Effect of fuel injection pressure and EGR techniques on various engine performance and emission characteristics on a CRDI diesel engine when run with linseed oil methyl ester

Abstract

Nowadays, owing to the reduction in petroleum supplies due to the growing oil demand, the search for alternate fuels has intensified. However, as alternate fuel choice grows, checking whether alternative fuels are suitable for use in engines has become time-consuming and expensive. Therefore, the usage of Linseed oil methyl ester (linseed biodiesel) in the common rail direct injection (CRDI) diesel engine was optimized for a smaller number of trials in this research. Response surface methodology (RSM) was employed for optimization. Input variables were chosen for LOME content in the blend, fuel injection pressure (FIP), exhaust gas recirculation (EGR) rates, and engine load while output parameters were selected for like indicated power (IP), indicated thermal efficiency (η(I)), indicated mean effective pressure (IMEP), hydrocarbon (HC), and NOx (Oxide of Nitrogen).The model layout employed in the analysis is focused on the matrix of the CCRD (central composite rotating design). The optimal input variables configuration is estimated at 5.45% LOME blend, 57.77 MPa FIP, 6.50% EGR, and 6.909 kg engine load leading to better efficiency together with reduced emissions. The optimized output of the engine at this input configurations are as IP 4.878 kW, IMEP 0.5886 MPa, indicated thermal efficiency 48.36%, HC 23.43 ppm vol., and NOx 533.15 ppm vol. Testing and optimum output response results are measured at acceptable input parameters and are considered to be within an acceptable error range. The findings of this analysis have shown that RSM is an appropriate technique for optimizing CRDI diesel engines.

Keywords

Linseed biodiesel performance emission common rail direct injection response surface methodology

Introduction

Increasing oil prices and facing restrictions on emissions from vehicles have intensified attention on the efficiency of the automobile sector. Besides, due to extensive usage, the rapid decline of fossil fuels has prompted the need for any minimum harmful emissions and sustainable fuel.¹ Renewable fuels such as alcohol, vegetable oils, animal fats, etc. have been claimed to be a preferable choice in the interest of substitute fuels. Vegetable oil is one the suitable choice for alternative fuel which almost fulfils the criteria of conventional fuel.² It is clear from past studies^3,4 that there are numerous problems involved with the usage of vegetable oils in diesel engines as a fuel. Vegetable oils generally have greater numbers of cetane number (CN) and low CV that prevent them from being used directly in the CI engine as fuel.⁵ some vegetable oils allow poor fuel flow due to having high viscosity and high molecular weight that leads to inadequate combustion, serious engine deposits, injector coking and clinging of the piston ring.⁶ To solve these challenges, many methods have been implemented, including preheating the oil, blending the oil with other fossil fuels or additives, micro-emulsification of oil, transesterification or thermal cracking / pyrolysis.⁷ Transesterification is the most generally used process for reducing oil viscosity and volume of oxygen.Viscosity and density of biodiesel (alkyl ester) are still higher than neat diesel even after the transesterification process.⁸ The high density of alkyl ester promoted the start of the fuel injections timing.⁹ Many reports show that the bulk modules of biodiesel are greater than conventional diesel, which enhances the timing of fuel injection.¹⁰ Throughout the case of biodiesel fuel, that could be the likely cause of the increased NOx.¹¹ Higher viscosity and biodiesel surface tension impact the fuel atomization by increasing the average fuel droplet size which in effect improves the penetration of the spray tip.¹² Many experiments have shown that viscosity is the dominant impact while density is the least on the average droplet size. Therefore, the first purpose is to reduce the viscosity of the fuel.¹³

The above-mentioned issue can be overcome by combining diesel with biodiesel, which brings the viscosity in an acceptable range. Injecting biodiesel at higher pressure is another way to improve atomization which will raise the atomization rate by growing the spray dispersal of biodiesel. In this test, linseed oil was used for the experiment because of its non-edibility, wide avail-ability, renew-ability, and environmental sustainability.¹⁴

Several researchers performed observational, optimization, and mathematical research to enhance the combustion and minimize the properties of hazardous emissions. Diesel engine efficiency and exhaust pollutants parameters depend on different factors such as amount of fuel supply, EGR technique, variable compression ratio (CR), fuel injection timing (FIT), and fuel injection pressure (FIP), etc. In a direct injection diesel engine, the fuel injection method is configured to obtain a high degree of atomization for the improved fuel injection to utilize the maximum air charge and to facilitate evaporation in a very short period and to boost the effectiveness of combustion. If the FIP is small, fuel particle sizes will rise and the time of ignition delay will also rise throughout the combustion. This state contributes to inadequate fuel combustion which allows the NOx, CO pollution to rise. As the intensity of injection rises the sizes of fuel particles may become low. The mixture of fuel and air which produces low smoke levels and CO emission is improved at the time of ignition delay.¹⁵ EGR technique decreases the flaming temperature and the oxygen content of the operating fluid within the engine cylinder which prohibited the formation of NOx.¹⁶ Therefore in this work, the effects of FIP and EGR were investigated.

The studies are experimentally looking into the impact of utilizing alternate fuels on engine pollution and vehicle efficiency characteristics. Yet, in terms of time and energy, those tests are not very successful. Researchers thus use computer-based methods to attain comparable efficiency and to discover through a limited number of trials the optimal rates of engine parameters.¹⁷ For this reason, prediction techniques, RSM is a proven method that several other systems have identified as a useful resource for optimization in numerous fields such as experimental process management and chemical analyzation, and it often decreases the number of tests needed, saves money, and requires lesser manpower. Several studies used RSM to investigate the impact of constant influences on the target reactions.¹⁸

Several studies used RSM to investigate the impact of constant influences on the target reactions. Gopal et al.¹⁹ applied RSM to evaluate optimum BSFC, NOx, and smoke, values by selecting the fuel blend content, EGR, and injection time as n-octanol/diesel fuel mixture input variables on a diesel engine. The 17% blending ratio, 20^o bTDC injection period, and EGR 10% were reported as optimal input configuration with a value of 0.967 R². BSFC, NOx, and smoke were reduced by 8%, 47.4%, and 21.08% respectively in these optimal conditions. Atman et al.²⁰ optimized the usage of diesel-butanol-cotton oil blends through RSM. Based on the operating conditions of the engine, optimum fuel blends were determined. A decrease in brake power, BTE, and BMEP were obtained in optimum fuel blends with an improvement in BFSC. The improved reduction was observed in the exhaust emissions. L.B. Moyo et al.²¹ studied the influence of reaction temperature, catalyst concentration and circulation flow rate in the development of biodiesel using the membrane reactor through the response surface methodology modelling. Experimental studies indicate that the greater the ratio of the catalyst to WCO, the greater the volume of free fatty acids (FFA). A maximum yield of 92.6 mole% for biodiesel at a temperature of 61^oC, circulation flow rate of 26 mL/min using a 1.3 wt % KOH catalyst concentration over a TiO2/Al₂O₃ membrane was achieved. A biodiesel yield of 94.03 mol % at 58.5^oC, circulation flow rate of 18.78 ml/min and 1.24 wt% catalyst concentrations was observed after membrane optimization. This study demonstrates that by optimizing various parameters, RSM can be used effectively to obtain higher yields for biodiesel production. Hirkude et al.²² established RSM experiments by using the blend of diesel and waste fried oil methyl ester to achieve optimum performance responses (BTE, EGT, BSFC, and smoke opacity (OP) for different process variables (CR, IT, and IP). The optimal parameter values for the least pollution and better efficiency were found to be 25^o bTDC IT (injection timing), 17.99 CR (compression ratio), and 250 bar IP (injection pressure). H.S. Pali et al.²³ concentrated on optimizing major features such as calorific value, viscosity, cetane number, and cold filter plug point (CFPP) through RSM on Sal biodiesel. Confirmation trails were obtained after the optimization tests and yield was achieved at 98.94%. Physicochemical properties such as calorific value 39.61 MJ/kg, viscosity 4.97 cSt, cetane number 61.91, and CFPP 0.024 °C were obtained in the same period.

Experimental researches carried out in different countries like India suggest that the efficiency and pollution parameters of the CI engine are influenced by many parameters like fuel injection pressure, blend ratio, EGR rates and the engine load. Nevertheless, the influence of the operating parameters is very important, and the researchers have extensively studied their impact on engine efficiency. From literature review no work found on CRDI diesel engine input parameter on which engine performance and emission were optimized based on linseed oil methyl ester. This work aims to optimize the variables responsible for engine efficiency as well as emission analysis for linseed methyl ester blends in CRDI diesel engines using the RSM optimization technique. For the present analysis, four factors were considered including blend%, engine load, FIP, and EGR rates to optimize the engine performance and emission characteristics. RSM is used to optimize the integration of the above variables to boost the indicated thermal efficiency η(Indicated), indicated mean effective pressure (IEMP), indicated power (IP) with the least emission of hydrocarbon (HC) and NOx. Therefore, linseed oil methyl can lead to better sustainability impact with enhanced economic interest.

Materials and methods

Biodiesel formation

Biodiesel derived from raw linseed oil by transesterification process was mixed with diesel which was bought from a private retailer in New Delhi. The reaction of the transesterification process is shown in Figure 1.

Figure 1.

The mechanism of the transesterification process.

In this process, 100 ml of linseed oil was placed in a beaker (with FFA less than 2 %) and warmed at 110oC to eliminate the content of water as shown in Figure 1. The oil is then left to cool down at about 60 °C and a mixture of methanol (20 ml) and KOH (1 gram) is slowly introduced to the oil and continuously stirred for around 30–40 minutes and covered with hard cardboard. It was then allowed to cool down for around 24 hours, without any agitation, in a separate funnel. Owing to the gravity effect, the mixture formed into two layers. The heavy portion settled down, i.e. glycerol, and fatty acid ester (linseed biodiesel) was the upper part. Lastly, the ester is washed 3 times to purify the linseed oil methyl ester (LOME) with warm water to eliminate impurities of methanol and KOH through water washing process. In this process, pure water at 45°C was spilled 50%(v/v) over the ester in a separating funnel and shaken smoothly and left for some time. The impurities and water have settled down and can be separated by separating funnels. The methyl ester was then heated for around 20 minutes at 110°C to evaporate the evidence of water and methanol inside the biodiesel.The oil (yield % = 88%) collected is fully miscible in fossil diesel. The 05 test samples are then marked as LOME5, LOME10, LOME15, and LOME20 based on a mixing ratio of 5%, 10%, 15%, and 20% with diesel fuel. Table 1 displays the important properties of different LOME blends. Density and kinematic viscosity is measured through Stabinger Viscometer.

Table 1.

The different properties of LOME blends.

Properties	Kinematic viscosity(mm²/sec)@40°C	Density(g/cm³)	Calorific value(kJ/kg)	Flashpoint(^oC)	Fire point(^oC)	Acid value(mg KOH/g)
ASTM methods	ASTM D7042	ASTM D7042	D6751 37	D0093	D7094	D6751
LOME	3.8894	0.8757	39100	174	187	0.341
LOME5	2.4988	0.8063	41750	62	67	–
LOME10	2.5874	0.8094	41400	78	84	–
LOME15	2.6872	0.8132	41050	96	101	–
LOME20	2.7384	0.8161	40600	102	105	–
Diesel	2.9783	0.8472	42100	55	64	–

Engine setup and methodology

The experimental set-up comprises of a single-cylinder CRDI-VCR diesel engine having 4-stokes, water-cooled system, fuel measurement devices, temperature specified thermocouples. The engine was run at a constant speed of 1500 rpm. The engine setup is shown in Figure 2. The system enables testing the efficiency of CRDI VCR engines with programmable ECU with varying compression ratios and multiple EGRs. Table 2 outlines the main engine characteristics used in this analysis. The engine was coupled with a water-cooled eddy current dynamometer which is used for measuring torque and brake power (BP). A Lab view based software “EngineSoft” was installed in the computer to measure the performance parameters.

Figure 2.

The setup of the CRDI diesel engine.

Table 2.

The important specification of the test engine.

Type	CRDI-VCR test engine
Engine	Kirloskar, 4-strokes, Single-cylinder, stroke 110 mm, 661 cc, bore 87.5 mm, water-cooled, Power 3.5 KW run at 1500 rpm.
CR range	12:1–18:1.
Dynamometer arm length	185 mm
Dynamometer	Water-cooled eddy current type
Connecting rod length	234 mm
Common rail	Having a pressure sensor with a valve for pressure regulating
EGR	Water cooled, Range 0–15%
Fuel tank	Capacity 15 litre with glass fuel metering column
Temperature sensor	PT100 and Thermocouple, Type K
Load sensor	Load cell, range 0–50 Kg, type strain gauge
Crank angle sensor	Make Kubler Germany, Resolution 1 Deg, Speed 5500 RPM with TDC pulse.
Inlet valve open	4.5 ^obTDC
Inlet valve closed	35.5 ^oaBDC
Exhaust valve open	35.5 ^obBDC
Exhaust valve closed	4.5 ^oaTDC
Software	“Enginesoft” Engine performance analysis software
Pump	Make Kirloskar, Type Monoblock
Overall dimensions	2000 W × 2500 D × 1500 H mm
Optional	Computerized diesel injection pressure measurement

Initially, the engine was run at a maximum speed of 1500 rpm with diesel fuel for 15 min with no load before the lubricating oil temperature increased to around 80 °C. In various test cycles, the same parameters have been applied during the experiment. With each run, the measurements were made 3 times, and the mean values were used to measure output parameters and exhaust pollutants. The performance characteristics IP, IMEP, and η(Indicated) were determined from measured results. The emissions parameters HC and NOx were measured through AVL gas analyzer “AVL DIX 650”. The resolution and accuracy of emissions parameters are listed in Table 3.

Table 3.

The Resolution and Accuracy of AVL gas analyzer for HC and NOx.

Parameters	Hydrocarbon (HC)	NOx
Units	ppm vol.	ppm vol.
Measuring	0–30000	0–5000
Resolution	≤2.000: 1	1
Accuracy	<2000: ±., ±3% o.M. ≥5000: ±5% o.M. ≥10,000: ±10% o.M	± 5 ± 1% o.M.

Each test was carried out at compression ration 18:1 and injection timing at 23^o CA bTDC (suggested by the manufacturer). All the tests were carried out by varying concentration of LOME in the blend at 5 levels (LOME5, LOME10, LOME15, LOME20, and diesel), FIP at 5 levels (50.0, 52.5, 55.0, 57.5, and 60.0 MPa), EGR rates at 5 levels (0%, 3.5%, 7%, 10.5%, and 14%), and engine loads at 5 levels (0 kg, 3 kg, 6 kg, 9 kg, and 12 kg).

Response surface methodology (RSM)

In the present analysis, RSM was used to model and evaluate reaction parameters to acquire engine characteristics. The ranges of the input parameters were chosen based on the acceptable limits under which the current engine would allow the modifications. The design of experiments was used to test engine efficiency over the full spectrum of input variable variance with the least number of tests. This approach optimizes the outcomes by establishing a balance between the input variables and output variables according to the input variables.^24,25

In RSM, the first move is to find an acceptable estimate of the actual operational relationship between the response and the number of independent variables as shown in Figure 3. The 2nd order model is seen in the given equation (1).

{Y = β}_{°} + \sum_{i = 1}^{k} β_{i} x_{i} + \sum_{i = 1}^{k} β_{i i} x_{i}^{2} + \sum_{j \geq 1}^{k} β_{i j} x_{i} x_{j} + ε

(1)

Figure 3.

The flowcharts of RSM techniques.

Throughout this effect, RSM is utilizing the least square method. The least-squares approach is used in evaluating the variables in the closely resembling polynomials. Optimization is then carried out on a surface that is designed.²⁶ Table 4 displays the levels and input factors. For this RSM model, the LOME content ratio, FIP, EGR rates, and engine load were considered as input variables and IP, IMEP, η(Indicated), HC, and NOx were taken as output responses. The flowcharts of RSM techniques are shown in Figure 4.

Table 4.

The coded levels of input variables.

Input factors units		Coded level
Input factors units		–2	–1	0	1	2
BLEND	%	0	5	10	15	20
FIP	MPa	50.0	52.5	55.0	57.5	60.0
EGR	%	0	3.5	7	10.5	14
LOAD	kg	0	3	6	9	12

Figure 4.

RSM optimizer.

Commercially accessible program MINITAB17 is used for the construction of an RSM model. Table 5 displays the descriptions of the tests carried out according to the CCRD process. It comprises a total of 31 experimental tests for the various input factors configurations. Optimization was conducted using RSM's derringers desirability method, where the solution with the maximum desirability was assumed the optimal one.

Table 5.

Layout of the experimental matrix.

Exp. No.	Blend ratio(% vol.) of LOME	FIP(MPa)	EGRRate (%)	Load(kg)	IP (kW)	IMEP(bar)	η (Indicated) (%)	HC (ppmv)	NOx (ppmv)
1	15	57.5	10.5	9	5.045	3.068	41.42	29	684
2	10	55.0	7	6	4.253	2.139	51.34	23	455
3	10	55.0	7	6	4.253	2.139	51.34	23	455
4	10	55.0	14	6	4.003	1.799	54.81	29	354
5	15	52.5	10.5	3	2.851	0.497	57.91	20	269
6	5	52.5	10.5	3	3.131	0.820	40.18	25	235
7	15	57.5	10.5	3	3.191	0.903	42.42	16	289
8	5	52.5	3.5	9	5.281	3.326	43.74	35	739
9	5	52.5	10.5	3	2.278	0.820	29.24	25	235
10	10	55.0	7	0	2.470	0.024	49.4	13	190
11	10	55.0	7	6	4.253	2.139	51.34	23	455
12	10	55.0	0	6	4.483	2.404	49.27	20	527
13	5	57.5	3.5	9	5.621	3.372	44.022	31	751
14	10	55.0	7	6	4.253	2.139	51.34	23	455
15	10	55.0	7	12	5.993	4.153	42.01	41	910
16	15	57.5	3.5	3	3.487	1.245	36.76	13	390
17	5	57.5	10.5	9	5.325	3.392	43.82	25	650
18	0	55.0	7	6	4.463	2.424	52.63	24	428
19	5	57.5	10.5	3	3.471	1.304	35.23	21	255
20	10	50.0	7	6	4.033	1.875	48.83	28	443
21	10	60.0	7	6	4.503	2.423	49.11	20	483
22	5	52.5	10.5	9	4.985	2.984	43.81	38	630
23	10	55.0	7	6	4.253	2.139	51.34	23	455
24	10	55.0	7	6	4.253	2.139	51.34	23	455
25	15	52.5	3.5	9	5.001	3.001	40.87	30	765
26	15	57.5	3.5	9	5.341	3.409	40.89	26	785
27	15	52.5	10.5	9	4.705	3.001	36.81	30	664
28	5	57.5	3.5	3	3.767	1.569	32.84	18	356
29	10	55.0	7	6	4.253	2.139	51.34	23	455
30	15	52.5	3.5	3	3.147	0.838	42.06	17	370
31	20	55.0	7	6	3.653	1.931	45.356	16	496

Results and discussions

Results from RSM

The variance analysis (ANOVA) was conducted in order to evaluate the model equation's statistical importance and reliability. The results of important individual words and their impact on the preferred response were also calculated by ANOVA. The results of ANOVA for engine performance and emission are shown in Tables 6 and 7. The findings indicate that the model has a strong degree of trust attributed to a lower P value (< 0.05). The P-value represents the possibility of a mistake that can be used to examine the reliability of every coefficient of regression. The P-value also shows the impact of each cross product on interaction. If the p-value is not greater than 0.05 then the model has a high impact on that parameter.²⁷ In this work, L² (quadratic impact of Load) has a high impact on the indicated thermal efficiency. Also, the L² (quadratic impact of load) has a great significance on HC and NOx emission. The “Lack of Fit” implies that the sample does not accurately explain the connection between the independent variables (B, P, E, and L) and the dependent variable IP, IMEP, η(Indicated), HC, and NOx.

Table 6.

The results of the variance of analysis for engine performance.

		Indicated thermal efficiency η (Indicated)			Indicated mean effective pressure (IMEP)			Indicated power (IP)
Sources	DF	Adj. SS	F-value	p-value	Adj. SS	F-value	p-value	Adj. SS	F-value	p-value
Model	14	740.82	1.29	0.309	35.2941	48.57	0.000	24.2554	48.57	0.000
Linear regression	4	144.52	0.88	0.497	0.2656	1.28	0.319	0.1825	1.28	0.319
B-Blend (%)	1	25.58	0.62	0.441	0.0395	0.76	0.396	0.0272	0.76	0.396
P-FIP (MPa)	1	95.41	2.33	0.147	0.0005	0.01	0.924	0.0004	0.01	0.922
E-EGR (%)	1	2.34	0.06	0.814	0.0323	0.62	0.442	0.0222	0.62	0.442
L-Load (kg)	1	10.36	0.25	0.622	0.1441	2.78	0.115	0.0991	2.78	0.115
Square	4	366.29	2.23	0.111	0.1046	0.50	0.734	0.0722	0.51	0.732
B²	1	106.55	2.60	0.126	0.1014	1.95	0.181	0.0700	1.96	0.180
P²	1	107.35	2.62	0.125	0.0004	0.01	0.934	0.0002	0.01	0.936
E²	1	39.22	0.96	0.342	0.0004	0.01	0.928	0.0003	0.01	0.926
L²	1	216.49	5.28	0.05	0.0016	0.03	0.862	0.0011	0.03	0.863
Way interaction	6	339.31	1.38	0.281	0.2311	0.74	0.624	0.1591	0.74	0.623
B×P	1	10.76	0.26	0.615	0.0360	0.69	0.417	0.0247	0.69	0.417
B×E	1	45.54	1.11	0.307	0.0180	0.35	0.565	0.0123	0.35	0.565
B×L	1	178.36	4.35	0.053	0.0356	0.69	0.420	0.0247	0.69	0.417
P×E	1	4.99	0.12	0.732	0.0180	0.35	0.565	0.0123	0.35	0.565
P×L	1	46.86	1.14	0.301	0.0360	0.69	0.417	0.0247	0.69	0.417
E×L	1	11.73	0.29	0.600	0.0180	0.35	0.565	0.0123	0.35	0.565
Error	16	655.68	*	*	0.8305	*	*	0.5707	*	*
Lack-of-fit	9	595.74	7.73	0.007	0.3011	0.44	0.874	0.2069	0.44	0.874
Total error	30	1396.50	*	*	36.1246	*	*	24.8261	*	*

DF: degree of freedom; Adj. SS: Adjusted sum of square; Adj. MS: adjusted mean square; *: nil.

Table 7.

The results of the variance of analysis for engine exhaust emission.

		Hydrocarbon (HC)			NOx
Sources	DF	Adj. SS	F-value	p-value	Adj. SS	F-value	p-value
Model	14	1261.59	25.82	0.000	1016033	158.35	0.000
Linear regression	4	22.14	1.59	0.226	1842	1.00	0.434
B-Blend (%)	1	6.15	1.76	0.203	0	0.00	0.978
P-FIP (MPa)	1	4.77	1.37	0.259	1299	2.83	0.112
E-EGR (%)	1	2.99	0.86	0.368	39	0.09	0.773
L-Load (kg)	1	4.97	1.43	0.250	323	0.70	0.414
Square	4	60.71	4.35	0.014	24740	13.50	0.000
B²	1	10.99	3.15	0.095	1360	2.97	0.104
P²	1	4.08	1.17	0.295	1460	3.19	0.093
E²	1	7.22	2.07	0.170	67	0.15	0.707
L²	1	36.27	10.39	0.005	23800	51.93	0.000
Way interaction	6	28.64	1.37	0.286	58	0.02	1.000
B*P	1	6.17	1.77	0.202	0	0.00	0.977
B*E	1	1.16	0.33	0.572	4	0.01	0.929
B*L	1	1.09	0.31	0.584	20	0.04	0.837
P*E	1	3.04	0.87	0.364	4	0.01	0.929
P*L	1	3.37	0.97	0.340	20	0.04	0.837
E*L	1	9.95	2.85	0.111	33	0.07	0.793
Error	16	55.83	*	*	7333	*	*
Lack-of-fit	9	55.83	*	*	7333	*	*
Total error	30	1317.42	*	*	10,23,366	*	*

*: nil.

The consistency of the exercise model has been checked by the determination coefficient (R²) as shown in Table 8. The determination coefficient (R²) for η(Indicated), IMEP, IP, HC, and NOx are 53.05%, 97.70%, 97.70%, 95.76%, and 98.66% and the determination coefficient modified (Adj. R²) are 11.97%, 95.69%, 95.69%, 92.05%, 98.66%, and 98.66% respectively. The provided coefficient value indicates the model’s strong accuracy. Also, for experimental relationships between variables and reaction, the model is very suitable.

Table 8.

The fitness of the model.

Model	S	R² (%)	Adj. R² (%)	Pred. R² (%)
Η (indicated)	6.40156	53.05	11.97	0.00
IMEP	0.227830	97.70	95.69	91.67
IP	0.188864	97.70	95.69	91.67
HC	1.86806	95.76	92.05	71.67
NOx	21.4080	99.28	98.66	95.37

Response optimizer

In the present experiment, the RSM optimizer is used to boost engine operating variables. The RSM optimizer is a method used to configure multi-targets such as IC engine instances. Here the key aim is to boost engine efficiency by maintaining the pollutants at their lowest value. NOx is a major pollutant that is produced during the combustion of biodiesel. Hence, minimize the variable (NOx) at its lowest value. The minimum amount of NOx is taken as the targets from the selection of measured situations.

Figure 4 displays RSM optimizer results. The optimum IP, IMEP, η(Indicated), HC, and NOx values are 4.87 kW, 0.5886 MPa 48.3661%, 23.43ppmv, 533.15ppmv respectively, with the corresponding optimum operational factors of a 5.45% LOME mix, 57.77 MPa FIP, 6.50% EGR, and 6.909 kg engine load. RSM may also be used to forecast and refine the parameters of the engine's function within the set of test values for variables. It lowers the number of experiments greatly while reducing time and energy.

The regression equations for various all responses valid at any input set can be determined by the submission of corresponding factors given by Table 9.

Table 9.

Regression equations coefficient for output responses.

Responses	NOx (ppm)	HC (ppm)	IEMP (MPa)	IP (KW)	η (Indicated)
Constant	3430	233	–2.4	–2.1	–841
Blend (%)	–0.7	–2.98	0.238	0.198	6.07
FIP (MPa)	–12.05	–0.730	0.0074	0.0063	3.27
EGR (%)	–11.0	3.03	–0.315	–0.261	–2.7
Load (kg)	35.9	4.46	0.759	0.629	–6.4
{Blend (%)}²	0.276	–0.0249	–0.00239	–0.00198	–0.0774
{FIP (MPa)}²	0.01146	0.000606	0.000006	0.000005	–0.00311
{EGR (%)}²	0.125	0.0411	–0.00032	–0.00027	–0.0958
{Load (kg)}²	3.212	0.1254	–0.00084	–0.00069	–0.306
Blend (%)×FIP (MPa)	–0.0013	0.00517	–0.000395	–0.000328	–0.0068
Blend (%)×EGR (%)	–0.030	0.0164	0.00204	0.00169	0.1025
Blend (%)×Load (kg)	–0.078	0.0181	–0.00328	–0.00273	–0.232
FIP (MPa)×EGR (%)	–0.0059	–0.00530	0.000407	0.000338	0.0068
FIP (MPa)×load (kg)	–0.0155	–0.00638	–0.000658	–0.000546	0.0238
EGR (%)×load (kg)	–0.145	–0.0799	0.00339	0.00281	–0.087

Each segment deals with the effect engine input variables have on engine output response. Four engine inputs (LOME blends, FIP, EGR rates, and engine load) are variables of engine regulation here. Under this section, MINITAB17 software created engine response surface plots for (IP, IMEP, η(Indicated), HC, and NOx) that are shown for all engine performance parameters in the subsequent subcategories. The MINITAB 17 software generates a surface plot of two engine controls while at the same time there are two other engine control variables defined as constant hold values. Such keep values are described in each engine reply surface plot's top corner towards the right and the CCRD table center value was taken automatically.

The impact of engine operating parameters on indicated power (IP)

The Indicated Power is the power generated in an IC Engine by combustion of the fuel within the cylinder. It is mainly the sum of brake power and frictional power. The combined effect of LOME blend content, FIP, EGR rates, and the load on the IP is shown in Figure 5. The hold values of LOME blend content, FIP, EGR rates, and the load are constant at 5.45%, 57.77 MPa, 6.505%, and 6.909 kg respectively. The surface plot and contour surface of the blend ratio and EGR rates for IP are shown in Figure 5(i) and (ii). It was observed that the IP of the diesel fuel is slightly greater than the entire blend% at all loads condition. This was likely due to an increase in density and viscosity but the low heating value of LOME blend compared with neat diesel.²⁸ A high BP will also be achieved using diesel fuel rather than LOME5, LOME10, LOME15, and LOME20. From the contour surface is clear that LOME20 produced minimum IP among all test fuel. It was also found that as the EGR rates increases, the IP of the engine gets reduced. The highest IP was achieved at a 6.5% EGR rate. EGR causes further exhaust gas volumes to be more in the combustion chamber instead of fresh air leading to dilution of the volume of fresh air. This results in the improper burning of fuel and produced less IP.

Figure 5.

Engine operating parameters impacts on indicated power (IP) (i) Surface plot of IP vs EGR, Blend (ii) Contour plot IP vs EGR, Blend (iii) Surface plot of IP vs EGR, FIP (iv) Contour plot IP vs EGR, FIP (v) Surface plot of IP vs FIP, Blend (vi) Contour plot IP vs FIP, Blend.

The surface plot and contour surface of FIP and EGR for IP are shown in Figure 5(iii) and (iv). It was found that IP increases with FIP values. It is mostly attributed to the greater degree of atomization at greater FIP and lowering of viscosity providing full combustion of fuel since the volume of fuel to be pumped is the same through varying FIP. Thus complete combustion of fuel generates a higher IP. Figure 5(v) and (vi) shows the change in the IP behaviour corresponding to blend ratio and FIP.

The impact of engine operating parameters on indicated mean effective pressure (IMEP)

The indicated mean effective pressure suggested is the constant pressure that will be needed during an engine's power stroke to perform a similar amount of work as the differing pressures received throughout the stroke perform. The combined effect of blend content and EGR for IMEP and its surface plot and contour surface are shown in Figure 6(i) and (ii) when the hold values are constant at 5.45% and 6.505% respectively. It was observed that IMEP decreases with increases in LOME content in the blend due to increases in indicated power (IP). IMEP is also gets affected by the EGR rates. As the EGR rates increase the corresponding value of IMEP reduces. This may be because of incomplete combustion of fuel due to the unavailability of fresh oxygen inside the combustion chamber.²⁹

Figure 6.

Engine operating parameters impacts on indicated mean effective pressure (IMEP) (i) Surface plot of IMEP vs EGR, Blend (ii) Contour plot IMEP vs EGR, Blend (iii) Surface plot of IMPEP vs EGR, FIP (iv) Contour plot IMEP vs EGR, FIP (v) Surface plot of IMEP vs FIP, Blend (vi) Contour plot IMEP vs FIP, Blend.

The surface plot and contour surface of FIP and EGR rates for IMEP display in Figure 6(iii) and (iv) when the hold values are constant at 57.77 MPa and 6.505% respectively. It was found that as the FIP increases the IMEP gets increases simultaneously. It may be attributed to the fact that it facilitates improved atomization of fuel within the cylinder as the FIP raises and provides a more surface closure to improve proper fuel combustion. The surface plot and contour surface of blend content and FIP for the IMEP display in Figure 6(v) and (vi) when the hold values are constant at 5.45% and 57.77 MPa respectively.

The impact of engine operating parameters on indicated thermal efficiency

The cumulative effect of LOME content in the blend, FIP, EGR and engine load on indicated thermal efficiency is displayed in Figure 7 with their surface plot and contour surface when the hold values are constant at 5.45%, 57.778 MPa, 6.505%, and 6.909 kg respectively. Figure 7(i) and (ii) shows the surface plot and contour surface of blend % and EGR rates for indicated thermal efficiency. It was found that indicated thermal efficiency reduces as blend concentration increases. It could be attributed to viscosity increases with the intensity of the blend resulting in lower atomization and incomplete combustion of fuel. Besides, greater LOME proportion in the blend induced a decrease in indicated thermal efficiency in comparison to diesel fuel owing to the lower energy content of LOME.³⁰ Indicated thermal efficiency can be shown to decrease significantly at higher EGR levels with all LOME/diesel blends.The increase in EGR obstructs the usual combustion process and reduces the burning capacity. It could be because the volume of fresh oxygen removed by the exhaust gas contributes to inappropriate combustion.

Figure 7.

Engine operating parameters impacts on indicated thermal efficiency (i) Surface plot of Indicated thermal efficiency vs EGR, Blend (ii) Contour plot of Indicated thermal efficiency vs EGR, Blend (iii) Surface plot of Indicated thermal efficiency vs EGR, FIP (iv) Contour plot of Indicated thermal efficiency vs EGR, FIP (v) Surface plot of of Indicated thermal efficiency vs FIP, Blend (vi) Contour plot of Indicated thermal efficiency vs FIP, Blend.

Figure 7(iii) and (iv) display the surface plot and contour surface of FIP and EGR rates. From this plot, it was observed that because of decreased viscosity, improved atomization, and greater combustion, the indicated thermal efficiency is enhanced at all load conditions with FIP. Optimal indicated thermal efficiency obtained at 57.7 MPa injection pressure. The surface plot and contour plot of LOME blend content and EGR rates for indicated thermal efficiency are shown in Figure 7(v) and (vi).

The impact of engine operating parameters on hydrocarbon emission (HC)

Figure 8(i) and (ii) shows the surface plot and contour surface of blend % and EGR rates for HC emission when the hold values are constant at 5.45% and 6.505%. It was noticed that the release of HC decreases with the increase in the ratio of LOME in the blend. Higher emissions of HC are due to an improvement in the volume of the blend which provides more oxygen than fossil diesel. More oxygen promotes proper combustion of fuel and minimum evidence of HC. It is assumed that the availability of oxygen in the fuel advances efficient combustion which contributes to a reduction of HC emissions. This drop shows the blend of diesel/linseed fuel being combusted more completely. HC emissions increase with higher rates of EGR. An increase in EGR levels results in a reduction of flame temperatures cause greater flame quenching regions where the ignition cannot take place easily.

Figure 8.

Engine operating parameters impacts on HC (i) Surface plot of HC vs EGR, Blend (ii) Contour plot of HC vs EGR, Blend (iii) Surface plot of HC vs EGR, FIP (iv) Contour plot of HC vs EGR, FIP (v) Surface plot of HC vs FIP, Blend (vi) Contour plot of HC vs FIP, Blend.

The surface plot and contour surface of FIP and EGR rates for HC emissions are shown in Figure 8(iii) and (iv) when the hold values are constant at 57.778 MPa and 6.505% respectively. From this surface plot it is seen that as the FIP rises, the HC reduces due to improved oxygen consumption, contributing to a higher rate of fuel combustion. Also, it could be attributed to the presence of more oxygen found in biodiesel, resulting in full fuel combustion. A similar trend of HC emission with FIP was shown by S V Channapattanaa et al.³¹ on Honne biodiesel.Figure 8(v) and (vi) display the surface plot and contour plot of blend % and FIP when the hold values are constant at 5.45% and 57.78 MPa.

The impact of engine operating parameters on NOx

The generation of NOx engine exhaust is primarily attributed to the high temperature reached during combustion within the cylinder, and this effect is caused by multiple input variables. The combined effect of LOME concentration in the blend, FIP, EGR, and engine load are shown in Figure 9. The surface plot and contour surface of blend %, EGR rates, and FIP are shown in Figure 9(i) to (vi) when the hold values are 5.45%, 6.505%, and 57.78 MPa. For all test fuel, the emergence of NOx with the biodiesel blend is considerably more than neat diesel. This trend is responsible due to enriching oxygen and the lower cetane number (CN) of LOME. The introduction of linseed biodiesel to diesel lowers the CN of the blend, leading to greater ignition delays. Therefore, during this period, much fuel mixture is injected within the cylinder, and when this higher fuel quantity is burned during the time of ignition delay, possibly cause high temperature, which is the suitable condition for the formation of NOx.

Figure 9.

Engine operating parameters impacts on NOx (i) Surface plot of NOx vs EGR, Blend (ii) Contour plot of NOx vs EGR, Blend (iii) Surface plot of NOx vs EGR, FIP (iv) Contour plot of NOx vs EGR, FIP (v) Surface plot of NOx vs FIP, Blend (vi) Contour plot of NOx vs FIP, Blend.

In CRDI diesel engines, the NOx emissions are minimized by adding the EGR technique. The explanations for rising NOx pollution in CI engines employing EGR are lowered oxygen content and reduced flaming temperature in the engine cylinder. Owing to a reduced oxygen supply, the reduction in NOx pollution is mostly attributed to an increase in in-cylinder temperature during the combustion cycle. Emissions of NOx decreased significantly at high EGR rates but resulted in higher BSFC and lower thermal efficiency.³² In order to get low NOx emissions in the exhaust, engine efficiency must be equilibrium. In terms of engine output, higher EGR levels are not recommended, as engine efficiency declines with higher EGR level. Yet engine NOx emissions decline with rises in EGR levels. Therefore 6.5% of EGR level is applied to get an optimized result between NOx emission and engine performance.

For FIP the levels of NOx are expected to rise. It could be because, at higher FIP, more fuel is burnt contributing to a high exhaust temperature which is a good environment for the production of NOx.

Validation of results

The goal of the RSM optimization is to improve the performance characteristics with the least emissions. For the present model, 5.45% of LOME content in the blend, 57.78 MPa FIP, 6.505 % EGR level, and 6.909 kg load on the engine is selected as the appropriate input parameters. The testing of optimal outputs is compulsory. Therefore, 3 tests have been registered and a mean value is taken at the optimum input configuration which involves LOME concentration in the blend, FIP, EGR, and the engine load as predicted by the RSM. In addition, the experiment results were conducted for confirmation along with the forecasted value observed with the aid of the model. The error %age is given in Table 10.

Table 10.

Verification test for RSM projected value with the experimental value.

Engine outcomes	Predicted value (RSM)	Experimental value	Error (%)
IP (kW)	4.879	4.78	2.03
IEMP (MPa)	0.5886	0.578	1.8
η(I) (%)	48.3661	46.368	4.13
HC (ppmv)	23.436	24	2.4
NOx (ppmv)	533.155	560	5.03

Note: Engine setting at 5.45% LOME blend ratio, FIP 57.78 MPa, EGR 6.505%, and load 6.909 kg.

Table 10 demonstrated that the developed models employing RSM for IP, IMEP, indicated thermal efficiency, HC, and NOx are reasonable to predict the LOME concentration in the blend, FIP, EGR, and engine load on the CRDI diesel engine's emissions and output. Forecasted mistake falls nearly 5% mark and should thus be regarded as appropriate.

Conclusions

The conclusions are made on the results of the 31 CRDI diesel engine test sets by concurrently changing the various input parameters. In this research, the influences of LOME content in the mixture, Fuel Injection Pressure (FIP), Exhaust Gas Recirculation (EGR) rates and Engine Load on Indicated Strength (IP), Indicated Thermal Efficiency, Indicated Mean Effective Pressure (IMEP), Hydrocarbon (HC) and NOx (Nitrogen Oxide) were assessed using RSM technique on a CRDI diesel engine. Variance analysis ( ANOVA) for the appropriateness and importance of the model was also conducted. The optimal and most effective optimization strategy has been shown to be the desirability technique of the RSM. It obtained maximum desirability of 0.64. This maximum desirability was achieved at 5.45% of LOME content in the blend, 57.78 MPa FIP, 6.505% EGR level, and 6.909 kg load on the engine which develops the corresponding output parameters as IP 4.878 kW, IMEP 0.5886 MPa, indicated thermal efficiency 48.36%, HC 23.43 ppm vol., and NOx 533.15 ppm vol. Raising the injection pressure led to an improved indicated thermal efficiency with lower HC and higher NOx emissions at all load conditions. Nevertheless, the consequences were invalidated when the injection pressure became too high. The determination coefficient (R²) for η(Indicated), IMEP, IP, HC, and NOx are 53.05%, 97.70%, 97.70%, 95.76%, and 98.66% and the determination coefficient modified (Adj. R²) are 11.97%, 95.69%, 95.69%, 92.05%, 98.66%, and 98.66% respectively. The average errors for IP, IMEP, indicate thermal efficiency, HC, and NOx were observed by 2.03%, 1.8%, 4.13%, 2.4%, and 5.03%, respectively from validation tests that verified between expected and experimental findings. Results have shown that RSM can be successfully used to predict CRDI diesel engine output and pollutants using linseed biodiesel fuel blends. Also, this work will assist engine designers and manufacturers in effective estimation of optimal engine configuration as a substitute fuel for the future, based on linseed biodiesel blends. In order to boost the efficiency of the diesel engine, the mixture of various non-edible vegetable oils should be used for probable opportunities.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Manish Kumar

Yashvir Singh

Manish Kumar is presently a research scholar at Delhi Technological University. He has also contributed in reviewing various articles.

Varun Kumar Singh is presently a research scholar at Delhi Technological University. He has also contributed in reviewing various articles.

Abhishek Sharma is presently working as a professor in Mechanical Engineering Department, G L Bajaj Institute of Technology and Management, Greater Noida, UP. He has published various research papers in peer reviewed journals. He is also reviewed various articles.

Naushad Ahmad Ansari is present working as a professor in Mechanical Engineering Department, Delhi Technological University, Delhi. He has published various research papers in peer reviewed journals.

Raghvendra Gautam is presently working as a professor in Mechanical Engineering Department, Delhi Technological University, Delhi. He has published various research papers in peer reviewed journals. He is also reviewed various articles.

Yashvir Singh is presently working as a professor in Mechanical Engineering Department, Graphic Era Deemed to be University, Dehradun, Uttarakhand. He has published various research papers in peer reviewed journals. He is also reviewed various articles.

Appendix

References

Murayama

Y-T

Miyamoto

, et al. Low carbon flower buildup, low smoke, and efficient diesel operation with vegetable oils by conversion to monoesters and blending with diesel oil or alcohols. SAE No. 8411, p.61.

Shrivastava

Verma

Pugazhendhi

An experimental evaluation of engine performance and emisssion characteristics of CI engine operated with roselle and karanja biodiesel. Fuel 2019; 254: 115652.

Adams

Peters

Rand

, et al. Investigation of soybean oil as a diesel fuel extender: endurance tests. J Am Oil Chem Soc 1983; 60: 1574–1579.

Saleh

Selim

MYE.

Improving the performance and emission characteristics of a diesel engine fueled by jojoba methyl ester-diesel-ethanol ternary blends. Fuel 2017; 207: 690–701.

Mosarof

Kalam

Masjuki

, et al. Optimization of performance, emission, friction and wear characteristics of palm and Calophyllum inophyllum biodiesel blends. Energy Convers Manage 2016; 118: 119–134.

Pramanik

Properties and use of Jatropha curcas oil and diesel fuel blends in compression ignition engine. Renewable Energy 2003; 28: 239–248.

Lin

Cunshan

Vittayapadung

, et al. Opportunities and challenges for biodiesel fuel. Appl Energy 2011; 88: 1020–1031.

Freedman

Pryde

Mounts

TL.

Variables affecting the yields of fatty esters from transesterified vegetable oils. J Am Oil Chem Soc 1984; 61: 1638–1643.

Szybist

Song

Alam

, et al. Biodiesel combustion, emissions and emission control. Fuel Process Technol 2007; 88: 679–691.

10.

Tat

Van Gerpen

Soylu

, et al. The speed of sound and isentropic bulk modulus of biodiesel at 21 _C from atmospheric pressure to 35 MPa. J Amer Oil Chem Soc 2000; 77: 285–289.

11.

Shrivastava

Salam

Verma

, et al. Experimental and empirical analysis of an IC engine operating with ternary blends of diesel, Karanja and Roselle biodiesel. Fuel 2020; 262: 116608.

12.

Ejim

Fleck

Amirfazli

Analytical study for atomization of biodiesels and their blends in a typical injector: surface tension and viscosity effects. Fuel 2007; 86: 1534–1544.

13.

Sharma

Ansari

Pal

, et al. Effect of biogas on the performance and emissions of diesel engine fuelled with biodiesel-ethanol blends through response surface methodology approach. Renewable Energy 2019; 141: 657–668.,

14.

Demirbas

Production of biodiesel fuels from linseed oil using methanol and ethanol in non-catalytic SCF conditions. Biomass Bioenergy 2009; 33: 113–118. 1

15.

Celıkten

An experimental investigation of the effect of the injection pressure on engine performance and exhaust emission in indirect injection diesel engines. Appl Therm Eng 2003; 23: 2051–2060.

16.

Pandian

Sivapirakasam

Udayakumar

Investigations on emission characteristics of the pongamia biodiesel–diesel blend fuelled twin cylinder compression ignition direct injection engine using exhaust gas recirculation methodology and dimethyl carbonate as additive. J Renewable Sustain Energy 2010; 2: 043110.

17.

Ganji

Chintala

Raju

VRK

, et al. Parametric study and optimization using RSM of DI diesel engine for lower emissions. J Braz Soc Mech Sci Eng 2017; 39: 671–680.

18.

Ileri

Karaoglan

Atmanli

Response surface methodology based prediction of engine performance and exhaust emissions of a diesel engine fuelled with canola oil methyl ester. J Renewable Sustain Energy 2013; 5: 033132.

19.

Gopal

Sathiyagnanam

Kumar

, et al. Prediction of emissions and performance of a diesel engine fueled with n-octanol/diesel blends using response surface methodology. J Clean Prod 2018; 184: 423–439.

20.

Atmanli

Ileri

Yilmaz

Optimization of diesel–butanol–vegetable oil blend ratios based on engine operating parameters. Energy 2016; 96: 569–580. 1

21.

Moyo

Iyuke

Muvhiiwa

, et al. Application of response surface methodology for optimization of biodiesel production parameters from waste cooking oil using a membrane reactor. South Afr J Chem Eng 2021; 35: 1–7.

22.

Hirkude

Padalkar

AS.

Performance optimization of CI engine fuelled with waste fried oil methyl ester-diesel blend using response surface methodology. Fuel 2014; 119: 266–273.

23.

Pali

Sharma

Singh

, et al. Sal biodiesel production using Indian abundant forest feedstock. Fuel 2020; 273: 117781117781.

24.

Bashir

Thiri

Yang

, et al. Purification of biodiesel via pre-washing of transesterified waste oil to produce less contaminated wastewater. J Clean Prod 2018; 180: 466–471.

25.

Khoobbakht

Karimi

Kheiralipour

Effects of biodiesel-ethanol-diesel blends on the performance indicators of a diesel engine: a study by response surface modeling. Appl Therm Eng 2019; 148: 1385–1394.

26.

Anwar

Rasul

Ashwath

Production optimization and quality assessment of papaya (Carica papaya) biodiesel with response surface methodology. Energy Convers Manage 2018; 156: 103–112.

27.

Simsek

Uslu

Determination of a diesel engine operating parameters powered with canola, safflower and waste vegetable oil based biodiesel combination using response surface methodology (RSM). Fuel 2020; 270: 117496.

28.

Hegde

Sharma

Raj

, et al. Assessing IP, FP and BP of diesel engine fueled with diesel and water-in-diesel emulsion using different surfactants. Invent Impact Combust 2014; 4: 187–191.

29.

Divekar

Tan

Chen

, et al. Characterization of exhaust gas recirculation for diesel low temperature combustion. IFAC-PapersOnLine 2015; 48: 45–51.

30.

Singh

Gandhi

Mahla

, et al. Experimental investigations on performance and emission characteristics of variable speed multi-cylinder compression ignition engine using diesel/argemone biodiesel blends. Energy Explorat Exploitation 2018; 36: 535–555.

31.

Channapattana

Pawar

Kamble

PG.

Effect of injection pressure on the performance and emission characteristics of VCR engine using honne biodiesel as a fuel. Mater Today: Proc 2015 ; 2: 1316–1325.

32.

Rao

KS.

Effect of EGR on diesel engine performance and exhaust emission running with cotton seed biodiesel. Int J Mech Mechatron Eng 2016; 16: 64–69.