Introduction
Due to the environmental pollution caused by combustion of fossil fuels and since they are nonrenewable, nowadays clean fuels such as biodiesel are of paramount importance. Biofuels with organic sources are assumed as one of the primary sources of renewable fuels or alternatives to fossil fuels. Combustion of plant-based fuels can be considered as an indirect application of solar energy. Biofuels mainly emit less pollution than fossil fuels and can be easily produced from the remains of plant materials. Engines are employed as power sources in industry and agriculture where many engineers are intended to investigate them and reduce their vibration problems. Vibrations are examined from different standpoints in engineering sciences. Tractors, as the most important sources of mechanical power on farms, have posed numerous problems associated with occupational health and safety of users. Engine vibration affects vehicle vibration, and is important for fault detection of engine and also driving comfort. Engine parts are damaged by vibration amplitude, and noise has negative effects on engine life (Saridemir and Polat, 2023). Vibration is caused by the combustion process caused by the explosive reaction of the air-fuel mixture under high pressure and temperature inside the combustion chamber. The type of used fuel affects it (Lima Júnior, Magalhaes, Ferreira, and Pereira, 2020).
Two biofuels, including biodiesel and bioethanol, can be used in diesel engines in pure or mixed forms since their features are identical to petrol and diesel (Ferella, Mazziotti Di Celso, De Michelis, Stanisci, and Vegliò, 2010). These biofuels require a vibration analysis before changing the engine fuel, which is the focus of this research. A wide range of studies have been performed to investigate the origins of engine vibrations and methods to reduce them (Taghizadeh-Alisaraeiet al., 2016). However, there are still many areas of study, including the effects of fuel types on vibrations, which have remained uninvestigated. The vibrations and strokes generated by using a particular fuel type in diesel engines can lead to significant wear and tear on engine components, skyrocketing maintenance costs, increased fuel consumption, and heightened discomfort and safety concerns for users (Salokhe, Majumder, and Islam, 1995). Compared to spark engines, more noise and vibrations are produced by diesel engines, especially in large engines with high ignition pressures and high compression ratios, having adverse effects on the users’ ears and bodies (Selim, 2001). Researchers showed that acceleration could be boosted by increasing the speed of engines for all mixtures of diesel and bioethanol fuels (Taghizadeh-Alisaraei and Rezaei-Asl, 2016). Other studies have also been performed on using biodiesel fuels bearing various fruitful results. Rahimi, Ghobadian, Yusaf, Najafi, and Khatamifar (2009) used a mixture of diesel and biodiesel fuel from sunflower oil and ethanol extracted from potato waste. The results demonstrated that the maximum power and torque of engines were slightly reduced using this fuel composition, while the average specific fuel consumption increased (Rahimi, Ghobadian, Yusaf, Najafi, and Khatamifar, 2009). Rice bran oil was used to produce biodiesel, which, when mixed with diesel and ethanol, was utilized as a fuel for diesel engines. The results showed that the highest thermal efficiency was obtained for 15% ethanol in the biodiesel and bioethanol blended diesel fuel (Subbaiahet al., 2010). Using a mix of ethanol and diesel can enhance the thermal efficiency of engines and reduce the particulates and NOx pollution (Hansen, Zhang, and Lyne, 2005). The results of relevant studies show that the vibration acceleration depends on the coordinate directions, engine speed, and fuel type. That is to say, the most severe vibrations occur in vertical directions, and vibrations are intensified by increasing the speed due to the increased number of courses and strokes per unit of time. Furthermore, they are reduced using pure biodiesel compared with pure diesel (Patelet al., 2016). Reducing the engine speed lessens both the tensional vibration of the crankshaft and the vibration acceleration (Patelet al., 2016).
Erdiwansyahet al. (2022) used different fuel blended with methanol and gasoline (between 5% and 10%) and three speeds of motor (1000, 1200, and 1400 rpm) on a four-cylinder gasoline engine. The findings indicate that fuel blends of methanol and gasoline exhibit the highest vibration levels at engine speeds between 1200 and 1400 rpm, compared to those observed at 1000 to 1200 rpm (Erdiwansyahet al., 2022). Cahyono and colleagues investigated the characteristics of biodiesel fuel blends infused with citronella oil, subsequently testing them on a single-cylinder diesel engine. The results showed that the inclusion of citronella oil in the blends led to a reduction in motor vibration when compared to B30 fuel (Cahyono, Semin, and Putri, 2024). Velmurugan and colleagues reduced the vibration of a four-stroke diesel engine by using a new fuel blend with the addition of biodiesel, palm oil, and nano particles such as titanium oxide (Velmurugan, Aathif Akmal, Paramasivam, and Thanikaikarasan, 2020). Chaitidis and colleagues measured the vibration of diesel motor in different engine rpm (850, 1150, and 2000) and variety of biodiesel blends (B20, B40, B60, and B80). The results indicated that increasing the proportion of biodiesel in fuel blends resulted in greater vibration and noise in the diesel motor (Chaitidis, Marhavilas, and Kanakaris, 2022). Susilo, Listiyono, and Khambali (2022) investigated the effect of a diesel-essential oil mixture on a diesel engine, related to engine performance, noise, and vibration. The experiments were performed on a 402 CC Dongfeng diesel engine. The results show that the B10 mixture has the lowest vibration at 1300 rpm (975.7 Hz) and the highest at 1900 rpm (989.8 Hz) (Susilo, Listiyono, and Khambali, 2022). Lima Júnioret al. (2020) analyzed vibrations of engines with two different fuels, Brazilian commercial diesel (B8) and Biodiesel (B100). They also analyzed the fuel consumption, level of particulate matter emissions, and engine temperature. The results showed that B8 diesel fuel had higher energy content than B100 biodiesel fuel across all three analyzed speeds. Total vibration of B8 diesel was 4.5% to 21% higher than B100 biodiesel, according to the speed range (Lima Júnioret al., 2020). In a research, the impacts of blends of biodiesel from waste cooking oil and various metal-oxide based nanoparticles (D100, B10, B10TiO2, B10Al2O3, and B10SiO2) on the vibration, noise characteristics, emission, combustion, and performance of a single-cylinder diesel engine were investigated (Ağbulut, Karagöz, Sarıdemir, and Öztürk, 2020). Their results showed that the highest vibration and noise data were seen in B10. The addition of nanoparticles into B10 slightly reduced these measurements (Ağbulutet al., 2020).
The acceleration of the front part of the tractor engine is significantly boosted through increasing the speed of the engine from 1000 to 2200 rpm in three vertical, horizontal, and lateral directions. Increasing the engine speed from 1400 to 2200 rpm yields an increase in the vibration acceleration of tractor engines due to the larger number of combustion courses and piston strokes per unit of time (Heidary, Hassan-Beygi, Ghobadian, and Alisaraei, 2013). Investigation of noise and vibrations caused by the combustion process in internal-combustion engines has always been of interest to researchers. Vibrations have direct destructive effects on engine components and users. Compared to petrol engines, more noise and vibrations are produced by diesel engines, producing adverse effects on engines and users (Guzzomi, Hesterman, and Stone, 2007; Keskin, 2010). The effects of a mixture of petrol and ethanol were investigated on the vibrations of cylinder blocks and noise of spark-ignition engines. The results showed that increasing ethanol in the mixture led to more engine vibrations and noise, especially between 1500 and 2500 rpm (Keskin, 2010). The vibrations of a mixture of biodiesel and diesel fuels in Perkins six-cylinder compression-ignition engines were studied before and after repairing the engines. The results indicated that the fuel mixture had significant impacts on the degree of vibrations such that 20% and 40% biodiesel mixtures offered the lowest vibrations. However, 15% and 30% biodiesel mixtures had the most severe vibrations. In addition, the results showed that the vibrations dropped dramatically after repairs (Taghizadeh-Alisaraei, Ghobadian, Tavakoli-Hashjin, and Mohtasebi, 2012). In another study, the results demonstrated that the mean of vibration acceleration increased in all gears and positions by increasing the rotational speed of the engine. The vibration acceleration varied along the coordinate directions. In general, the vibration acceleration was highest in the vertical direction, followed by the lateral direction, and lowest in the longitudinal direction (Mirnezami, Hassan-Beygi, Banakar, and Ghobadian, 2017). Engine speed is the most crucial factor affecting engine power. Furthermore, the quantity of biodiesel fuel and engine speed are the most important indices of the amount of specific fuel consumption (SFC) (Buragohain and Mahanta, 2008; Mirnezamiet al., 2017; Xu, Yin, Liu, and Jia, 2017).
The internal combustion engine is an important source of noise and vibration in a tractor. Given that various researches have been conducted on engine vibrations, the majority of tests on engines have been performed using fossil fuels. However, no comprehensive research has been performed on the utilization of different combinations of biodiesel, bioethanol, and diesel fuels, and their effects on the vibrations (acceleration) of engines. Hence, in the present study, the vibrations of the engine of a MF285 tractor were measured in three directions at speeds of 1000, 1600, and 2000 rpm for ten different fuel levels obtained from different compositions of biodiesel, bioethanol, and diesel fuels. The volumetric percentages of biodiesel used in fuel combinations are 5, 10, and 15, and 2, 4, and 6 for bioethanol. To analyze the effects of the concerned parameters on engine vibrations, the response surface methodology (RSM) and artificial neural network fuzzy inference system (ANFIS) are applied. After optimization, the best fuel is obtained in terms of having the least vibration of the engine by the RSM and ANFIS. Meanwhile, all the predicted results are validated against the experimental data.
Materials and Methods
In the present study, a rapeseed-based biodiesel fuel was produced in a biodiesel production workshop using hydrodynamic cavitation reactors. Fig. 1 shows the process of producing biodiesel and its components.
Fig. 1. Biodiesel production system by hydrodynamic cavitations reactor (1- Oil tank, 2- Metoxed tank, 3- Homogenization tank, 4- Product tank, 5- Positive displacement pump, 6- Electro motor, 7- Hydrodinamic cavitations reactor, 8- Inverter, and 9- Pressure gauge
To produce biodiesel, rapeseed liquid oil, 99% pure methanol alcohol with a molar ratio of 6:1, and sodium hydroxide catalyst equal to one weight percent of oil were used. The conditions of biodiesel production included maintaining a temperature of 50 °C, a five-minute reaction time, a mixing intensity of 8000 rpm, and a pump flow rate of 0.83 L min-1. Following various stages of purification, including separating glycerin, methanol and water washing, 98% of the produced biodiesel was methyl ester. Then, 10 different fuel mixtures were prepared from biodiesel (Bi), bioethanol (Ej), and diesel (Dk). These compounds are as follows: B5E2D93, B5E4D91, B5E6D89, B15E2D83, B15E4D81, B15E6D79, B25E2D73, B25E4D71, B25E6D69, B0E0D100, where the subscripts represent the volumetric percentage in the fuel mixture. To conduct the experiment, a single-differential MF 285 tractor, manufactured by the Iranian Tractor manufacturing Company, was used. To collect the signals of engine vibrations, a VM120 vibration meter, made by Monitran Co., was employed. Three engine speeds of 1000, 1600, and 2000 rpm were considered as well as three directions x, y, and z. The experiments were carried out on dirt roads in constant and controlled manners, and the tractor was serviced prior to the experiments. The experiments on fuel mixtures were conducted in the stationary mode for all fuel mixtures in the non-load state. Moreover, the experiment equipment was installed on the right side of the engine in the driver’s line of sight, with no apparatus connected to the tractor. Vibrations were measured in three engine speeds, three repetitions, and three directions. A total of 270 experiments were conducted with respect to the 10 fuel levels. To test the given fuels, the pipes of the fueling circuit were cut off from the input and output of the secondary fuel filter. By creating a sub-circuit, the concerned fuels that were stored in a reservoir on the right side of the hood of the tractor entered the rotary injection pumps (like a watermill). Fueling was slow or disrupted because the primary pump (triangular pump) was disconnected from the fuel transfer to the injection pump, necessitating deaeration. To resolve this problem, the fuel reservoir was to be located at a higher point than the location of the injection pump in order for the fuel to be directed into the fueling circuit by the gravitational force. Therefore, benefiting from this information, the fueling operation was precisely conducted without any air in the course of the fueling circuit. A fuel return pipe was mounted on the top of the auxiliary fuel tank to transfer the rejected fuel into the reservoir. To commence the next experiment, the tractor was given sufficient time to consume its previous fuel in order to avoid fuel blending. After collection and storage of data related to the vibration acceleration of the MF 285 tractor engine, it was necessary to analyze them at different levels using RSM and ANFIS.
Modeling through RSM
RSM is a collection of mathematical and statistical modeling methods used for multiple regression analysis and applied to examine the relationship between one or more measured parameters. In this study, a three-dimensional hysterical data design was used for evaluation. To investigate the effects of the concerned parameters on engine vibrations in RSM, the three parameters including percentage of biodiesel variations, percentage of bioethanol variations, and engine speed in rpm were assumed as inputs where the mean of engine vibration in m s-2 was assumed as the output. Encoded and real levels of independent variables are presented in Table 1. The experiment design was arranged with three factors and three levels, including 21 factor points to adjust a second-order response surface. In order to validate the model, 9 input vectors from the test data were added to the model, and the output parameter was compared with the RSM predictions. Design-Expert software version 7.0.0 has been used to analyze and model the experimental data as well as designing respective diagrams.
| Run | Label in RSM | Average vibration (m s-2) | |||||
|---|---|---|---|---|---|---|---|
| A | B | C | |||||
| Biodiesel (%) | Code in RSM | Bioethanol (%) | Code in RSM | Engine speed (rpm) | Code in RSM | ||
| 1 | 5 | -1 | 2 | -1 | 1000 | -1 | 2.31 |
| 2 | 5 | -1 | 2 | -1 | 1600 | 0.2 | 3.54 |
| 3 | 5 | -1 | 2 | -1 | 2000 | 1 | 4.91 |
| 4 | 5 | -1 | 4 | 0 | 1000 | -1 | 3.10 |
| 5 | 5 | -1 | 4 | 0 | 1600 | 0.2 | 3.89 |
| 6 | 5 | -1 | 4 | 0 | 2000 | 1 | 5.48 |
| 7 | 5 | -1 | 6 | 1 | 1000 | -1 | 2.05 |
| 8 | 5 | -1 | 6 | 1 | 1600 | 0.2 | 3.13 |
| 9 | 5 | -1 | 6 | 1 | 2000 | 1 | 5.05 |
| 10 | 15 | 0 | 2 | -1 | 1000 | -1 | 2.14 |
| 11 | 15 | 0 | 2 | -1 | 1600 | 0.2 | 3.07 |
| 12 | 15 | 0 | 2 | -1 | 2000 | 1 | 4.78 |
| 13 | 15 | 0 | 4 | 0 | 1000 | -1 | 1.91 |
| 14 | 15 | 0 | 4 | 0 | 1600 | 0.2 | 3.27 |
| 15 | 15 | 0 | 4 | 0 | 2000 | 1 | 4.66 |
| 16 | 15 | 0 | 6 | 1 | 1000 | -1 | 2.03 |
| 17 | 15 | 0 | 6 | 1 | 1600 | 0.2 | 3.17 |
| 18 | 15 | 0 | 6 | 1 | 2000 | 1 | 4.48 |
| 19 | 25 | 1 | 2 | -1 | 1000 | -1 | 2.09 |
| 20 | 25 | 1 | 2 | -1 | 1600 | 0.2 | 3.15 |
| 21 | 25 | 1 | 2 | -1 | 2000 | 1 | 3.88 |
| 22 | 25 | 1 | 4 | 0 | 1000 | -1 | 1.71 |
| 23 | 25 | 1 | 4 | 0 | 1600 | 0.2 | 2.78 |
| 24 | 25 | 1 | 4 | 0 | 2000 | 1 | 4.68 |
| 25 | 25 | 1 | 6 | 1 | 1000 | -1 | 1.97 |
| 26 | 25 | 1 | 6 | 1 | 1600 | 0.2 | 3.08 |
| 27 | 25 | 1 | 6 | 1 | 2000 | 1 | 4.49 |
| 28 | 0 | -1.5 | 0 | -2 | 1000 | -1 | 2.65 |
| 29 | 0 | -1.5 | 0 | -2 | 1600 | 0.2 | 4.23 |
| 30 | 0 | -1.5 | 0 | -2 | 2000 | 1 | 4.25 |
Modeling through ANFIS
This system, an optimum combination of fuzzy logic and artificial neural network (Buragohain and Mahanta, 2008; Metin Ertunc and Hosoz, 2008), is invaluable for solving nonlinear problems in agricultural engineering applications (Arkhipov, Krueger, and Kurtener, 2008; Buragohain and Mahanta, 2008; Cheng, Cheng, and Lee, 2002). Furthermore, it is capable of establishing and inferring the linguistic concepts of nonlinear relationships between inputs and outputs (Naderlooet al., 2012). Data analysis was performed using ANFIS to meet two objectives. The first objective was to simulate the effects of the changes of various parameters on engine vibrations, and the second one was to present a predictive model.
Five important and effective factors were optimized in the ANFIS model: 1) type of input fuzzy sets (input membership functions include trimf, trapmf, gbellmf, gaussmf, gauss2mf, pimf, dsigmf, or psigmf), 2) number of input fuzzy sets (number of membership functions ranges between 2 and 6), 3) type of output fuzzy sets(output membership functions include linear or constant), 4) type of optimization (two optimization methods include backpropagation or hybrid), and 5) number of epochs (3 to 100). Thirty percent of the data (9 data sets) were used as the test set, and the remaining 70% (21 data sets) were assumed as the training set. The ANFIS models were developed using MATLAB R2013b with ANFIS Toolbox. Ultimately, the best obtained ANFIS model had three inputs (biodiesel, bioethanol, and engine speed) and one output (mean of engine vibrations). There were three Gaussian input fuzzy sets and a linear output fuzzy set. Additionally, a hybrid optimization method was applied with 30 epochs. In the end, the actual results were compared with the predicted results by the ANFIS model.
Results and Discussion
All measured data and the mean of engine vibrations are shown in Table 1.
The bar graph for the analysis of the measured data is shown in Figure 2. In this graph, the mean of engine vibrations is shown for different fuel mixtures over the intended periods.
As outlined in the graph, increasing the engine speed leads to a rise in the vibrations of the body of the engine due to a larger number of combustion courses and piston strokes per unit of time (Ghaderi, Naderloo, Javadikia, Mostafaei, and Rabbani, 2019; Javadikia, Naderloo, Safrangian, Mostafaei, and Mohtasebi, 2016; Safrangian, Naderloo, Javadikia, Mostafaei, and Mohtasebi, 2017).
Fig. 2. Total vibration values (vibration outcome) for various fuels under different period
This result was in good agreement with the results of previous studies (Heidaryet al., 2013; Patelet al., 2016; Salokheet al., 1995; Taghizadeh-Alisaraeiet al., 2012, 2016). Furthermore, Fig. 2 also indicates that pure diesel fuel had the most vibrations among all of the concerned fuels, and B5E4D91 fuel had also the highest extent of vibrations among the fuel mixtures.
Taghi zadeh-ali saraeiet al. (2012) investigated nine different fuel combinations at seven engine speeds and showed that the lowest vibration was for B40 and B20, and the highest vibration was associated with B15, B30, and B50 fuels (Taghizadeh-Alisaraeiet al., 2012). On the contrary, Heidaryet al. in 2013 obtained the opposite results, so that B100 and B5 fuels had the least vibrations, and B10 fuel produced the highest extent of vibration (Heidaryet al., 2013). The results obtained from the present study are closer to the results obtained by Heidary et al. In another study, the use of mustard oil biodiesel fuel reduced the vibration of the engine due to the enhanced oxygen quality (Tüccar, 2021). In explaining the cause of the changes in engine vibration in the combination of different fuels, it can be said that since the cetane number of ethanol is lower than biodiesel and diesel and the cetane number of biodiesel is slightly higher than diesel, the cetane number of the fuel is affected by different fuel combinations. This indicates that a higher cetane number of the fuel correlates with a shorter ignition delay and leads to reduced cylinder pressure at the piston peak, which in turn increases engine vibration and knocking (Ağbulutet al., 2020; Anand, Sharma, and Mehta, 2011; Debnath, Sahoo, and Saha, 2013; Emiroğlu and Şen, 2018; Radhakrishnan, Munuswamy, Devarajan, and Mahalingam, 2019).
The analytical results of RSM are shown in Table 2.
| Source | Sum of squares | df | Mean square | Value | Prob > F |
|---|---|---|---|---|---|
| Model A-Biodiesel | 33.41834621 | 9 | 3.713149579 | 35.43764 | < 0.0001 |
| 1.830436943 | 1 | 1.830436943 | 17.46937 | 0.0005 | |
| B-Bioethanol | 0.02233838 | 1 | 0.02233838 | 0.213194 | 0.6493 |
| C-Engine speed | 30.0939629 | 1 | 30.0939629 | 287.2115 | < 0.0001 |
| AB | 0.105229453 | 1 | 0.105229453 | 1.004291 | 0.3282 |
| AC | 0.002668351 | 1 | 0.002668351 | 0.025466 | 0.8748 |
| BC | 0.281520349 | 1 | 0.281520349 | 2.68678 | 0.1168 |
| A^2 | 0.117845216 | 1 | 0.117845216 | 1.124694 | 0.3015 |
| B^2 | 0.446491739 | 1 | 0.446491739 | 4.261238 | 0.0522 |
| C^2 | 0.795042817 | 1 | 0.795042817 | 7.587748 | 0.0122 |
| Residual | 2.095596278 | 20 | 0.104779814 |
The coefficient of determination (R2) in RSM was determined to be 0.9410. The real values of the mean of engine vibrations are illustrated in Fig. 3 versus the predicted values by RSM. The large and small graphs show the model's performance on the entire data set and the test set, respectively, comparing actual outcomes to predictions. Table 3 shows the performance of the best model in ANFIS. As specified in the table, although the correlation coefficient of the model for the test set is a little lower than that of the training set, the performance of the model is appropriate and acceptable compared to the entire data.
Fig. 3. Actual values and predicted values of mean vibration developed by RSM model; the smaller graph is for the test dataset
| Mean absolute error | Mean squared error | P-value | Correlation coefficient | |
|---|---|---|---|---|
| Results for test data | 0.338051 | 0.189907 | 0.000768 | 0.906 |
| Results for total data | 0.142017 | 0.077838 | 3.26×10-18 | 0.967 |
| Results for training data | 0.001847 | 7.46×10-4 | 4.77×10-44 | 0.999 |
Fig. 4 demonstrates the real and predicted values of the mean of engine vibrations using the ANFIS model. Similarly, the large and small graphs are related to the entire data and test set, respectively, i.e., the performance of the model compared to the predicted status.
Fig. 4. Actual values and predicted values of mean vibration developed by ANFIS model; the smaller graph is for the test dataset
In general, the model derived from ANFIS is an appropriate model capable of investigating the effects of the input parameters on the vibrations of the engine. In addition, regarding the good performance of the model in the test, it can predict engine vibrations for various inputs that have not been tested already. The values of R2 for the test set in RSM and ANFIS models were obtained as 0.86 and 0.82, respectively. A comparative analysis of the performance between RSM and ANFIS was conducted by plotting the residual graph, or model error, for both models, as illustrated in Fig. 5.
Fig. 5. The residuals graphs of ANFIS and RSM models
As shown in Figure 5, the error range of the ANFIS model was larger. However, the RSM model exhibited greater fluctuations in its error range, while the ANFIS model demonstrated impressively low errors in samples 19-27. Nevertheless, the performance of both models was acceptable and similar in many ways.
The disturbance graph of three input parameters compared to the mean of engine vibrations is shown for both models in Fig. 6. The figure depicts the interactions between the input parameters and the entire engine vibrations. The intersection of these graphs was their central point, indicating the mean variation range for each of the input parameters. Also, it showed interactions of the parameters as well as their effects on the mean of total vibrations, i.e., vertical axis. Also, the interactions between the input parameters can be studied, too.
Fig. 6. The disturbance graph of the effects of the input parameters on the mean of engine vibrations
According to Fig. 6, increasing the engine speed yielded an increase in the mean of engine vibrations in both models, which was in good agreement with the results of previous studies. In both models, with the increase of biodiesel concentration, the mean of engine vibrations was reduced, which was in good agreement with the results of other studies (Patelet al., 2016; Taghizadeh-Alisaraeiet al., 2012; Zöldy, 2011). In the RSM model, as bioethanol concentration increased from two to four percent, the mean of engine vibrations slightly increased. Moreover, as bioethanol concentration increased from four to six percent, the mean of engine vibrations was slightly reduced. These results are slightly inconsistent with the results of a study indicating that increasing the percentage of ethanol led to a permanent increase in vibrations of the engine. This might be attributed to the use of industrial ethanol instead of bioethanol or the ethanol blended with gasoline used in diesel engines. However, in the present study, a mixture of bioethanol, diesel, and biodiesel was used in a diesel engine (Keskin, 2010). It is worth noting that the slopes of the graphs of engine speeds in both models were much steeper than the slopes of those of both biodiesel and bioethanol graphs. In other words, the increase in vibrations due to a rise in the engine speed was higher than that by reducing the biodiesel and bioethanol concentrations. The effects of various parameters on the mean of engine vibrations are shown in the three-dimensional colored graphs in Figs. 7 to 9.
Fig. 7. Three-dimensional graph of biodiesel and bioethanol changes on average vibrations of the engine in ANFIS (right) and RSM (left) models
Fig. 8. Three-dimensional graph of biodiesel and engine speed changes on average vibrations of the engine in ANFIS (right) and RSM (left) models
Fig. 9. Three-dimensional graphs of bioethanol and engine speed changes on average vibrations of the engine in ANFIS (right) and RSM (left) models
Fig. 7 shows that as bioethanol and biodiesel concentrations increased, the mean of engine vibrations was reduced. The highest vibrations occurred for the concentrations of bioethanol and biodiesel in the mixture of 4% and 25%, respectively. The critical point of vibration occurred once the fuel was composed of five percent biodiesel and four percent bioethanol (B5E4D91), with the largest mean of engine vibrations of all fuel mixtures.
As seen in Fig. 7, increasing biodiesel concentration, the mean of engine vibrations was significantly reduced in both models. However, it is not possible to have the same interpretation on bioethanol since the mean of engine vibrations gradually increased and then decreased by increasing bioethanol concentration in RSM model. In ANFIS model, increasing bioethanol concentration percentage from zero to two, two to four, and four to six, the mean of engine vibrations almost remained constant, dropped, and increased, respectively.
The obtained results from the study conducted by Taghizadeh-Alisaraei and Rezaei-Asl (2016) showed that the addition of bioethanol to diesel fuel with 6% concentration, as compared with pure diesel fuel, increased the torque and engine power by an average of 8.3%. In comparison with pure diesel, applying this fuel mix increased the acceleration of the engine block by 4.79%, while an increase in bioethanol initially reduced the vibration. It seems that with an increase in bioethanol concentration in diesel fuel by more than 8%, the combustion delay will increase and the engine will operate irregularly. In other words, the pressure changes inside the cylinder and strokes occur. By increasing the concentration of bioethanol in pure diesel fuel, the vibration of the engine increases, which is in good agreement with the obtained results (Taghizadeh-Alisaraei and Rezaei-Asl, 2016).
As illustrated in Fig. 8, decreasing engine speed along with increasing biodiesel concentration, the mean of engine vibrations was reduced in both ANFIS and RSM models. The steep slope of engine speed-mean of engine vibrations diagram compared to the gentle slope of biodiesel-mean of engine vibrations diagram indicated that reducing the engine speed has more dominant effects on reducing vibrations. In 2017, Çelebi et al. showed that the vibration of the engine block increased by increasing engine speed, which corresponds to the result obtained (Çelebiet al., 2017).
As it can be observed in Fig. 9, in both models, the mean of engine vibrations increased with the rise of engine speed and fall of bioethanol. The lowest vibrations occurred for the concentration of bioethanol and engine speed of 4% and 1200 rpm, respectively. In 2016, Syed et al. investigated the effect of ethanol fuel mixes (E2, E5, E10, E15, and E20) on the vibrations of the engine block at the speeds of 1200, 1600, 2000, and 2400 rpm. They concluded that the lowest vibration of the engine block was for the E10 fuel mixture, while the highest vibration was for the E20 fuel mixture, which is in great agreement with the results of the present study (Javed, Murthy, Baig, and Rao, 2016).
Results of RSM Optimization
In RSM optimization, to identify the best fuel combination with the lowest engine vibrations, the conditions represented in Fig. 10 were applied. In other words, the optimization was aimed to find an appropriate fuel combination with the lowest engine vibrations. In RSM, B25E6D69 was selected as the appropriate fuel mixture. Regarding the maximum amount of biodiesel in the mixture, it could be inferred that the replacement of diesel fuel with biodiesel by 25% not only is possible but also can result in lower vibration effects.
Fig. 10. The optimization conditions towards identifying the best mixture with the lowest engine vibrations
Conclusion
Regarding the limited sources of diesel fuel and substitution of organic oils or waste, the present study aimed to investigate the engine vibrations of MF285 tractor caused by replacing a part of diesel fuel with biodiesel produced from rapeseed oil and bioethanol. The effects of ten fuel mixtures, made from different combinations of diesel, biodiesel, and bioethanol fuels with different volumetric percentages over different periods, on the tractor engine vibrations were analyzed through ANFIS and RSM models. The obtained results revealed that as the engine speed increased, the vibrations dramatically increased. Furthermore, increasing the bioethanol concentration, the vibrations were diminished. However, by increasing bioethanol concentration, the vibrations increased to some extent and then dropped. To reach the lowest engine vibrations, the best fuels in RSM and ANFIS models were B25E6D69 and B25E4D71 fuels, respectively, in which the volumetric percentage of biodiesel was 25%.
Acknowledgments
The authors would like to acknowledge the financial support from the Ministry of Science, Research and Technology, Tehran, Iran, and the Vice Chancellor for Research and Technology of Razi University of Kermanshah.
Conflict of Interest: The authors declare no competing interests.
Author Contributions
A. Safrangian: Data acquisition, Review and editing services
H. Javadikia: Supervision, Conceptualization, Methodology, Technical advice, Data acquisition, Data pre and post processing, Statistical analysis, Software services, Numerical/computer simulation, Validation, Visualization, Text mining, Review and editing services
L. Naderloo: Supervision, Conceptualization, Methodology, Technical advice, Data acquisition, Data pre and post processing, Statistical analysis, Software services, Numerical/computer simulation, Validation, Visualization, Text mining, Review and editing services
M. Mostafaei: Guidance on biodiesel production
S. S. Mohtasebi: Technical advice, Preparation of measuring devices.
References
- Ağbulut, Ü., Karagöz, M., Sarıdemir, S., and Öztürk, A. (2020). Impact of various metal-oxide based nanoparticles and biodiesel blends on the combustion, performance, emission, vibration and noise characteristics of a CI engine. Fuel, 270(February).DOI
- Anand, K., Sharma, R. P., and Mehta, P. S. (2011). Experimental investigations on combustion, performance and emissions characteristics of neat karanji biodiesel and its methanol blend in a diesel engine. Biomass and Bioenergy, 35(1), 533-541.DOI
- Arkhipov, M., Krueger, E., and Kurtener, D. (2008). Evaluation of Ecological Conditions Using Bioindicators: Application of Fuzzy Modeling BT - Computational Science and Its Applications – ICCSA 2008. In O. Gervasi, B. Murgante, A. Laganà, D. Taniar, Y. Mun, and M. L. Gavrilova (Eds.) (pp. 491–500). Berlin, Heidelberg: Springer Berlin Heidelberg.DOI
- Buragohain, M., and Mahanta, C. (2008). A novel approach for ANFIS modelling based on full factorial design. Applied Soft Computing, 8(1), 609-625.DOI
- Cahyono, B., Semin, S., and Putri, I. C. (2024). Experimental Investigation of Citronella Oil as Bioadditive in Biodiesel Fuel on Diesel Engine Performance, Vibration and Emissions. International Journal of Marine Engineering Innovation and Research, 9(2).DOI
- Çelebi, K., Uludamar, E., Tosun, E., Yıldızhan, Ş., Aydın, K., and Özcanlı, M. (2017). Experimental and artificial neural network approach of noise and vibration characteristic of an unmodified diesel engine fuelled with conventional diesel, and biodiesel blends with natural gas addition. Fuel, 197, 159-173.DOI
- Chaitidis, G. D., Marhavilas, P. K., and Kanakaris, V. (2022). Potential Effects on Human Safety and Health from Infrasound and Audible Frequencies Generated by Vibrations of Diesel Engines Using Biofuel Blends at the Workplaces of Sustainable Engineering Systems. Sustainability (Switzerland), 14(13).DOI
- Cheng, C.-B., Cheng, C.-J., and Lee, E. S. (2002). Neuro-fuzzy and genetic algorithm in multiple response optimization. Computers and Mathematics with Applications, 44(12), 1503-1514.DOI
- Debnath, B. K., Sahoo, N., and Saha, U. K. (2013). Adjusting the operating characteristics to improve the performance of an emulsified palm oil methyl ester run diesel engine. Energy Conversion and Management, 69, 191-198.DOI
- Emiroğlu, A. O., and Şen, M. (2018). Combustion, performance and emission characteristics of various alcohol blends in a single cylinder diesel engine. Fuel, 212, 34-40.DOI
- Erdiwansyah, M., Sani, M. S. M., Mamat, R., Gani, A., and Munawir, S. (2022). Effect of Vibration and Noise on Spark Ignition Engines of Methanol Fuel Blended with Gasoline. Journal of Scientific and Industrial Research, 81(1), 32–38.DOI
- Ferella, F., Mazziotti Di Celso, G., De Michelis, I., Stanisci, V., and Vegliò, F. (2010). Optimization of the transesterification reaction in biodiesel production. Fuel, 89(1), 36-42.DOI
- Ghaderi, M., Naderloo, L., Javadikia, H., Mostafaei, M., and Rabbani, H. (2019). Different blends of biodiesel, bioethanol, diesel and noise pollution emitted by stationary and moving MF285 tractor. Journal of Environmental Health Science and Engineering, 17(2), 743-752.DOI
- Guzzomi, A. L., Hesterman, D. C., and Stone, B. J. (2007). The effect of piston friction on the torsional natural frequency of a reciprocating engine. Mechanical Systems and Signal Processing, 21(7), 2833-2837.DOI
- Hansen, A. C., Zhang, Q., and Lyne, P. W. L. (2005). Ethanol-diesel fuel blends - A review. Bioresource Technology, 96(3), 277-285.DOI
- Heidary, B., Hassan-Beygi, R., Ghobadian, B., and Alisaraei, A. T. (2013). Vibration analysis of a small diesel engine using diesel-biodiesel fuel blends. Agricultural Engineering International: The CIGR e-Journal, 15(3), 117-126.
- Javadikia, H., Naderloo, L., Safrangian, A., Mostafaei, M., and Mohtasebi, S. S. (2016). Analyzing the MF285 tractor engine vibrations caused by different combinations of biodiesel, bioethanol and diesel as fuel by Response Surface Method. Iranian Journal of Biosystems Engineering, 47(2), 329-335.DOI
- Javed, S., Murthy, Y. V. V. S., Baig, R. U., and Rao, T. N. (2016). Vibration analysis of a diesel engine using biodiesel fuel blended with nano particles by dual fueling of hydrogen. Journal of Natural Gas Science and Engineering, 33, 217-230.DOI
- Keskin, A. (2010). The influence of ethanol-gasoline blends on spark ignition engine vibration characteristics and noise emissions. Energy Sources, Part A: Recovery, Utilization and Environmental Effects, 32(20), 1851-1860.DOI
- Lima Júnior, J., Magalhaes, R. R., Ferreira, D. D., and Pereira, P. H. C. (2020). Vibration Level Evaluation of Engines Fueled With Brazilian Commercial Diesel and Biodiesel. Theoretical and Applied Engineering, 4(3), 1-11.DOI
- Metin Ertunc, H., and Hosoz, M. (2008). Comparative analysis of an evaporative condenser using artificial neural network and adaptive neuro-fuzzy inference system. International Journal of Refrigeration, 31(8), 1426-1436.DOI
- Mirnezami, S. V., Hassan-Beygi, S. R., Banakar, A., and Ghobadian, B. (2017). Modelling total weighted vibration of a trailer seat pulled by a two-wheel tractor consumed diesel–biodiesel fuel blends using ANFIS methodology. Neural Computing and Applications, 28(December), 1197-1206.DOI
- Naderloo, L., Alimardani, R., Omid, M., Sarmadian, F., Javadikia, P., Torabi, M. Y., and Alimardani, F. (2012). Application of ANFIS to predict crop yield based on different energy inputs. Measurement: Journal of the International Measurement Confederation, 45(6), 1406-1413.DOI
- Patel, C., Agarwal, A. K., Tiwari, N., Lee, S., Lee, C. S., and Park, S. (2016). Combustion, noise, vibrations and spray characterization for Karanja biodiesel fuelled engine. Applied Thermal Engineering, 106, 506-517.DOI
- Radhakrishnan, S., Munuswamy, D. B., Devarajan, Y., and Mahalingam, A. (2019). Performance, emission and combustion study on neat biodiesel and water blends fuelled research diesel engine. Heat and Mass Transfer, 55(4), 1229-1237.DOI
- Rahimi, H., Ghobadian, B., Yusaf, T., Najafi, G., and Khatamifar, M. (2009). Diesterol: An environment-friendly IC engine fuel. Renewable Energy, 34(1), 335-342.DOI
- Safrangian, A., Naderloo, L., Javadikia, H., Mostafaei, M., and Mohtasebi, S. S. (2017). Investigating the engine vibration in MF285 tractor effected by different blends of biodiesel fuel using statistical methods and ANFIS. Journal of Agricultural Machinery, 7(1), 165-176.DOI
- Salokhe, V. M., Majumder, B., and Islam, M. S. (1995). Vibration characteristics of a power tiller. Journal of Terramechanics, 32(4), 181-197.DOI
- Saridemir, S., and Polat, F. (2023). Experimental Analysis of The Impacts of Biodiesel-Diesel Fuel Mixtures on Engine Vibration, Noise and Combustion. Düzce University Journal of Science and Technology, 11(5), 2352-2364.DOI
- Selim, M. Y. E. (2001). Pressure-time characteristics in diesel engine fueled with natural gas. Renewable Energy, 22(4), 473-489.DOI
- Subbaiah, G. V. V., Gopal, K. R. R., Hussain, S. A. A., Prasad, B. D. D., Reddy, K. T. T., and Andhra Pradesh. (2010). Rice bran oil biodiesel as an additive in diesel-ethanol blends for diesel engines. International Journal of Recent Research and Applied Studies, 3(3), 334-342.
- Susilo, S. H., Listiyono, L., and Khambali, K. (2022). Analysis of the Effect of Dieselessential Oil Fuel Mixture on the Performance, Noise, Vibration of Diesel Engines. Eastern-European Journal of Enterprise Technologies, 4(6–118), 16-21.DOI
- Taghizadeh-Alisaraei, A., Ghobadian, B., Tavakoli-Hashjin, T., and Mohtasebi, S. S. (2012). Vibration analysis of a diesel engine using biodiesel and petrodiesel fuel blends. Fuel, 102, 414-422.DOI
- Taghizadeh-Alisaraei, A., Ghobadian, B., Tavakoli-Hashjin, T., Mohtasebi, S. S., Rezaei-asl, A., and Azadbakht, M. (2016). Characterization of engine’s combustion-vibration using diesel and biodiesel fuel blends by time-frequency methods: A case study. Renewable Energy, 95, 422-432.DOI
- Taghizadeh-Alisaraei, A., and Rezaei-Asl, A. (2016). The effect of added ethanol to diesel fuel on performance, vibration, combustion and knocking of a CI engine. Fuel, 185, 718-733.DOI
- Tüccar, G. (2021). Experimental study on vibration and noise characteristics of a diesel engine fueled with mustard oil biodiesel and hydrogen gas mixtures. Biofuels, 12(5), 537-542.DOI
- Velmurugan, V., Aathif Akmal, S. M., Paramasivam, V., and Thanikaikarasan, S. (2020). Prediction of vibration and exhaust gas emission characteristics using palm oil with nano particle diesel fuel. Materials Today: Proceedings, 21(xxxx), 800-805.DOI
- Xu, H., Yin, B., Liu, S., and Jia, H. (2017). Performance optimization of diesel engine fueled with diesel–jatropha curcas biodiesel blend using response surface methodology. Journal of Mechanical Science and Technology, 31(8), 4051-4059.DOI
- Zöldy, M. (2011). Ethanol-biodiesel-diesel blends as a diesel extender option on compression ignition engines. Transport, 26(3), 303-309.DOI
