The Effect of Back Pressure Change on Exhaust Emissions According to the Confluence Geometry of a Dual Exhaust System in Idling

Abstract

In this study, a pressure transducer was installed in an exhaust system to analyze the effect of the change in back pressure according to the change of the confluence geometry of an exhaust pipe system on an exhaust emission. In addition, to perform exhaust gas measurement, the system was warmed up for about 40 s on the chassis dynamometer, and exhaust gas and back pressure measurements were performed simultaneously. In the back pressure measurement results, it was possible to confirm the difference in back pressure according to the change in the confluence shape. In addition, it was also confirmed that there was a clear difference in the exhaust emission measurement result. In particular, the H-type exhaust pipe system showed the highest pressure in the exhaust pipe due to the influence of the confluence geometry. Due to this influence, THC showed the highest measured value in the exhaust emission result. However, the X-type exhaust pipe system showed the lowest pressure due to the influence of the confluence geometry. Due to this influence, the THC showed the lowest measured value in the exhaust emission result. Therefore, through the conclusion of this study, an optimal exhaust system to reduce THC was proposed, and the importance of back pressure in exhaust system design was confirmed.

1. Introduction

In recent years, high-performance vehicles have been widely distributed due to the dramatically increasing development of automotive technologies. The recently launched expensive vehicles with V6 engines have adopted dual exhaust pipe systems to maximize engine performance with optimal back pressure design and evoke the emotions of consumers. In the exhaust pipe system, the exhaust gas from each exhaust pipe mounted in the engine is combined to maintain the exhaust gas pressure in a stable state. The change in the confluence geometry of the pipes significantly affects the engine output, exhaust sound, and exhaust emissions. The excellent performance and quality of the recently launched high-performance vehicles have also increased the expectations of many consumers. Consumers are attempting to tune their own vehicles in various methods for higher satisfaction, and motoring enthusiasts are tuning and operating their cars using off-the-shelf products and customized services, or through online shopping sites to suit their tastes [1]. The exhaust pipe system is tuned mostly to improve driving performance through the power performance upgrade of the engine or to acquire unique exhaust sound that can evoke emotions while driving.

Such tuning, however, is performed without considering environmental regulations in most cases. The size of the global automotive tuning market is known to reach approximately KRW 100 trillion. This is also evidence that many vehicle owners are attempting to perform tuning for their own preferences. In the future, however, a more careful approach to exhaust system tuning is required. This is because carbon neutrality has been declared a worldwide goal in preparation for climate change; environmental regulations on internal combustion engine vehicles are expected to be reinforced by the policies of countries that have plans to stop using fossil fuels due to global warming [2,3].

Therefore, the emission of harmful exhaust gas must be considered for tuning so as not to violate eco-friendly policies and at least Ultra Low Emission Vehicle (ULEV) standards in the Unites States for exhaust gas must be met. In addition, the design of the exhaust pipe system that emits engine exhaust gas is crucial. In low-speed rotation conditions, it is necessary to consider efficient exhaust gas emission, excellent fuel efficiency, and low noise and vibration through the formation of stable back pressure. At high speed, the flow area that can emit exhaust gas at a constant flow velocity and the geometry of the exhaust pipe must be considered for excellent driving performance. It is very difficult, however, for the exhaust pipe system to satisfy both the low- and high-speed ranges.

The flow area secured to exhibit optimal performance in the high-speed range may cause an increase in harmful exhaust gas in low-speed rotation conditions, such as the idling state, due to the backflow of combustion exhaust gas into engine cylinders [4]. This result is caused by the backflow of exhaust gas during the valve overlap or its combustion with exhaust gas that has not been discharged from cylinders. The valve overlap refers to the simultaneous opening of intake and exhaust valves at the top dead center (TDC) between the end of the exhaust stroke and the beginning of the intake stroke. Its benefit is the suction of a large amount of air due to the increased volumetric efficiency of the cylinder at high speed, and its drawback is a reduction in volumetric efficiency at low speed [5].

Thus, with respect to the valve overlap of the gasoline engine, various studies have been conducted on the effects of variable valve timing (VVT) and variable valve lift (VVL), i.e., devices related to the opening and closing of intake and exhaust valves inside cylinders, on exhaust gas emission [6,7,8,9,10,11,12].

It was confirmed that the residual gas in a cylinder affects combustion characteristics by lowering the flame speed. While the incomplete combustion caused by residual gas has the benefit of reducing NOx emissions by decreasing the internal temperature of the cylinder, it increases harmful exhaust gas, such as THC and CO, and causes an increase in vehicle body vibration due to unstable engine rotation under low-speed rotation conditions, such as the idling state [13,14,15,16,17,18,19,20,21,22,23].

Thus, various studies on the valve overlap have been conducted to improve fuel efficiency and reduce harmful exhaust gas with focus on the opening/closing timing and delay of intake and exhaust valves, and various experimental studies have been conducted on the components of the exhaust pipe system, such as the catalytic converter, main resonator, and sub resonator, as well as intake pipes and air filters related to the intake system [24,25,26,27,28,29].

It was difficult, however, to find various studies on the pipes of the exhaust pipe system despite their important purpose of connecting each component. Therefore, in this study, experimental research was conducted to examine the effect of the back pressure change according to the confluence geometry of an exhaust pipe system on exhaust emissions. The experimental conditions were based on the idling state, which is affected by the drawback of the valve overlap.

The main purposes of this study is to analyze the effect of the back pressure change on the emission of harmful exhaust gas and to propose the optimal geometry by analyzing the relationship based on the experiment results. Y Pipe, X Pipe, and H Pipe were selected as representative confluence geometry of the exhaust pipe system. For Y Pipe, two exhaust pipes are combined into one, which is again separated into two exhaust pipes. It exhibits the best power performance in the low-speed rotation range by maintaining the optimal back pressure as it has the smallest flow area at the exhaust gas confluence point. In the case of X Pipe, two exhaust pipes merge in the shape of X to maintain back pressure and share pressure and then separate again into two pipes. It shows the best power performance with the optimal back pressure from medium to high speed range. In the case of H Pipe, two exhaust pipes are separated, but they are connected with a pipe that acts as a bridge in the shape of H to maintain back pressure and share pressure. Its benefit is that the tone of the exhaust sound from the engine is generally rough enough to evoke the emotions of tuning owners [1].

2. Prototype Design

In this study, it was necessary to fabricate exhaust system prototypes for the experiment, and each exhaust system was modeled using the UG NX12 software. Each designed exhaust system had Y Pipe, X Pipe or H Pipe as confluence geometry [1]. The designed pipe geometry is a representative shape geometry in the automobile tuning market. Y Pipe is a basic pipe geometry, and X and H Pipe are high-end types and are widely used in high-performance vehicles. Therefore, when designing, each geometry was considered not to be interfered with by the vehicle’s lower frame and drive shaft to be used in the experiment. A sub resonator of the same model was used.

Figure 1 shows the designed exhaust systems and their components. The confluence position of exhaust gas, which is the main characteristic of the exhaust pipe system, was 600 mm for Y Pipe and 800 mm for X Pipe and H Pipe from the origin in the X-axis direction.

Additionally, the pipe center distance of the H pipe in the Y-axis direction from point 4 is about 117 mm. The difference in confluence position was caused by differences in geometry and the size of the off-the-shelf pipes used. Therefore, the cross-sectional area of flow was different at the confluence position.

Table 1. Sizes of exhaust pipe used for design and manufacture.

Type (Unit)	Y-Type	X-Type	H-Type
Pipe size (mm)	63.5	50.8	50.8
Cross section (mm2)	2874.75	3262.98	3589.02

shows the outer diameters and cross-sectional areas of the stainless steel structural pipes used.

3. Experiment Method

In the experiment of this study, pressure transducers were installed to measure the pressure of exhaust gas that flows inside the exhaust pipe. The pressure difference according to the confluence geometry was analyzed based on the exhaust pressure measurement results, and harmful exhaust gas was analyzed according to the confluence geometry using an exhaust gas analyzer. The vehicle used in the experiment is a rear-wheel-drive sedan equipped with a 3696 cc naturally aspirated gasoline engine and a 7-speed transmission. The specifications of the vehicle engine are shown in

Table 2. Specifications of the V6 engine used for the experiment.

Description	Specification
Type of Engine	DOHC 24-valve
Cylinder Type	V6
Bore, stroke	95.5, 86.0 mm
Compression ratio(:1)	11.0
Maximum Power	330 hp/7000
Maximum Torque	36.8 kgf∙m/5200

.

For the experiment, the vehicle was set on the chassis dynamometer as shown in Figure 2. The back pressure results measured by each pressure transducer were collected using a data logger, and the exhaust gas results were collected using a gas analyzer.

3.1. Back Pressure Measurement

For back pressure measurement, the pressure transducer and data logger were set up as shown in Figure 2, and it was confirmed that the ambient temperature was about 20 °C and the humidity was about 50% during the measurement.

The pressure transducers were installed at points 3, 4, and 6 in Figure 1. Point 3 had a distance of 100 mm from the X-axis origin, while point 4 had a distance of 800 mm and point 6 had a distance of 2400 mm. Figure 3 shows a schematic diagram of a pressure transducer installed in an exhaust pipe using a stainless steel screw nipple and socket.

The specifications of the pressure transducer used for the measurement are shown in

Table 3. Specifications for installed pressure transducers.

Description	Specification
Transducer type	Piezo resistive
Measurable Range	−30~30 kPa
Accuracy	±0.25
Operating temperature range	−20~100 °C
Output type	4~20 mA (2 wire)

, and the specifications of the data logger are shown in

Table 4. Specifications for installed data loggers.

Description	Specification
Number of input channels	20 channels
Accuracy	Voltage ± 0.1%
Sampling interval	10 ms (1 channel), 50 ms (4 channel)
Operating environment	0 to 45 °C, 5 to 85% RH

. The exhaust pressure of the vehicle was measured in the idling state in shift mode N. The engine RPM was 650 RPM ± 20, and measurements were repeated three times for 3 min at each measurement location.

3.2. Exhaust Emission Measurement

The equipment used to measure exhaust gas can measure harmful exhaust gas components emitted from the vehicle’s engine.

Table 5. Specifications of emission analyzer.

Description	Specification
Measuring item	CO, THC, CO2, O2, λ, AFR, NOx
Measuring method	CO, THC, CO2: NDIR, O2, NOx: Electrochemical cell
Repeatability	Less than ±2% FS
Operating temperature range	0~40 °C

shows the specifications of a gas analyzer using non-dispersive infrared (NDIR) spectroscopy. The measurement items analyzed in this experiment are CO, CO₂, THC, and excess air factor (λ).

The measurement method followed the idle test method. One exhaust pipe was arbitrarily selected, and the measurement probe was measured after being inserted 30 cm or more. The measurement procedure is shown in Figure 4. For the warm-up, the driving speed of 40 km/h was maintained for about 40 s, and for the idling test, after the chassis dynamometer roller stopped and the RPM stabilized, it was measured for 10 s. Measured values were arithmetic mean, and measurements were repeated three times in compliance with the experimental procedure.

4. Results and Discussion

4.1. Analysis of the Measured Back Pressure

In the back pressure measurement experiment, the pressure of exhaust gas inside the actual exhaust pipes was measured using the pressure transducers. As a result of checking the temperature of Bank 1 (Catalytic converter) using a vehicle scanner during warm-up according to the procedure in Figure 4 before measurement, the temperature increased from about 85 °C to about 300 °C.

Figure 5A–C show the time series plots of back pressure using the data measured at measurement positions 1, 2, and 3 for approximately three minutes. In the measurement state, the sampling interval was 50 ms. In the measurement results, H Pipe exhibited the highest pressure at measurement position 1, Y Pipe at measurement position 2, and H Pipe at measurement position 3. Based on the measurement results, the analysis of the time-series plots revealed that X Pipe had the highest standard deviation (0.773) at measurement position 1, Y Pipe (0.227) at measurement position 2, and X pipe (0.189) at measurement position 3. As for the lowest standard deviation, Y Pipe had a value of 0.625 at measurement position 1, X Pipe had a value of 0.174 at measurement position 2, and H Pipe had a value of 0.158 at measurement position 3.

Therefore, the standard deviation was shown to check the spread based on the arithmetic mean value in each of the measured back pressure results shown in Figure 5A–C.

The standard deviation result at the measurement position (A) showed a relatively higher value than the result confirmed at the measurement positions (B) and (C). This result is shown because it is directly affected by the exhaust pulsation, and it is confirmed as a position that can be affected by the valve overlap.

Moreover, the standard deviation was decreased at the measurement positions (B) and (C). From this result, it can be seen that the exhaust pulsation caused by the exhaust gas confluence was stabilized.

Figure 6A–C show the time-series plots of back pressure at each measurement position using the average data of three measurements performed for approximately three minutes and the total mean values.

Figure 6A shows the results measured at measurement position 1. The first, second, and third measurements showed that relatively high pressure was applied in H Pipe and relatively low pressure in X Pipe.

H Pipe exhibited the highest mean back pressure of 0.273 kPa. Y Pipe showed 0.245 kPa, which was approximately 10.3% lower, and X Pipe showed the lowest pressure of 0.169 kPa, which was approximately 38.1% lower.

Figure 6B shows the results measured at measurement position 2. The first, second, and third measurements showed that relatively high pressure was applied in H Pipe and relatively low pressure in X Pipe. H Pipe exhibited the highest mean back pressure of 0.362 kPa. Y Pipe showed 0.321 kPa, which was approximately 11.3% lower, and X Pipe showed the lowest pressure of 0.144 kPa, which was approximately 60.2% lower.

Figure 6C shows the results measured at measurement position 3. The first, second, and third measurements showed that relatively high pressure was applied in Y Pipe and relatively low pressure was applied in X Pipe. Y Pipe showed the highest mean back pressure of 0.288 kPa. H Pipe exhibited 0.280 kPa, which was approximately 2.8% lower, and X Pipe exhibited the lowest pressure of 0.256 kPa, which was approximately 11.1% lower.

The experimental results confirmed the characteristics of the pressure change according to the difference in the confluence geometry. Therefore, the measured back pressure results were compared with reference to Bernoulli’s theorem.

The H pipe showing the maximum pressure in Figure 6A,B is predicted to flow the exhaust gas at the slowest flow rate. In particular, the exhaust gas flowing from each exhaust pipe has a reduced flow velocity due to the disturbance of the flow resistance, such as a vortex generated at the confluence point.

As a result, the pressure increased by 32.6% more in measurement position 2 than in measurement position 1.

The X pipe with the lowest pressure was expected to flow the exhaust gas at a fast and stable rate. In particular, the exhaust gas flowing from each exhaust pipe naturally flowed to the confluence point, and the decrease in flow velocity was minimized.

As a result, the pressure decreased by 14.8% more in measurement position 2 than in measurement position 1.

The Y pipe showed a higher pressure than the arithmetic mean of the X pipe. The exhaust gas is expected to flow at a fast and stable flow rate inside the exhaust pipe.

However, when the exhaust gas flowing from each exhaust pipe naturally passed through the confluence point, the pressure increased due to the reduced flow area.

As a result, the pressure increased by 31% more in measurement position 2 than in measurement position 1.

4.2. Analysis of the Measured Exhaust Emissions

The components of the exhaust gas emitted after combustion inside the cylinders were analyzed using the low-speed idling test method. The items that were compared and analyzed were CO, CO₂, THC, and excess air factor (λ).

Figure 7A shows the CO results according to the confluence geometry of the exhaust pipe system. Overall, the mean CO result was found to be 0.01% based on three measurements. The CO emission was close to zero due to the influence of the lean AFR values compared above and complete combustion.

Figure 7B shows the CO₂ results according to the confluence geometry of the exhaust pipe system. Based on three measurements, Y Pipe exhibited a maximum CO₂ result of 14.3% and a mean value of 14.27%. Both X Pipe and H Pipe showed a maximum value of 14.5% and a mean value of 14.5%.

Therefore, the total mean CO₂ emission was found to be 14.4% under the influence of complete combustion because each exhaust system exhibited leaner AFR compared to the theoretical AFR in the idling state.

Figure 7C shows the THC results according to the confluence geometry of the exhaust pipe system. THC is representative of harmful gas that is emitted in the idling state. Based on three measurements, H Pipe exhibited relatively high values. The maximum value was 64.88 ppm and the mean value was 63.45 ppm. X Pipe showed relatively low values. The minimum value was 51.68 ppm, which was 25.6% lower compared to the maximum value, and the mean value was 52.25 ppm, which was 21.4% lower compared to that of H Pipe. In the case of Y Pipe, the mean value was 59.05 ppm, which was 7.4% lower compared to that of H Pipe.

Figure 7D shows the result of comparing the excess air factor (λ) according to the change in the confluence geometry of the exhaust pipe system.

In the result of the analysis based on three measurements, Y Pipe showed a maximum value of 1.031 and an arithmetic mean value of 1.028. In X Pipe and H Pipe, the arithmetic mean values of three times were 1.019 and 1.015, respectively.

Therefore, compared to the theoretical air intake, the actual air intake was confirmed to be the leanest in the Y pipe.

All of the measurement results satisfied ULEV standards in the Unites States for exhaust gas. In the idling state, however, the low exhaust gas temperature caused by lean excess air factor (λ) and ideal CO₂ emission through complete combustion may increase THC emissions by interfering with combustion as opposed to oxygen.

This is because combustion exhaust gas flows back into the cylinders again due to the pressure difference during the valve overlap and accelerates incomplete combustion together with the residual gas remaining in the cylinders.

In Figure 6, for back pressure measurement results (A) and (B), the highest pressure is shown in the H pipe, and the lowest pressure is shown in the X pipe.

Based on this result, when compared with (C), THC in Figure 7, it can be seen that the decrease in back pressure is related to the decrease in THC.

In order to clearly confirm this relationship, the results of back pressure measurement 1 and THC were used to show the scatter diagram, as shown in Figure 8. The scatter diagram shows that the THC emission increases as the back pressure increases. The result showing the lowest emission in the X Pipe group showing the lowest back pressure can be confirmed.

Therefore, a relatively large amount of combustion exhaust gas may flow back into the cylinders when a large pressure is formed inside the exhaust pipe.

The experiment results also confirmed that there was a large difference in THC emissions depending on the back pressure even though only the geometry of the exhaust pipe system was changed under the same conditions.

5. Conclusions

In this study, measurement of the back pressure of an exhaust pipe system and the test of exhaust emissions using a gas analyzer were conducted to analyze the effect of the back pressure change according to the confluence geometry of the exhaust pipe system on exhaust emissions in an idling state.

The most important back pressure measurement position in the back pressure measurement result is number 1. The reason is that it is a location that is greatly affected by changes in the confluence geometry.

Therefore, in the case of H Pipe, which exhibited the highest back pressure at measurement location number 1, a maximum of 64.88 ppm was shown in the result of THC exhaust emission. The reason is that the flow area of the exhaust pipe is the largest, but the flow resistance, such as a vortex, formed at the confluence point. This is because the flow resistance formed at the confluence position prevented the exhaust gas from flowing naturally and increased the pressure inside the exhaust pipe as the flow rate decreased due to interference and resistance. Therefore, the highest back pressure was 0.362 kPa at position 2 where the back pressure was measured. In addition, in the case of X Pipe, which showed the lowest back pressure at measurement location 1, the minimum value of 51.68 ppm was reduced by 21.4% in the result of THC exhaust emission. In the case of X Pipe, the flow of exhaust gas was naturally formed at the confluence position, creating a high flow rate, and as a result, the lowest back pressure of 0.144 kPa at measurement position 2 was shown. The Y pipe showed a higher pressure than the X pipe, and the THC emission increased by 13%. In the case of Y pipe, exhaust gas flows naturally at the confluence position, but the pressure inside the exhaust pipe is 0.321 kpa, which is 11.3% lower than that of H pipe due to the reduced flow area and long residence time.

Therefore, it was found through this study that the change in back pressure at measurement position 1, according to the change in the geometry confluence, affects the discharge of THC. These results were influenced by the valve overlap mentioned in the introduction. It results from a part of the exhaust gas that was burned back flowing into the cylinder under the influence of the back pressure formed inside the exhaust pipe. As a result, the THC emission increased from the exhaust pipe system where the relatively high back pressure was measured. These results are because the combustion performance decreased as the exhaust gas exhausted back to the cylinder under the influence of the valve overlap mentioned in the introduction. Therefore, the X Pipe, which showed the lowest back pressure in the experimental results of this study, minimized the exhaust gas THC emission by minimizing the backflow of the burned exhaust gas, and this result can be evaluated as the confluence geometry of the optimized dual exhaust pipe system.