Experimental Study of Kinetic to Thermal Energy Conversion with Fluid Agitation for a Wind-Powered Heat Generator

Abstract

In this paper, a heat generator with fluid agitation is developed and experimentally studied. This heat generator can convert kinetic energy from a wind turbine directly to thermal energy through the process of viscous dissipation—a process achieved through the agitation of the working fluid inside a container. In the experimental study, an electric motor (instead of a wind turbine) was used to provide the kinetic energy input to the heat generator. The torque, rotational speed, and temperature rise in the fluid were measured. Using the measured quantities, the efficiency of kinetic energy to sensible heat conversion was calculated. Experiments were conducted to investigate the effects of different impellers, rotational speeds, and working fluids, including distilled water, ethylene glycol (EG), and their respective nanofluids, with ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ nanoparticles at different concentrations. The study also found that the temperature rise in fluids due to viscous dissipation was influenced by the specific heat of the fluid, suggesting that the heat generator can be optimized for energy storage with high-specific-heat fluids, such as water, or for achieving a higher temperature rise with low-specific-heat fluids, such as ethylene glycol. The experimental results indicated that the heat generator was up to 90% efficient in converting kinetic energy to thermal energy. The study revealed that, for constant power input, the heat dissipation rate depends solely on the vessel’s geometry, not the fluid properties. Optimizing the impeller design and baffles within the vessel is crucial for maximizing power input. For applications, a wind turbine can power this heat generator to provide heat to a house or a commercial building.

1. Introduction

The world energy consumption is growing at the highest rate in history, and with the recent shunning of non-renewable energy resources, the bulk of this new energy demand needs to be met by renewable energy resources [1]. Wind and solar energy are the most popular renewable energy sources. There has been significant investment and research in recent decades. Modern wind turbines are generally used for electricity generation. However, the final form of energy required by the user in many cases is thermal energy [2]. For example, an average household in Canada uses 80% of their total energy consumption for heating purposes [3]. Although electrical energy conversion to thermal energy is a high-efficiency process, electricity generation from wind turbines is usually low, pushing down the overall efficiency of heating with wind. Directly producing thermal energy from the wind is more cost-effective than the wind–electricity–heat approach. The direct conversion of kinetic-to-thermal energy eliminates the complex units of electrical generators, power converters, etc., leading to low cost and improved reliability [4]. Such a technology also enhances the economic viability of small wind turbines that may be inefficient in producing electricity [5].

There are several ways to design a heat generator that converts the kinetic energy from a wind turbine directly to thermal energy, such as solid or liquid friction [6], air compression [7], and induction through an alternating magnetic field [8]. This paper proposes a novel heat generator that converts the kinetic energy from wind directly into thermal energy through the agitation of a working fluid in a container. This process is known as viscous dissipation. The concept of heat generation using viscous dissipation is nothing new—James Prescott Joule used the same process to prove the conversion of mechanical work into heat, leading to the development of the first law of thermodynamics [9]. This wind-powered heat generator can be deployed in any cold region with abundant wind energy. For Canada, particularly the northern communities, coastal regions, and remote islands can be primary customers. It was estimated that 60 GW of power can be harnessed from onshore wind energy in Newfoundland alone [10]. The innovative wind-powered thermal energy technology proposed in this study would benefit economic prosperity, environmental sustainability, and social well-being for many similarly cold regions in the world.

Several researchers have attempted to theoretically model the wind-powered heat generator system under idealized conditions [11,12,13]. Unfortunately, these simulation studies are not backed by any experimental results. A few papers only focused on optimizing the torque–speed characteristics of the wind turbine and did not address the mechanism behind heat generation [11,12]. Another paper discussed harnessing the heat from viscous dissipation to power a maritime distiller [13]. However, this study only looked at the problem theoretically and provided no experimental data. The closest attempt at the wind to thermal conversion in the literature is an experimental study by researchers at Northwest A&F University in Shaanxi, China [14]. This study investigated the rise in temperature and the efficiency of such a system for three different base fluids. While this study concluded the viability of such a concept, the kinetic-to-thermal energy conversion efficiency was low and not what theoretical analysis would conclude. This discrepancy could be attributed to the complex method of power transmission employed in this study. Their design relied on a system with many belts and couplings, which are known to introduce significant frictional losses. In contrast, our study utilizes a simpler design with fewer moving parts, thereby minimizing frictional losses and potentially achieving a higher conversion efficiency. Another experiment conducted by chemical engineering researchers at IIT-Delhi investigated impeller design characteristics using temperature rise data from their experiments [15]. In that experimental setup, heat was just a byproduct of the stirring action. The results from this study corroborate well with the theoretical analysis results. A conference paper in 2013 provided valuable experimental data on the temperature rise due to mechanical work [16]. This paper proposed a “cavitation heat pump”, which could heat the liquid for use at home. The authors assumed that the cavitation process produced heat but missed the viscous dissipation phenomena.

More recent studies on wind-powered thermal generators have investigated the wind-turbine start-up issue [17], rotor design [18], and different potential applications [19,20]. Thermodynamic analysis in [17] indicated a potential temperature increase of 5 °C in a 1 L fluid within an hour. In another study [18], a wind-energy-stirring heater was designed and experimentally tested. CFD simulations were used to analyze the impact of the rotor design and fluid properties on heating performance, revealing that increased rotor layers, higher liquid levels, and fluids with high density, low viscosity, and low-specific-heat capacity are desired for temperature increases of the working fluids. A wind-powered stirring heater using a paraffin oil and water mixture was developed for winter heating in northeast China [19]. This system bypasses electricity generation, offering higher efficiency and reduced emissions compared to traditional coal- or straw-burning methods. Another interesting application, proposed by Chen et al. [20], involves a system that directly converts offshore wind energy to heat, stores it thermochemically, and transports it to shore for on-demand use.

Previous studies have tried different working fluids in the thermal generator, including nanofluids. A nanofluid is a fluid that contains colloidally suspended solid particles of size between 1 nm and 100 nm in a base fluid. Ever since Choi’s innovative work on nanofluids in 1995 [21], the interest of the research community has been kindled into the amazing thermal fluid property of nanofluids. A nanofluid is observed to have better heat transfer properties compared to the base fluid [22]. Because of the small size of nanoparticles, the nanofluids act as a single-phase fluid instead of a two-phase mixture [23]. Essential requirements for nanofluids include durability, stability, and even suspension of nanoparticles in the fluid. There are multiple examples of nonmetallic nanoparticles and metallic nanoparticles that could be used to prepare nanofluids [23]. Some popular base fluids are water, ethylene glycol (EG), transformer oil, and mineral oil.

Despite the advantages and novelty of the wind-powered heat generator with fluid agitation, the kinetic-to-thermal energy conversion mechanism is still not well understood. The effectiveness and efficiency of this energy conversion depend on many factors, including the heat generator’s geometrical parameters and the working fluid’s thermal fluidic properties. The effects of these properties have not been systematically investigated. In this paper, a heat generator powered by fluid agitation is developed and experimentally studied to understand the impact of these properties. This heat generator converts kinetic energy (e.g., that from a wind turbine) directly to thermal energy through the agitation of the working fluid inside a steel container. This heat generator uses an optimized flat-blade turbine (FBT) impeller and a fully baffled configuration. An electric motor is used to provide the kinetic energy input to the heat generator. A torque sensor, a tachometer, and thermocouples are used to measure the torque, rotational speed (RPM), and temperature rise in the fluid. The efficiency of kinetic energy to sensible heat conversion is calculated using the measured data. Different working fluids are investigated, including distilled water, ethylene glycol (EG), and their respective nanofluids, with ${\mathrm{A}\mathrm{l}}_2{\mathbf{O}}_3$ nanoparticles at different concentrations. ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ nanoparticles are selected for the study due to the vast availability of literature on their properties. The results from these experiments provide essential insight into the energy conversion process and valuable guidance for the design of a heat generator with wind-powered fluid agitation.

2. Design of The Heat Generator and Experimental Setup

2.1. Theoretical Analysis

Heat will be generated once a fluid is agitated [24], and this heat will be stored by the fluid if there are negligible losses to the environment. The literature also indicates that all the mechanical power that goes into the system will be converted into heat, as no energy is stored in the energy cascade process [25]. In this study, an electric motor (instead of a wind turbine) was used to provide kinetic energy input to the heat generator. If the power is constant during the whole experiment, it can be multiplied by time to obtain the total kinetic energy input. Ideally, this kinetic energy input and the total heat stored should be equal if there is no heat loss to the environment. The design of the heat generator aims to maximize the heat generated by viscous dissipation.

Flow turbulence inside the vessel comes from the stirring of the liquid by an impeller driven by the electric motor. Some commercially available impellers are propellers, turbines, paddles, and high-shear impellers [26]. These impellers can be divided into two major groups depending on the flow regime they create—either axial or radial. The radial-flow-type impellers apply greater shear stress to the fluid than their axial-flow-type counterparts. The flow pattern for each type is visualized in Figure 1.

The impeller power consumption is a function of rotational speed, external forces, such as gravitational force, fluid properties, such as density and viscosity, along with impeller and vessel geometry. Dimensional analysis was used to obtain twelve dimensionless groups that can correlate the power consumption in a vessel. The dimensionless groups obtained by the Buckingham PI theory are noted below:

$$\mathrm{f}\left({\displaystyle \frac{{\mathrm{D}}^2\mathsf{\omega }\mathsf{\rho }}{\mathsf{\mu }}},{\displaystyle \frac{{\mathrm{D}\mathsf{\omega }}^2}{\mathrm{g}}},{\displaystyle \frac{\mathrm{P}}{\mathsf{\rho }{\mathsf{\omega }}^3{\mathrm{D}}^5}},{\displaystyle \frac{\mathrm{D}}{{\mathrm{D}}_{\mathrm{T}}}},{\displaystyle \frac{\mathrm{D}}{\mathrm{Z}}},{\displaystyle \frac{\mathrm{D}}{\mathrm{C}}},{\displaystyle \frac{\mathrm{D}}{\mathrm{p}}},{\displaystyle \frac{\mathrm{D}}{\mathrm{w}}},{\displaystyle \frac{\mathrm{D}}{\mathrm{l}}},{\displaystyle \frac{{\mathrm{n}}_2}{{\mathrm{n}}_1}},{\displaystyle \frac{\mathrm{D}}{\mathrm{B}}},{\displaystyle \frac{\mathrm{D}}{{\mathrm{l}}_{\mathrm{b}}}}\right)=0$$

(1)

This equation is by no means universal—it is valid only for a cylindrical vessel with an impeller placed at the center. For multiple impellers or a different tank shape, more dimensionless groups need to be introduced. The first three terms in the equation are the Reynolds number, Froude number, and Power number. The Froude number (Fr) quantifies the relative importance of inertial forces and gravitational forces acting on a fluid. The Froude number essentially compares the fluid’s momentum to the gravitational force pulling it down. A high Fr indicates that inertia dominates, leading to faster, supercritical flow. Conversely, a low Fr signifies gravity’s influence, resulting in slower, subcritical flow. The Power number relates the mechanical power input to an impeller (device transferring energy to the fluid) and the properties of the fluid being mixed. The Power number reflects the efficiency of the impeller in transferring energy to the fluid based on its density and the impeller’s geometry and rotational speed. A higher Power number indicates that a larger power input is required to achieve the desired mixing for a specific fluid–impeller combination. Conversely, a low Po signifies a more efficient energy transfer.

The equality of these groups ensures dynamic and kinematic similarity between two geometrically similar vessels. The last nine groups are responsible for the geometric similarity. The following relation summarizes the Reynolds, Froude, and Power numbers’ relationship for a geometrically similar vessel:

$${\mathrm{N}}_{\mathrm{p}}\propto {\left(\mathrm{R}\mathrm{e}\right)}^{\mathrm{a}}{\left(\mathrm{F}\mathrm{r}\right)}^{\mathrm{b}}$$

(2)

The Froude number’s effect is almost negligible for a fully baffled system [14]. This claim can easily be verified from the fact that no vortexes are formed in a baffled system when compared to an unbaffled system. Because the power dissipation of a baffled system is higher than that of an unbaffled system, the Froude number is disregarded for the remainder of the investigation. It is interesting to note that the Power number for a baffled system becomes constant at a high Reynolds number, very similar to the coefficient of drag for a sphere when plotted against the Reynolds number [27]. Figure 2 shows a plot between the Power number and Reynolds number, which is a popular choice by many researchers to present their findings. While it is easy to understand and interpret, this graph has led to a widespread misconception that only the Reynolds number is needed to conclude the power input needed by a system. It is important to note that this dynamic similarity is only valid for a particular geometric configuration, and that any change in any of the geometric parameters will yield a different result.

Bates et al. [28] conducted a comprehensive comparison among many different radial-type impellers. They observed a decrease in the Power number with an increase in the Reynolds number. As the flow became fully turbulent, the Power number became constant [29]. For a given impeller type, as the number of blades, width of a blade, or length of blades increased, so did the Power number. Different types of impellers have different Power numbers, with the Rushton turbine having the highest Power number and the flat-blade turbine (FBT) coming a close second, with only a negligible difference. Any curvature or pitch change in the FBT blade decreases that turbine’s maximum Power number compared to a generic FBT. Due to this negligible difference in Power number and the ease of manufacturing, the flat-blade turbine (FBT) was chosen for our study.

The impeller’s position at varying depths of the liquid inside the vessel has a negligible effect on the Power number for any value above C/D = 1. Bates et al. noted that clearance, on the other hand, impacts the Power number. They noted that the difference in performance varied between the different types of impellers, where the Power number increased with increasing clearance for the disk-type impeller but decreased for the pitched-type impeller. For the flat type of impeller, the clearance of C/D = 1 is enough to reach a reasonably high Power number.

In a vessel without baffles, there exist two types of flow regions: for the region around the impeller, the liquid rotates as a whole, and for the region from there to the vessel wall, a free vortex exists. Aiba [30] found that this phenomenon is independent of the type of impeller. He also noted that the flow pattern is independent of the impeller speed. This is because, in an unbaffled vessel, the tangential velocity component is high compared to other components.

For a fully baffled vessel, the tangential velocity decreases while the radial velocity remains unchanged [31]. A fully baffled configuration for a given vessel is when the relative Power number of baffled vessels is highest compared to the unbaffled vessel’s Power number. The introduction of baffles also increases the Power number of a given impeller [28]. Hence, for this experimental study, a fully baffled vessel was used. It was found that the baffle ratio of 0.4 is optimum for any turbine diameter, and that any increase in the length of the baffle [28] after that point will decrease the Power number. Nagata et al. [32] observed the same trend, but they proposed 0.5 as the optimum baffle ratio. The baffle ratio is defined as:

$$\mathrm{B}\mathrm{R}={\displaystyle \frac{{\mathrm{n}}_{\mathrm{b}}{\mathrm{w}}_{\mathrm{b}}}{{\mathrm{D}}_{\mathrm{T}}}}$$

(3)

where $n_b$ is the number of baffles and ${\displaystyle \frac{{\mathrm{w}}_{\mathrm{b}}}{{\mathrm{D}}_{\mathrm{T}}}}$ is the ratio of baffle length to the tank diameter. If four baffles are used, where each baffle’s length is ten percent of the tank’s diameter, this configuration is considered fully baffled.

2.2. Design and Fabrication

The goal of this study was to design a heat generator that would allow high power input. A cylindrical configuration was chosen because of the ease of manufacturing and the fact that most experimental studies and handbook data are available for cylindrical vessels. The length-to-height ratio was kept at 1:1. Four baffles, each ${\displaystyle \frac{1}{10}}\mathrm{t}\mathrm{h}$ of the vessel diameter, were selected to form a fully baffled configuration. The impeller was a flat-blade type, and its clearance from the bottom of the vessel was ${\displaystyle \frac{1}{3}}$ of the vessel height. All these criteria were selected following the theories discussed in Section 2.1 to maximize the power input to the heat generator. The heat generator was designed for a volume of five liters, and this volume guides the dimensions of the vessel. To investigate the effect of the change in the impeller diameter, two different impellers were used in the experimentation, as shown in Figure 3.

2.3. Experimental Setup and Instrumentation

The experimental setup is shown in Figure 4. The heat generator has two main parts: the bottom part is the heat generator vessel with baffles and the impeller in the working fluid, and the upper part includes the motor and torque sensor housing. The two parts were joined together using mechanical fasteners and clamps. The motor and torque sensor were joined together using Lovejoy Jaw coupling, as these couplings are good at tolerating slight misalignment and, hence, reduce vibration. A standard solid coupling was used to attach the torque sensor to the impeller shaft. The vessel was insulated using Reflectix ^® (Markleville, IN, USA) double-reflective insulation [33].

Four thermocouples (T type, Omega Canada, St-Eustache, QC, Canada) were used to measure the temperature of the liquids in the heat generator. The thermocouples were inserted at different depths (2, 3, 5, and 7 inches) into the liquids through holes on the upper cover of the vessel, all approximately 2 inches from the center, with equal distance between and 90 degrees apart from each other. Each thermocouple probe had a stainless-steel sheath material that could protect it from the working fluids. The averaged temperatures from readings of the four thermocouples were used in reporting the fluid temperature, though it was found that there were no noticeable differences between the four.

While the specific heats for water, ethylene glycol, and their mixture are readily available in the literature, there is no credible resource for the specific heat of the used nanofluids. In this study, differential scanning calorimetry (DSC) was conducted to determine the specific heat of the nanofluids, where a Mettler-Toledo DSC1 was used.

2.4. Uncertainty Analysis

Measurement uncertainties of the major independent variables in this study are listed in

Table 1. Measurement uncertainties.

Parameter	Uncertainty
Temperature, T	$\pm$0.5 $℃$
Specific heat, $c_p$	$\pm$0.17 $\mathrm{J}/\mathrm{k}\mathrm{g}\bullet \mathrm{K}$
Torque, $\tau$	$\pm$0.4 lb-in
Mass,$m$	$\pm$11 g
RPM	$\pm 5.5$ RPM

. The total uncertainty for each variable included both bias and precision errors:

$$\mathrm{U}=\sqrt{{\mathrm{P}}^2+{\mathrm{B}}^2}$$

(4)

where $P$ is the precision limit and $B$ is the bias limit.

The uncertainty of a dependent variable can be calculated from the uncertainties of independent variables using the Kline and McClintock method [34]. The major dependent variable in this study was the efficiency of energy conversion, $\eta$, from the input kinetic energy to thermal energy (sensible heat) in the fluid, as defined in Equation (5):

$$\mathsf{\eta }={\displaystyle \frac{\mathrm{m}{\mathrm{c}}_{\mathrm{p}}\left({\mathrm{T}}_{\mathrm{f}}{-\mathrm{T}}_{\mathrm{i}}\right)}{\mathsf{\tau }·{\mathsf{\omega }}_{\mathrm{c}\mathrm{y}\mathrm{c}}}}$$

(5)

The total uncertainty of this dependent variable, ${\mathrm{U}}_{\mathsf{\eta }}$, can be calculated by Equations (6) and (7). Here, the temperature difference, $\mathsf{\Delta }\mathrm{T}={\mathrm{T}}_{\mathrm{f}}{-\mathrm{T}}_{\mathrm{i}}$, is used to simplify the process:

$${\left({\displaystyle \frac{{\mathrm{P}}_{\mathsf{\eta }}}{\mathsf{\eta }}}\right)}^2={\left({\displaystyle \frac{{\mathrm{P}}_{\mathrm{m}}}{\mathrm{m}}}\right)}^2+{\left({\displaystyle \frac{{\mathrm{P}}_{{\mathrm{c}}_{\mathrm{p}}}}{{\mathrm{c}}_{\mathrm{p}}}}\right)}^2+{\left({\displaystyle \frac{{\mathrm{P}}_{{\mathrm{T}}_{\mathrm{f}}}}{\mathsf{\Delta }\mathrm{T}}}\right)}^2+{\left({\displaystyle \frac{{\mathrm{P}}_{{\mathrm{T}}_{\mathrm{i}}}}{\mathsf{\Delta }\mathrm{T}}}\right)}^2+{\left({\displaystyle \frac{{\mathrm{P}}_{\mathsf{\tau }}}{\mathsf{\tau }}}\right)}^2+{\left({\displaystyle \frac{{\mathrm{P}}_{{\mathsf{\omega }}_{\mathrm{c}\mathrm{y}\mathrm{c}}}}{{\mathsf{\omega }}_{\mathrm{c}\mathrm{y}\mathrm{c}}}}\right)}^2$$

(6)

$${\left({\displaystyle \frac{{\mathrm{B}}_{\mathsf{\eta }}}{\mathsf{\eta }}}\right)}^2={\left({\displaystyle \frac{{\mathrm{B}}_{\mathrm{m}}}{\mathrm{m}}}\right)}^2+{\left({\displaystyle \frac{{\mathrm{B}}_{{\mathrm{c}}_{\mathrm{p}}}}{{\mathrm{c}}_{\mathrm{p}}}}\right)}^2+{\left({\displaystyle \frac{{\mathrm{B}}_{{\mathrm{T}}_{\mathrm{f}}}}{\mathsf{\Delta }\mathrm{T}}}\right)}^2+{\left({\displaystyle \frac{{\mathrm{B}}_{{\mathrm{T}}_{\mathrm{i}}}}{\mathsf{\Delta }\mathrm{T}}}\right)}^2+{\left({\displaystyle \frac{{\mathrm{B}}_{\mathsf{\tau }}}{\mathsf{\tau }}}\right)}^2+{\left({\displaystyle \frac{{\mathrm{B}}_{{\mathsf{\omega }}_{\mathrm{c}\mathrm{y}\mathrm{c}}}}{{\mathsf{\omega }}_{\mathrm{c}\mathrm{y}\mathrm{c}}}}\right)}^2-2{\left({\displaystyle \frac{{\mathrm{B}}_{{\mathrm{T}}_{\mathrm{f}}}^{\prime }}{\mathsf{\Delta }\mathrm{T}}}\right)}^2{\left({\displaystyle \frac{{\mathrm{B}}_{{\mathrm{T}}_{\mathrm{i}}}^{\prime }}{\mathsf{\Delta }\mathrm{T}}}\right)}^2$$

(7)

where ${\mathrm{B}}_{{\mathrm{T}}_{\mathrm{f}}}^{\prime }$ and ${\mathrm{B}}_{{\mathrm{T}}_{\mathrm{i}}}^{\prime }$ are portions of ${\mathrm{B}}_{{\mathrm{T}}_{\mathrm{f}}}$ and ${\mathrm{B}}_{{\mathrm{T}}_{\mathrm{i}}}$, and because they arise from identical error sources, we can assume them to be perfectly correlated [32]. This simplifies the equation to:

$${\left({\displaystyle \frac{B_{\eta }}{\eta }}\right)}^2={\left({\displaystyle \frac{B_m}{m}}\right)}^2+{\left({\displaystyle \frac{B_{c_p}}{c_p}}\right)}^2+{\left({\displaystyle \frac{B_{\tau }}{\tau }}\right)}^2+{\left({\displaystyle \frac{B_{{\omega }_{cyc}}}{{\omega }_{cyc}}}\right)}^2$$

(8)

The uncertainty associated with the efficiency was up to 42% for experiments with lower rotational speeds and higher-specific-heat fluids, which caused a smaller temperature rise. The uncertainty was much lower for experiments with a higher temperature rise and lower-specific-heat fluids.

3. Results and Discussion

3.1. DSC Measurements of Specific Heat of the Fluids

The specific heat values along with temperature rise values were used to calculate the amount of heat generated due to viscous dissipation. The DSC results for water-based nanofluids are shown in Figure 5. To verify the accuracy of the DSC equipment, an experiment was run on water. The specific heat values from the experiment were in close agreement with values from the literature, testifying to the DSC’s reliability.

For the nanofluids, the specific heat for the 0.50% ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$–water nanofluid was higher than water, and the specific heat value decreased with the addition of more nanoparticles. The lowest specific heat was for the 1.00% ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$–water nanofluid. The mechanism behind the increase and then the decrease in specific heat with nanoparticle addition is widely reported in other studies [36,37], but it is not understood properly. Some researchers suggest that the nanolayer’s presence surrounding the nanoparticles is responsible for this increase in heat capacity at lower-volume concentrations, but this theory is not widely accepted [37].

The experimental results indicated that the specific heat decreased with the increase in temperature. While this trend was expected, as evident from the literature, the effect was more pronounced in this study. Mass loss might be responsible for this exaggerated trend, but overall, the difference was within the margin of error expected from the equipment.

The DSC results for the EG-based nanofluids are presented in Figure 6. The DSC values for EG were not in good agreement with values from the literature. This is because the EG used in the experiments was not one hundred percent pure but contained trace elements of impurities, while the specific heat values for EG found in the literature are for pure EG itself. One valuable takeaway from the comparison between the two is that their rate of change with increasing temperature was nearly identical. This trend was expected, as there was no mass loss during EG experimentation on DSC equipment.

For the nanofluids, the specific heat for the 0.50% ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$–EG nanofluid was higher than EG, but the specific heat value decreased with the addition of more nanoparticles. The lowest specific heat in this series was for the 1.00% ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$–EG nanofluid. This trend agrees with that in the water-based nanofluid data. From the experimental results, the specific heat for the nanofluids slightly increased with the increase in temperature.

3.2. Rate of Temperature Rise

3.2.1. Water-Based ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ Nanofluid

The first series of experiments were conducted with water as a base fluid and with varying concentrations of ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ nanoparticles. Eight experiments were conducted with the water-based nanofluids. Temperature rise data and torque values for four different concentrations (0.00%, 0.50%, 0.75%, and 1.00%) of nanofluids at two different rotational speeds (750 RPM and 1000 RPM) were recorded. Figure 7 and Figure 8 show the temperature increases of liquids inside the vessel within 60 min of fluid agitation. The initial temperatures of the fluids were approximately the same as the room temperature, within $\pm 5 \mathrm{℃}$ variations for all the experiments to ensure the room temperature did not have a significant impact on the results. It can be noted that the temperature increases were almost linear. The final value of temperature rise at the end of the experiment was not statistically significant to indicate that the nanoparticle concentration had any effect on the temperature rise. For the 750 RPM rotational speed experiments, the temperature rise values at the end of the experiments for all the nanofluid concentrations were within $\pm 0.2 \mathrm{℃}$. The average increase in temperature was noted to be 2.1 $\mathrm{℃}$. In the 1000 RPM rotational speed experiments, the temperature rise values at the end of the experiments for all the nanofluid concentrations were within $\pm 0.5 \mathrm{℃}$. The experiment with 0.75% ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ had a slightly lower temperature rise than the other experiments, but the average total temperature rise was around 4.3 $\mathrm{℃}$.

3.2.2. Ethylene Glycol (EG)-Based ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ Nanofluid

Eight experiments were conducted with ethylene glycol (EG) as the base fluid, with varying nanoparticle concentrations (0.00%, 0.50%, 0.75%, and 1.00%) and two different rotational speeds (750 RPM and 1000 RPM). Figure 9 and Figure 10 show the temperature increases of liquids inside the vessel within 60 min of fluid agitation. The temperature increases were almost linear, and the final value of temperature rise at the end of the experiment was affected only slightly by the nanoparticle concentration. For the experiments with the 750 RPM rotational speed, there was an almost identical temperature rise trend for pure EG and the 0.5% ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$–EG nanofluid. For 0.75% and 1.00% concentrations, the final temperature rise decreased slightly. In the experiments with the 1000 RPM rotational speed, the temperature rise increased slightly for the 0.50% ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$–EG nanofluid compared to pure EG. However, with the increasing nanoparticle concentration, the temperature rise started to decrease slightly.

The linear temperature rise was expected because the power input to the vessel was constant throughout the experiment. For a constant mass and nearly constant specific heat, a linear rise in temperature for constant power input is expected. Although the specific heat of a fluid is a function of temperature and varied slightly with the change in the fluid temperature, this change was not significant enough to have any measurable effect. Comparing the results from Figure 7 and Figure 8 with Figure 9 and Figure 10, a clear correlation between rotational speed and temperature rise was evident: higher rotational speeds resulted in higher temperature rises. This is because higher speeds increased the agitation and viscous dissipation within the fluid, leading to more efficient conversion of kinetic energy to thermal energy.

The addition of nanoparticles did change the temperature rise trend, but the change was insignificant and did not make economic sense. For this reason, nanofluids are not suggested for the commercial development of heat generators due to the high cost of nanoparticles and insignificant benefit of temperature rise.

3.2.3. Effect of Base Fluid on Temperature Rise

The next series of experiments were conducted with different water and EG mixtures as base fluids as shown in Figure 11 and Figure 12. The rationale behind choosing multiple base fluids was to identify the temperature rise trend and total energy conversion efficiency with varying fluid properties, such as viscosity and specific heat.

The difference in temperature rise can be explained when the specific heat values from

Table 2. Specific heat values of base fluids.

Fluid	${\mathit{C}}_{\mathit{p}}$$(\mathbf{J}/\mathbf{k}\mathbf{g}\bullet \mathbf{K}$)
Water	4.18
EG/Water (20:80)	4.03
EG/Water (50:50)	3.64
EG/Water (80:20)	3.06
EG	2.45

are considered: the fluid with the lowest specific heat had the highest temperature rise, and vice versa. The only exception to this trend was pure water, which had the second-lowest temperature rise instead of the lowest temperature rise. This was due to the much higher efficiency of energy conversion for pure water than that for the 20% EG and 80% water mixture.

3.2.4. Effect of Impeller Diameter on Temperature Rise

Experiments were conducted to better understand the impact of the impeller diameter on the fluid temperature rise. The impeller used until this point (impeller 1) had a 5 cm diameter, and this diameter was increased to 7 cm for impeller 2, and experiments were run in water. The temperature rise data for 750 RPM and 1000 RPM are represented in Figure 13 and Figure 14, respectively. It can be observed that temperature rose for the impeller with a larger diameter increase for both rotational speeds. This trend is in line with the theoretical analysis in Section 2, as an increase in diameter increased the impeller Power number, and hence the system consumed a greater power. As the impeller diameter increased, the Power number also increased. The Power number is a dimensionless parameter that relates the impeller power input to the rotational speed and fluid properties. A higher Power number signifies greater energy transfer from the impeller to the fluid. With a larger diameter, the impeller sweeps a larger volume of fluid per rotation, requiring more power to maintain the same rotational speed. This additional energy translates to a greater increase in fluid temperature.

Finally, the temperature rise data after 60 min of fluid agitation are plotted in Figure 15 to summarize all the experiments discussed earlier. Temperature rise was the highest for the EG-based nanofluids, and it was lowest for the water-based nanofluids, with a mixture of base fluids in between. It is also clear that temperature rise was always higher for higher rotational speeds and larger impeller diameters.

3.3. Energy Conversion Efficiency of the Heat Generator

This section analyzes the conversion efficiency of kinetic energy to heat (or thermal energy) in the heat generator, as shown in Equation (5). The energy conversion efficiency of all the experiments is plotted in Figure 16. The first eleven experiments plotted are for impeller 1, while the last experiment is for impeller 2. The energy conversion process was highly efficient, where some experiments had an efficiency of more than 90%, and even the least efficient experiments had an efficiency of 70%.

It is also clear that the energy conversion efficiency was almost always higher for 750 RPM, i.e., the lower rotational speed conditions. There were only two experiments where the efficiency for 750 RPM and 1000 RPM was almost the same, but it was never higher for the 1000 RPM conditions. Another trend in the results was that the efficiency for water-based fluids was higher than that for EG-based fluids, and the efficiency of mixed base fluids was in between. It can also be noted that the efficiency of an impeller with a larger diameter was lower than that for a smaller-diameter impeller, even if other conditions were kept the same.

All these trends can be explained by looking at how the efficiency is defined in Equation (5). For this analysis, only the sensible heat was calculated using the temperature rise of the fluids. The other two thermal energy outputs were not included in this calculation, i.e., (a) sensible heat with temperature rise of the metal vessel, and (b) latent heat due to vaporization of the fluids, even with perfect thermal insulation of the vessel. These energy outputs led to the fact that the conversion of kinetic energy to sensible heat of the fluids was less than 100%.

Cavitation occurs when the local pressure of a liquid goes below its vapor pressure. This causes a phase change from liquid to vapor. This process requires a large amount of thermal energy (the latent heat of evaporation) from the surrounding liquid and causes the local temperature to decrease. This phenomenon is called the thermal effect of cavitation [42]. In this study, the impeller was operating at a high rotational speed, which may have caused the local pressure of the fluids to go below their vapor pressure, hence causing the evaporation. An amount of energy, $Q_e=m_{loss}h$, was lost due to the evaporation and loss of the vapor through the imperfect sealing of the vessel, where $h$ is the specific latent heat of vaporization. Cavitation may damage the impeller and evaporate the liquid. Hence, cavitation must be avoided for the best results. A possible solution would be to keep the rotational speed under an upper limit so that the local pressure of the liquid does not go below the vapor pressure.

With higher rotational speeds or larger impellers, a high degree of turbulence is created in the fluids, which is more likely to cause cavitation and energy loss, leading to lower efficiency of sensible heat conversion for the fluids. The vapor pressure of water at 25 $℃$ was 3.17 Kpa [43], while the vapor pressure of EG at 25 $℃$ was 0.012 KPa. The lower vapor pressure of EG makes it easier to result in cavitation, hence causing lower efficiency, as defined in Equation (5).

Overall, the experimental results pointed toward the fact that most of the kinetic energy was converted into thermal energy, and the conversion was in line with what was expected from the theoretical analysis.

3.4. Power Input from Torque and Rotational Speed

The torque and rotational speed values can be used to calculate the power input of the heat generator. During an experiment, the RPM value was kept constant at 750 RPM or 1000 RPM, and the torque value varied across different experiments. The torque values for all the experiments using the standard 5 cm-diameter impeller are plotted in Figure 17. The results confirmed that torque values were higher for higher rotational speeds. It is also interesting to note that for each rotational speed (750 RPM and 1000 RPM), the torque values, and consequently the power input, were almost constant (0.15 Nm and 0.25 Nm) across all eleven experiments. As a comparison, the torque values for the 7 cm-diameter impeller were 0.299 Nm for the 750 RPM experiment, and 0.360 Nm for the 1000 RPM experiment. It can be inferred from the data that power input depends on the geometric properties of the vessel and impeller but is independent of fluid properties. If one is interested in increasing the power input, the best way is to focus on the geometric optimization of the heat generator, not the fluid property optimization. Some of the ways of optimizing the power input for the heat generator include increasing the diameter of the impeller, using an impeller with a high Power number, and using a fully baffled configuration.

4. Conclusions and Recommendations

The conversion of kinetic energy directly into heat (thermal energy) is a fundamental phenomenon. This study is one of the first experimental investigations on this energy conversion. A heat generator with fluid agitation was designed, manufactured, and experimentally investigated. Throughout this research, the following conclusions and recommendations were identified for further development of this technology:

• The experiments showed an average temperature rise of 2.1 °C at 750 RPM and 4.3 °C at 1000 RPM for water-based nanofluids and an average temperature rise of 2.7 °C at 750 RPM and 5.0 °C at 1000 RPM for EG-based nanofluids.
• The theoretical conversion of kinetic energy into heat in a closed system is one hundred percent. The experimental results in this study demonstrated that more than 90% of the kinetic energy was converted into sensible heat in the absence of cavitation. Cavitation will cause evaporation of the liquid. The vapor escapes the system and will cause energy loss. Cavitation may also damage the impeller and, hence, it must be avoided.
• When a constant power was supplied by an electric motor, the power dissipation was independent of the fluid properties, and it was only dependent on the geometric parameters of the heat generator. The geometric parameters, such as impeller shape, size, and baffles of the vessel, need to be optimized to increase the power input.
• The nanofluids did not affect the power dissipation characteristics in the heat generator. While nanofluids would improve the heat transfer characteristics for user applications, their high cost does not make economic sense for a practical heat generator.
• The temperature rise in the fluids due to viscous dissipation depended on the specific heat of the fluid. The heat generator can be optimized for energy storage by using a high-specific-heat fluid, such as water, or it can be optimized for high-temperature rise by using a low-specific-heat fluid, such as EG.
• This heat generator provided an efficient and cost-effective alternative for heating in wind-rich, cold regions, reducing reliance on fossil fuels and lowering greenhouse gas emissions. Its simplicity and fewer moving parts imply lower maintenance costs and improved reliability.
• Improvements can be made in several ways. The cavitation issue can be addressed by either lowering the impeller speed or redesigning it. The device’s efficiency can be further amplified by integrating it as a pre-heater for existing water-heating systems. Additionally, Computational Fluid Dynamics (CFD) simulations can be used to optimize the design. Developing a self-starting wind turbine specifically for this system would create a fully autonomous and sustainable solution.