (PHEV) and battery electric vehicles (BEV), tends to increase [8].

The electric motor is one of an electric vehicle’s most essential components, whose temperature significantly impacts the electric motor’s power density [9]. The resistivity of coil materials like copper will rise with increased motor temperature, impacting winding loss. As resistivity increases, there will be a decrease in electrical voltage flowing through the coil, which will affect the motor’s performance [10]. A higher temperature will also increase the resistivity of the aluminum material, which raises the resistance of the rotor and stator and reduces the motor’s capacity to operate at full load and under magnetic field saturation. As the stator and rotor resistance increases with increasing temperature, total losses increase and efficiency decreases [11]. Deterioration of the insulation materials, burnout of motors, reduction in efficiency, and demagnetization of magnets are all consequences of poor thermal management [12].

In-wheel motors are a kind of electric motor found in electric vehicles. Electric motors can be installed on each vehicle wheel using in-wheel motor technology. Each wheel can rotate in the opposite direction, increasing vehicle stability control or VSC [13]. Compactness, a high interior vehicle space utilization ratio, a reduced center of gravity, exceptional driving stability, easy and intelligent control, and many other advantages are all provided by in-wheel motors [14]. Apart from these advantages, in-wheel motors also have disadvantages regarding heat transfer. Due to the small working space, temperature increases immediately impact its lifespan and efficiency [15].

Some investigators have looked into ways to cool electric motors, like making grooves on the outside of the motor housing [16], installing heat pipe loops [17], using spiral water jackets [18], and using liquid cooling channels [19]. Oil can be used as a medium to cool electric motors and performs better than water in cooling electric motors [20]. Adding a radiator to the internal oil circulation cooling system can reduce the electric motor temperature by 20 °C compared to without using a radiator [21]. Inserting phase change material (PCM) into the hollow cavity of the electric motor housing can reduce the temperature of the electric motor. The rise in temperature of an electric motor with a housing equipped with PCM is smaller than a housing of the same size without PCM [22]. Adding silicone gelatin potting material in the gap between the end winding and the electric motor casing reduces the temperature of the electric motor [23]. An axial flux permanent magnet (PM) generator with potting positioned between the end winding and the frame can lower the temperature by 10% compared to an axial flux PM generator without potting [24].

The stator of an electric motor has a higher temperature than the rotor, with the stator copper coils having the maximum temperature during operation [25]. When an electric vehicle accelerates, the temperature in the winding will increase significantly, even exceeding the insulation level [26]. The wholly closed construction of the in-wheel motor with no incoming and outgoing air makes the heat removal process more difficult. This research proposes a novel cooling system to reduce the temperature of the in-wheel motor. The fan blade is installed on the inner housing of the external type of electric motor (in-wheel motor) to reduce heat in the in-wheel motor, which has yet to be studied. The fan blades (wafters) increase airflow in the end space inside the in-wheel motor.

In previous research, as conducted by Kang et al. [27], wafters are applied to an internal permanent magnet synchronous machine (IPSM) with an inner rotor type. The wafter is installed on the rotor so that when the rotor rotates, the wafter also rotates, causing the forced airflow at the end region to increase, thus helping the electric motor cooling process. Satrustegui et al. [28] apply a wafter to electric motors with internal rotor type and wafter installed on the rotor. However, the type of electric motor used is an induction motor. Staton et al. [29] also apply wafters to electric motors with internal rotor type and aluminum wafters installed on the rotor. The type of electric motor used is an induction motor. Micallef et al. [30] researched wafters in totally enclosed fan-cooled (TEFC) induction machines with internal rotor type, and the rotor is the location of the wafter placement. Computational fluid dynamics (CFD) analysis was performed by Jungreuthmayer et al. [31] in the case of using fan blades installed on the internal permanent magnet synchronous machine, internal rotor type, resulting in decreased temperature in the electric motor. Among these studies, no one has researched the use of wafters in electric motors with external rotor types and wafters installed in the housing. Electric motors with the external rotor type are more challenging to cool because the winding and stator construction, the components of electric motors that experience the highest heat, are located inside the engine, so they do not come into direct contact with outside air. Aside from that, using wafters to cool electric motors with brushless direct current motors (BLDC) has never been attempted, as previous research only examined internal permanent magnet synchronous machines (IPMSM) and induction motors regarding the use of wafters.

2. Materials and Methods

2.1. Description of the Proposed Cooling System

This study uses an in-wheel brushless direct current (BLDC) motor with an outer rotor topology and permanent magnet. The electric motor has a rated power of 1500 watts, consisting of 48 poles and 54 slots.

Table 1. Specifications of the in-wheel motor BLDC under study.

Parameter	Value	Parameter	Value
Rotor type	Outer rotor	Stator outer diameter	258 mm
Rated power	1.5 kW	Air gap length	1 mm
Number of poles	48	Magnet thickness	2 mm
Number of slots	54	Axle diameter	28 mm
Rotor outer diameter	271 mm	Slot width	9 mm
Slot opening	1.7 mm	Slot depth	17 mm
Magnet length	30 mm	Stator lamination length	30 mm
Rotor length	50 mm	Bearing diameter	45 mm

lists the specifications for the in-wheel motor utilized in the study, and Figure 1 depicts the in-wheel motor’s construction as seen from the radial and axial directions. Figure 2 depicts the inside of the in-wheel motor and some of the electric motor’s essential components, including the coil, permanent magnet, stator, rotor, and housing.

The proposed cooling system involves fan blades (wafters) installed on the housing inside, as seen in Figure 3. Figure 3a shows six fan blades (wafters) installed in the in-wheel motor housing, and Figure 3b shows the dimensions of a wafter. Wafters use 3D printing technology from polylactic acid (PLA) material. Because the weight of the fan blades (wafters) is just 24 g, adding them to the housing does not considerably escalate the mass of the in-wheel motor. Using fan blades (wafters) also does not affect the motor’s electromagnetic performance [29].

2.2. Measurement Setup

There were two stages of the testing. The initial stage involves taking the in-wheel motor’s temperature under normal circumstances or without changing the in-wheel motor housing. The second stage involves measuring the temperature of the in-wheel motor after modifications (adding fan blades or wafters to the housing interior). Figure 4 displays the experimental setup used for both testing stages.

The electrical energy used in this investigation to power the electric motor is 48 V DC. Tests were carried out under no load conditions and tested at a constant speed for each experiment. The rotational speed of the in-wheel motor during experimental testing was 485 rpm due to mechanical limitations. The average ambient temperature at the time of testing was 28 °C. The in-wheel motor’s temperature is known using an infrared (IR) thermal camera. The brand of thermal camera used is Fluke Ti401 Pro with accuracy ±2. Fluke Corporation, an American industrial test, measurement, and diagnostic equipment manufacturer, manufactures Fluke Ti401 Pro. The headquarters of Fluke Corporation is in Washington, United States. Figure 5 shows an example of temperature distribution on an electric motor using a thermal imager.

2.3. Thermal Models

A thermal model is required to confirm experimental test results and provide information about the thermal characteristics of multiple in-wheel motor components. Several thermal modeling techniques exist for electric motors: numerical methods (finite element analysis and computational fluid dynamics) and lumped circuit models [32]. Thermal analysis with finite element analysis (FEA) and computational fluid dynamics (CFD) is more time-consuming than other methods. It demands a lot of computer power, but its main advantage is that it can analyze devices with any geometry and use any cooling system [33]. Meanwhile, the lumped method shows better and faster results in analyzing the thermal behavior of electric motors. Numerical approaches (FEA and CFD) are vulnerable to external convection parameters, which are difficult to estimate and must be weighed against experimental results first. In the lumped method, the thermal resistance is calculated using theoretical equations and does not need to consider experimental results [34]. This research uses lumped circuit models to model the thermal characteristics of in-wheel motors and uses ANSYS Motor-CAD software.

In electric motor thermal analysis, the following are two of the trickiest elements:

• The convection heat transfer forecast is primarily concerned with the outside surface of the motor and the interior air gap. CFD can solve this with much effort. Meanwhile, using lumped circuit models, the correct convection heat transfer estimate only uses natural and forced convection correlation.
• Defining internal interface resistance, especially gap size on the motor. Because of the significant aspect ratio elements, this aspect is challenging to represent using CFD. The benefit of adopting a lumped circuit approach is that it requires little effort to execute numerous calculations with varied interface gaps.

Using the well-established convection heat transfer correlation (natural and forced convection correlation), Motor-CAD automatically computes convection from all the motor’s external surfaces [32]. The general form of convection correlation for natural convection is as follows:

$$\mathrm{N}\mathrm{u}=\mathrm{a}{\left(\mathrm{G}\mathrm{r}·\mathrm{P}\mathrm{r}\right)}^{\mathrm{b}}$$

(1)

Nu represents the Nusselt number, a and b are constants, Gr represents the Grashof number, and Pr represents the Prandtl number. Almost all natural convection correlation applications are related to the housing’s external cooling system. The following is the general form of convection correlation for forced convection:

$$\mathrm{N}\mathrm{u}=\mathrm{a}{\left(\mathrm{R}\mathrm{e}\right)}^{\mathrm{b}}{\left(\mathrm{P}\mathrm{r}\right)}^{\mathrm{c}}$$

(2)

Re denotes the Reynolds number, and a, b, and c are constants. When the Reynolds number exceeds 2300, the flow transitions typically from laminar to turbulent, and when the Reynolds number exceeds 5000, the flow becomes entirely turbulent. When the Reynolds number exceeds 500,000, there is a general transition from laminar to turbulent flow for geometries with external flow.

Staton et al. [35] define the end space as the area within the end shields that contains the end winding and cover. The end spaces are one of the most challenging areas of machine cooling to accurately predict because fluid flow is affected by a variety of factors, including the shape and length of the end winding, added fanning effects caused by simple fans and end-ring wafters, the surface finish of the rotor’s end sections, and turbulence. Several writers have investigated the cooling of interior surfaces near end windings. In many circumstances, they recommend using a formulation like the one below:

$$\mathrm{h}={\mathrm{k}}_1\left(1+{\mathrm{k}}_2·{\mathrm{v}\mathrm{e}\mathrm{l}}^{\mathrm{k}3}\right)$$

(3)

where $\mathrm{h}$ is the heat transfer coefficient $\left(\mathrm{W}/{\mathrm{m}}^2/\mathrm{C}\right)$, ${\mathrm{k}}_1$, ${\mathrm{k}}_2$, ${\mathrm{k}}_3$ are curve fit coefficients, and vel is the local fluid velocity (m/s). The ${\mathrm{k}}_1$ term accounts for natural convection, and the ${\mathrm{k}}_1\times {\mathrm{k}}_2\times {\mathrm{v}\mathrm{e}\mathrm{l}}^{\mathrm{k}3}$ term accounts for the added forced convection due to rotation.

To make thermal analysis on in-wheel motors easier and faster, the intricate design of the in-wheel motor, as shown in Figure 2, is simplified into building the in-wheel motor, as shown in Figure 6. Bolts, for example, are not examined because they have no significant effect on thermal analysis. The dimensions of the in-wheel motor are the same as those of the actual electric motor, as shown in Figure 6.

Table 2. Winding specifications of the in-wheel motor.

Parameter	Value	Parameter	Value
Coil style	Stranded	Wedge model	Wound space
Divider type	Overlapping	Slot area	134.3 mm2
Wire diameter	0.79 mm	Winding area	123.1 mm2
Conductors per slot	180	Winding depth	15.747 mm
Winding type	Lap	Path type	Central
Phases	3	Turns	6
Throw	1	Parallel paths	1
Liner thickness	0.25 mm	Winding layers	2

lists the winding requirements for the in-wheel motor.

Table 3. The miscellaneous data used for the thermal simulation.

Parameter	Value
Motor orientation	Horizontal
Lamination stacking factor (stator)	0.97
Lamination stacking factor (rotor)	0.97
Ambient temperature	28 $°\mathrm{C}$
Radiation emissivity	0.9
Interface gap between stator lamination stack and axle	0.03 mm
Interface gap between rotor and housing	0.005 mm
Interface gap between housing and endcap	0.005 mm
Interface gap between magnet and rotor lamination stack	0.005 mm
Effective bearing gap (front)	0.4 mm
Effective bearing gap (rear)	0.4 mm
Interface gap between bearing and front endcap	0.0073 mm
Interface gap between bearing and gap	0.0112 mm

shows the miscellaneous data used for the thermal simulation. The environmental temperature used was 28 $°\mathrm{C}$. This temperature is the average temperature during experimental testing. Except for the ambient temperature, other data, as shown in Table 3, are default data in Ansys Motor-CAD software.

Knowing the thermal properties of electric motor materials is very important in analyzing the thermal characteristics of the electric motor. The housing material on the test motor is aluminum. The stator and rotor lamination material is silicon steel (M350-50A), the magnet material is neodymium N30UH, the winding material is copper, and the axle material is stainless steel.

Table 4. Thermal properties of electric motor [31].

Component	Value	Thermal Conductivity (W/m·°C)	Specific Heat (J/kg·°C)	Density (kg/m3)
Housing	Aluminum	168	833	2790
Rotor	M350-50A	30	460	7650
Stator	M350-50A	30	460	7650
Winding	Copper	401	385	8933
Magnet	N30UH	7.6	460	7500
Axle	Stainless Steel	15.1	480	8055

shows the thermal properties of these materials.

The effect of using fan blades (wafters) on the in-wheel motor housing to reduce the temperature of the electric motor, if using Ansys Motor-CAD software, can be achieved by changing the air velocity resulting from the use of the wafters. Therefore, the air velocity at the end space of the electric motor when installing the wafters on the housing needs to be known first. Changes in the area of the internal surface housing of the in-wheel motor due to the use of wafters also affect the thermal characteristics of the electric motor.

2.4. Data Analysis Technique

Perform a normality test to see whether the obtained data follow a normal distribution. The assumptions for a normality test are as follows: H0 for data that follow a normal distribution and H1 for data that do not follow a normal distribution. The Kolmogorov–Smirnov formula is used in this study using the following formulation:

$$\mathrm{D}=\mathrm{m}\mathrm{a}\mathrm{x}\left\{{\mathrm{D}}^{+},{\mathrm{D}}^{-}\right\}$$

(4)

$${\mathrm{D}}^{+}={\mathrm{m}\mathrm{a}\mathrm{x}}_{\mathrm{i}}\left\{\frac{\mathrm{i}}{\mathrm{n}}-{\mathrm{Z}}_{\left(\mathrm{i}\right)}\right\}$$

(5)

$${\mathrm{D}}^{-}={\mathrm{m}\mathrm{a}\mathrm{x}}_{\mathrm{i}}\left\{{\mathrm{Z}}_{\left(\mathrm{i}\right)}-\left(\mathrm{i}-1\right)/\mathrm{n}\right\}$$

(6)

$$\mathrm{Z}=\mathrm{F}\left({\mathrm{X}}_{\left(\mathrm{i}\right)}\right)$$

(7)

where

$\mathrm{F}\left({\mathrm{X}}_{\left(\mathrm{i}\right)}\right)$ is the probability distribution function of the normal distribution;

${\mathrm{X}}_{\left(\mathrm{i}\right)}$ is ith order statistics of a random sample, $1\le \mathrm{i}\le \mathrm{n}$;

n is the sample size.

After the data show a normal distribution, the next stage is to carry out a 2-sample t-test to determine whether the two data groups have similarities or significant differences. Calculate the confidence interval (CI) using the following equation:

$$\left({\overline{X}}_1-{\overline{X}}_2\right)-t_{\frac{a}{2}}\left(S\right) to \left({\overline{X}}_1-{\overline{X}}_2\right)+t_{\frac{a}{2}}\left(S\right)$$

(8)

where

${\overline{X}}_1$ = the first sample’s mean

${\overline{X}}_2$ = the second sample’s mean

$t_{\frac{a}{2}}$ = t distribution’s inverse cumulative probability at 1 − α/2

A = 1 − confidence level/100

s = the computed sample standard deviation for the test statistic

$$t=\frac{\left({\overline{X}}_1-{\overline{X}}_2\right)-{\delta }_o}{s}$$

(9)

For unequal variances, the sample standard deviation of ${\overline{X}}_1-{\overline{X}}_2$ is:

$$s=\sqrt{\frac{s_1^2}{n_1}+\frac{s_2^2}{n_2}}$$

(10)

The degrees of freedom are:

$$DF=\frac{{\left({VAR}_1+{VAR}_2\right)}^2}{\frac{{VAR}_1^2}{n_{1-1}}+\frac{{VAR}_2^2}{n_2-1}}$$

(11)

The pooled variance is used to estimate the common variance when the variances are equal:

$$s_p^2=\frac{\left(n_1-1\right)s_1^2+\left(n_2-1\right)s_2^2}{n_1+n_2-2}$$

(12)

The standard deviation of ${\overline{X}}_1-{\overline{X}}_2$ is estimated by:

$$s=s_p\sqrt{\frac{1}{n_1}+\frac{1}{n_2}}$$

(13)

The test statistic degrees of freedom are:

$$DF=n_1+n_2-2$$

(14)

where

${\overline{X}}_1$ = the first sample’s mean

${\overline{X}}_2$ = the second sample’s mean

$s$ = sample standard deviation of ${\overline{X}}_1-{\overline{X}}_2$

${\delta }_o$ = discrepancy between the two population means hypothesized

$s_1$ = the first sample’s standard deviation

$s_2$ = the second sample’s standard deviation

$n_1$ = sample size of the first sample

$n_2$ = sample size of the second sample

${VAR}_1$ = $\frac{s_1^2}{n_1}$

${VAR}_2$ = $\frac{s_2^2}{n_2}$

The research employed the usage of the p-value to examine the hypothesis. The p-value is a quantitative measure obtained from a statistical test that assesses the likelihood of receiving a particular set of observations if the null hypothesis is true. The smaller the p-value, the more likely we will reject the null hypothesis. Examples of hypotheses in this research include:

Null hypothesis (H0)—There is no effect or difference between in-wheel motors with wafters and in-wheel motors without wafters.

Alternative hypothesis (HA or H1)—Some effect or difference exists between in-wheel motors with and without wafters.

Given the experimental data, a p-value indicates how believable the null hypothesis is. Specifically, assuming the null hypothesis is true, the p-value tells us the probability of obtaining an effect at least as significant as the one we observed in the experimental data. If the p-value of a hypothesis test is sufficiently low, we can reject the null hypothesis. Specifically, when we conduct a hypothesis test, we must choose a significance level (alpha) at the outset. Common choices for significance levels are 0.01 and 0.05. We can reject the null hypothesis if the p-value exceeds our significance level. Otherwise, if the p-value equals or exceeds our significance level, we fail to reject the null hypothesis.

3. Results and Discussion

Figure 7 depicts the temperature distribution in the electric motor housing per unit of time due to experimental tests on in-wheel motors. Temperature measurements were carried out on the in-wheel motor in default conditions (without housing modification) and conditions after modification (adding fan blades or wafters to the inner surface of the housing). In-wheel motors have two housings named front housing and rear housing in this study. The rear housing is marked with a power cable and a hall sensor cable to differentiate the front and rear housing, as seen in Figure 4. The average temperature of the rear housing is relatively higher than the front housing. The average temperature of the front housing of the electric motor in default condition is 34.92 °C, and the average temperature of the rear housing in default condition is 35.19 °C. The average temperature of the front housing of the electric motor in the modification condition is 33.42 °C, and the average temperature of the rear housing in the modification condition is 33.58 °C. The temperature in the rear housing of the in-wheel motor is higher than the front housing temperature because there are power and hall sensor cables. These electrical components slightly increase the temperature of the rear housing of the in-wheel motor, as the thermal image shows in Figure 8. Even though the average temperature in the rear housing is slightly higher than the front housing temperature if tested statistically using a two-sample t-test, the p-value between default (front housing) and default (rear housing) is 0.764. The p-value between the modified front and rear housing is 0.838. These two p-values are higher than the significance level of 0.05, meaning there is no significant difference between the front and rear housing temperatures. So, for the subsequent analysis, the front and rear housing temperatures are assumed to be the same.

Figure 9 depicts the temperature in the electric motor housing with fan blades (wafters) and the temperature without wafters from experimental test results. There is no load on the motor during testing except from the electric motor components. The in-wheel motor speed is 485 rpm from the start to the end of the test. Testing time is 11,400 s or 190 min. Installing fan blades (wafters) on the in-wheel motor housing lowered the temperature in the motor housing, as shown in Figure 9. The average temperature of the in-wheel motor housing without wafters is 34.92 $°\mathrm{C}$, whereas the average temperature after wafters is 33.42 $°\mathrm{C}$. Using wafters on the housing reduced the temperature by 1.493 $°\mathrm{C}$. Using a two-sample t-statistical test on the electric motor housing temperature without wafters and with wafters obtained a p-value of 0.082. The p-value is higher than the significance level of 0.05, meaning that the temperature difference in the in-wheel motor housing without and using wafters is insignificant. The less significant reduction in temperature on the in-wheel motor after using wafters is because the design of the wafters used is not optimal. More research is needed to determine the best wafter design.

Figure 10 compares the temperature of the electric motor housing from experimental and simulation tests for the default housing. Based on the two-sample t-test, the p-value was obtained at 0.404, which means it is much higher than the significance level of 0.05, so the temperature data from simulation and experimental testing are similar. The average temperature of the default in-wheel motor when tested experimentally is 34.92 °C, and the average temperature of the default in-wheel motor when tested by simulation is 35.63 °C, so if we refer to the mean, then there is an error between experimental testing and simulation for default in-wheel motor of 1.99%.

Figure 11 compares the in-wheel motor housing temperature from experimental and simulation tests for the modification housing. Based on the two-sample t-test, the p-value was obtained at 0.927, which means it is much higher than the significance level of 0.05, so the temperature data from simulation and experimental testing are similar.

Figure 12 depicts the thermal simulation findings for major in-wheel motor components, including the winding, rotor, magnet, and housing. Figure 12 also compares the thermal characteristics of the motor after installing fan blades (wafters) in the housing with the thermal characteristics of the motor in default conditions. The temperature in the winding is substantially higher than in the rotor, magnet, and electric motor housing. The use of fan blades (wafters) has reduced the temperature at the winding. Based on the two-sample t-test on winding temperature, the p-value difference between an electric motor using wafters and an electric motor without wafters is 0.000. This p-value is lower than the significance level of 0.05, meaning there is a significant difference between the winding temperature of the electric motor with wafters and that of the electric motor without wafters. So, the decrease in winding temperature due to using wafters on the housing is significant. Using wafters in the housing reduced the average temperature in the winding by 3.047 °C, or a reduction of 4.34%. The stable state temperature on the default in-wheel motor is reached after the motor has operated for 56,280 s or 938 min (15.6 h), while on the modified in-wheel motor, the stable state temperature is reached after the motor has operated for 55,720 s or 929 s (15.5 h). Using wafters in the in-wheel motor housing can speed up the stable state temperature of the electric motor by 9 min compared to in-wheel motors without wafters.

Figure 13 displays the temperature of the permanent magnet in-wheel motor as determined using thermal modeling. An electric motor’s wafters decrease the permanent magnet’s temperature compared to an electric motor without wafters. Based on the two-sample t-test on permanent magnet temperature, the p-value between the permanent magnet temperature of the modified electric motor and the permanent magnet temperature of the default electric motor is 0.002. The p-value is below the significance level of 0.05, indicating a substantial disparity in magnet temperature between the electric motor with and without wafters. The utilization of wafters in this investigation results in a decrease of 0.93 °C in the mean temperature of the permanent magnet in-wheel motor compared to the temperature of the permanent magnet in-wheel motor without wafters. Using wafters in the in-wheel motor housing results in a 2.3% decrease in temperature on the permanent magnet.

The decrease in permanent magnet temperature of 2.3% from using wafters in this study is smaller than the decrease in permanent magnet temperature from using wafters, as observed by Kang et al. [27]. Kang et al. [27] apply wafters to the interior of a permanent magnet synchronous motor (IPMSM) with an internal rotor type where the tilt angle of the fan blades (wafters) is varied and equipped with vent holes. This results in a decrease in temperature on the permanent magnet by 7%. The reduction in winding temperature of 3.047 °C achieved by implementing wafters in this study is somewhat smaller than the reduction in winding temperature achieved by Satrustegui et al. [28] through wafters. Satrustegui et al. [28] utilize a wafter to modify electric motors of the internal rotor type by installing the wafter on the rotor. The specific electric motor utilized is an induction motor and can reduce the temperature of the winding by 16.2 °C. The relatively minor decrease in temperature seen in this study compared to previous research provides an incentive to enhance the performance of the wafters by improving their design, notwithstanding the utilization of a BLDC electric motor with an outer rotor configuration.

4. Conclusions

This study proposes a novel cooling solution for an electric motor by mounting the fan blade on the inner housing of the in-wheel motor (an electric motor with the rotor on the outside and the stator on the inside). Fan blades (wafters) are made from polylactic acid (PLA) and made using 3D printer technology. Wafters are placed on a 1500 W in-wheel motor with 48 poles and 52 slots. To assess the effect of wafters on the thermal parameters of an electric motor, the researchers used experimental approaches and thermal simulation. The motor rotates at 500 rpm with no load condition during the experimental test. The temperature of the electric motor is measured using an infrared thermal imager. Thermal simulation uses the lumped circuit model method and uses Ansys Motor-CAD software. Applying wafters on the inner surface of the in-wheel motor housing significantly reduces the winding temperature by 3.047 °C, a temperature reduction of 4.34%. Using wafters in the in-wheel motor housing also speeds up the stable state temperature of the electric motor by 9 min compared to in-wheel motors without wafters. Further research is needed to improve the performance of the fan blades (wafters) in the in-wheel motor housing in the form of design changes, using different materials, or increasing the surface roughness of the wafters. Electric motors also need to be tested under various load conditions and operational conditions to determine the ability of the wafters to overcome overheating.