Numerical Study on Prediction of Icing Phenomena in Fresh Air and Blow-by Gas Mixing Region of Diesel Engine under High Velocity of Intake Air Condition

Abstract

The icing of an intake pipe that might happen in an actual vehicle was numerically predicted in this study. For various operating conditions, the amount of icing was estimated, and the variables influencing the amount of icing were identified. We compared the factors that affected icing: relative humidity, air temperature, and inlet velocity. Seven RPM and load conditions, an intake temperature range of 253–268 K, and a relative humidity range of 65–85% were used for the case studies. To verify the model accuracy, wind tunnel test results from chassis dynometer tests were compared to the data from simulations. The flow analysis was performed using the numerical analytical tool ANSYS Fluent (2019 R1), while the amount of condensed water and icing was predicted using FENSAP-ICE, a program that analyzes and predicts icing phenomena under mechanical systems. The ambient temperature, relative humidity, and inlet air velocity had the biggest effects on the icing rate. The total amount of icing increased for similar BB and input air velocities. When the input air and BB velocities are the same, the variables influencing icing are the ambient temperature and relative humidity. The amount of ice was less affected by outside temperature and relative humidity when the rpm was high, and the inlet air velocity also had an impact.

1. Introduction

Compared with that in the past, the driving environment of vehicles has diversified, and exhaust regulations have been tightened. This diversification can be attributed to two reasons. First, globalization of the automotive industry means that each country’s climatic conditions must be considered. Secondly, extreme weather conditions have been observed because of global warming. In addition, exhaust regulations have become more stringent worldwide, the proportions of pure internal combustion engine vehicles and eco-friendly vehicles have decreased and increased, respectively. However, eco-friendly vehicles lag behind internal combustion engine vehicles in terms of efficiency and convenience. The internal combustion engine vehicle industry has developed technologies to comply with tighter emission regulations.

Typical technologies include exhaust gas recirculation (EGR), various post-processing devices, and positive crankcase ventilation technology. The harmful gases emitted by internal combustion engines can be classified into exhaust gases and blow-by gas (BB). Exhaust gases from gasoline engines are treated through a three-way catalyst (TWC), while diesel engines undergo after-treatment systems before being discharged to the outside [1,2]. BB is the gas discharged through a gap between the cylinder and piston during the compression or explosion stroke and accumulates in the crankcase owing to the mixture of unburned and combustion gases with engine oil. BB is usually discharged quickly because its high acidity causes rust to develop inside the engine and dilutes engine oil. Previously, BB was released into the atmosphere; however, to reduce air pollution, it is now recirculated into the combustion chamber for complete combustion. Delprete et al. conducted a sealing analysis based on three parameters: ring spacing, mass and tension between cylinders and pistons, and static torsion [3]. Although the ring between the cylinder and piston was responsible for controlling the BB discharge, the second ring was the most important. The inertial force is proportional to the mass and acceleration; thus, the lower the mass of the ring, the better. However, it can be destroyed when subjected to a high engine velocity and radial load; therefore, further studies have been conducted on other parameters, such as the ring face and lubricating oil patterns. Cavallaro et al. simulated the thermal conditions to evaluate BB [4]. They observed that the thermal bore expansion directly affected the ring gap, and the temperature of the bore and piston affected the gas pressure and flow around the ring pack. Ebner and Jaschek developed a damper to introduce the orifice measurement principle into a BB measuring devices and studied the improvement in BB measurement efficiency by separating the contaminants in BB from sticky condensed water [5]. Edelbauer et al. conducted a study on BB flow using computational fluid dynamics (CFDs) to reduce crankcase ventilation loss and quantify BB emissions at low and medium engine velocities [6]. They confirmed that the BB losses were negligible at low and medium velocities and significant in the high-velocity range. Rakopoulos et al. conducted a study using a CFD code to predict the BB discharged according to the cylinder pressure around the ring pack and clearance velocity between the cylinder and piston head [7]. The accuracy regarding complex physical phenomena is increasing with high-performance computing and commercial CFD simulations [8,9,10]. Mitchell et al. compared the blow-by emission characteristics of diesel- and oxygen-containing fuels and determined that oxygenated fuels during cold starts improved fuel performance and suppressed BB emissions [11]. Nabi et al. compared the results of a mixture of diesel and oxygenated blends in terms of BB, and accordingly, a significant amount of BB emissions were emitted in a diesel cold-start environment [12]. In particular, the PM emissions were higher than those of NO_x, and the BB contained a mixture of exhaust gas and oil. Pagnozzi et al. proposed an air/oil separation method using a ventilation valve inside the crankcase for vehicles that did not reinject the BB into the intake manifold [13]. Kolhe et al. developed a new oil separation system to prevent the cyclone-type oil mist systems utilized in vehicles by transferring more oil to the intake system under transient velocity and load conditions [14]. The BB discharged from the vehicle was circulated in a closed loop system. Air/oil separation and rings have been studied, but there is a lack of research on the effect of an insufficient intake of air.

A portion of the exhaust gas discharged after combustion inside the cylinder is utilized in the EGR. EGR is a commonly employed technology for reducing NO_x emissions from internal combustion engines. The application of EGR according to the operation area is related to smoke emissions and reduction in thermal efficacy. Divekar et al. studied the effect of EGR on achieving ultralow NO_x emissions while minimizing smoke and efficiency penalties [15]. Studies have indicated that NO_x reduction is directly correlated with intake port dilution due to EGR and is independent of the fueling strategy, intake boots, and engine load level. Xie et al. conducted a study to compare the effects of hot and cooled EGR. Hot EGR improves fuel consumption, NO_x, and PM in the engine, whereas cooled EGR accelerates flame development and propagation velocity [16]. Shen et al. studied the effects of fuel consumption, combustion processes, and emissions to compare high pressure exhaust gas recirculation (HP EGR) and low pressure exhaust gas recirculation (LP EGR) fixed on a turbocharged GDI engine. The optimal combustion process is an important factor in reducing BSFC owing to the dilution effects of EGR, and it is suggested that the application of LP EGR at low velocities and the use of HP EGR at high velocities will improve BSFC and emissions [17].

EGR is an emission reduction technology employed to comply with emission regulations. HP EGR allows the immediate use of gases released after combustion. It has the advantage of a short travel path and does not significantly affect the durability of the turbocharger (TC). The performance of the HP EGR cooler is important because it utilizes high-temperature gas after the exhaust. Bourgoin et al. utilized an HP EGR cooler at a low temperature and mixed a large amount of exhaust with intake air to significantly reduce NO_x [18]. However, acidic condensate is generated in the low to middle velocity range, and this can corrode parts of the cooler. Lujan et al. developed a mathematical model to predict the condensation that occurs when HP EGR is utilized in low temperature environments [19]. Condensate generation depends on the engine operating conditions, ambient humidity, temperature, and pressure. It is confirmed that condensation occurs depending on the pressure and humidity. However, the increased use of HP EGR has the disadvantage of reducing the amount of exhaust sent to the TC, which reduces the torque of the turbine, thereby preventing it from generating the required output. LP EGR recirculates the gas that has passed through the TC and after the treatment device. Although it is less reactive than HP EGR, it has sufficient time to mix with intake air and allows the use of significant amounts of exhaust. However, LP EGR has a long supply line and temperature difference, resulting in condensate water passing through the compressor of the TC, affecting the durability of the compressor and the overall intake system. Oh et al. identified the condensate generation characteristics inside an intercooler under various operating conditions and studied the variables affecting the generation [20]. Typically, the mixed gas that enters the intercooler comprises LP EGR and air. The LP EGR contains NO_x, which tends to increase in acidity as EGR usage increases. Ko et al. studied compressor and intercooler corrosion using condensates produced by LP EGR [21]. Galindo et al. developed a model to evaluate the LP EGR condensates and determined that the initial temperature of the cooling water had the greatest impact on condensate generation [22]. Karstadt et al. determined that the compressor wheels were affected by the condensate generated by LP EGR. Accordingly, the larger the liquid volume, the more the physical damage endured; however, the vane can be protected by the film coating [23].

Internal combustion engine vehicles have adopted dual-loop EGR technology to offset the disadvantages of HP and LP EGR. Zamboni and Capobianco evaluated various operating modes with HP and LP EGR rates to improve the maximum NO_x and soot emissions by experimenting with various modes of operation applying HP or LP EGR and dual-loop EGR [24]. The optimal amount was mapped to each operating condition to suit the various operating conditions. Both fuel economy and NO_x emissions can be improved using dual-loop EGR compared to HP EGR alone. Park et al. utilized a 1D simulation to convert the model from a basic engine model with an HP EGR system to a dual-loop EGR system, which improved vehicle performance by increasing the fuel economy and reducing NO_x [25]. Park had also predicted the performance and emissions characteristics under optimal mitigation strategies for dual-loop EGR applications [26]. By applying multi-objective Pareto optimization method with model-based control to minimize NO_x and brake specific fuel consumption, optimal EGR split ratios were proposed by introducing the EGR split index [25,26,27].

The above-mentioned studies have been conducted with a focus on the importance of LP EGR as the need for it increases. On the other hand, the problem of condensation due to LP EGR usage has been known for a long time, but the academic approach has recently become increasingly important as field issues with condensation have arisen accompanying icing problems in engine system. There are a few studies that have linked academic results on condensation and icing in automobile engines to field issues, and the study by Yoon et al. provided almost the only results that correspond to the study prior to this study [28]. Yoon et al. studied the amount of ice generated when blow-by gas (BB) mixed with HP EGR and LP EGR meets sub-zero air in a low- and middle-velocity region in the intake system.

In this study, condensation and icing phenomena are introduced using numerical validation and prediction based on experimental results under wind tunnel tests with a chassis dynamo, while previous research results have focused on the low rpm to middle rpm range representing low to middle inlet air conditions. The results showed that different factors appeared to be affecting the condensation and icing under high-inlet air velocity conditions compared to low- and middle-inlet air velocity conditions. Figure 1 shows the overall research flow and main results from the CFD analysis that can express the key methods of the present study. By applying a three-step simulation process, which is ANSYS fluent (2019 R1), FENSAP-Drop 3D (2019 R1), and FENSAP-ICE (2019 R1), the icing phenomena can be validated and predicted, having good agreement with wind tunnel test results.

2. Methodologies

2.1. Modeling Overview

The vehicle was usually operated under various rpm and load conditions. When driven in a low-temperature environment, the intake air flow had the same temperature as the outside temperature, whereas the BB from the crankcase was discharged at a higher temperature than the intake air. The discharged BB contained several mixtures and condensates. The condensed water was generated due to temperature differences between the inlet air and gas temperature of blow-by gas which contained water vapor as the combustion product. Condensate was generated in the area where the BB encountered cold air, and it froze and began when driving for a long time in a low-temperature environment. This obstructed the BB outlet and hindered its smooth release to upstream of the compressor. If the BB was not discharged smoothly, pressure in crankcase may increase, and overall efficiency of engine performance will be lost. Therefore, it is very important to predict the icing amount, icing rate, and icing location depending on numerical simulations. Figure 2 shows the mesh quality and main result contour for geometry. The mixing zone where cold air and BB meet was designed with a fine mesh; three layers of inflation were placed on the wall surface. Starting from the construction of simple mass balance calculations, appropriate factors could be removed that can have an effect on icing phenomena covering various engine parameters we should consider, such as equivalence ratio, chemical formulas to describe air and fuel properties, EGR rates, and thermodynamic properties at each part of the engine cycle system. Then, the determined factors were used as input boundary conditions for parametric studies under a CFD environment.

2.2. Mass Balance Equations

Figure 3 illustrates a geometry of the icing prediction zone in the intake system where air, blow-by gas, and LP EGR is mixed at the upstream of the compressor of the target engine system and the schematic of the mass balance diagram, indicating the incoming and outgoing flows. Assuming that the RPM and load is constant, each part has the same mass flow rate upon entry and exit. ${\dot{m}}_{xx}$ is the mass flow rate which is obtained from the chassis dynamometer test under the wind tunnel. X_xx indicates the mass fraction of flow rate at each part. The engine utilized in this study is a compression ignition engine in which fuel was injected during the compression stroke. X_bb is expressed in terms of the amount of m_bb/total mixed gas, considering that some of the cylinder gas escapes to the BB. A detailed explanation of the mass balance equation and CFD modeling can be found in the previous paper [28].

Table 1. Engine specifications and operating conditions.

Item	Specification
Stroke	83 mm
Bore	99 mm
Displacement	2000 cc
Compression ratio	16:1
Operation load	7 point
Blow-by gas composition	CO2, H2O, O2, N2

represents the target engine specifications for this study. Seven cases were determined to measure the real amount of ice that occurred when a positive temperature coefficient (PTC) heater was used. The exhaust gas in the BB is composed of CO₂, H₂O, O₂, and N₂. Only the major components that remained after the complete combustion of the hydrocarbon fuel were considered. Figure 2 illustrates a schematic diagram of the mixing zone boundary condition and temperature sensors, where air enters the intake pipe and the BB enters the S-pipe. The area where the air and blow-by gas are mixed is referred to as the mixing zone, which is the area where condensed water is generated and freezes when it hits a cold wall. In the mixing zone, the air and BB were mixed before exiting the outlet. Each wall material property value including the heat transfer coefficient for thermal conductivity were input as boundary conditions. The maximum temperature of the PTC heater and the temperature at the end of the breather hose were 393 and 353 K. The wall surface temperature around the mixing zone was 293 K. Geometry for the CFD domain was modelled from the mixing zone to the mixed and exit areas. The reliability of the model was determined by comparing the actual vehicle wind tunnel analysis results with CFD results.

Table 2. Parameter ranges.

Load	Inlet Air		Total Point
	Temperature [K]	Relative Humidity [%]
7 load	268	65 75 85	84 point
	263
	258
	253

indicates that air temperature and humidity are used as external factors which can have an effect on the icing phenomena. Temperatures from 268 K to 253 K having 5 K intervals and humidity levels from 65% to 85% with 10% intervals were evaluated to predict the condensation and icing amounts.

2.3. Geometry Modeling

The shape employed in this study is the region where air and BB gases are mixed in the intake pipe at the front stage of the compressor. BB gas flows into the S-pipe, passes through the breather hose, and meets and mixes with sub-zero air in the mixing zone. The material of the S-pipe wall surface is synthetic rubber, and the breather hose comprises aluminum, which conducts the heat of the PTC heater. Therefore, the heat of the PTC heater resolves the breather hose outlet icing. However, when sub-zero air and BB of high temperature and humidity are mixed, and a temperature or phase change in the gas occurs and lasts for a long time, the cold wall causes icing on the intake pipe wall inside the surface. The tetrahedral mesh of the walls and fluid region of the analysis system were created using the commercial software ANSYS-R19.0. In the CFD analysis, the mesh type was selected by considering the setup time, computational expense, and computational diffusion. The shape is relatively simple, but a tetrahedral mesh was chosen because the breather hose was inserted into the intake pipe. Three layers of inflation were placed on the wall surface and the tetrahedral mesh element and maximum sizes were set to 6 mm and 9 mm, respectively. The mesh metric employs orthogonal quality to determine the time of the flow analysis, and the range for the orthogonal quality is 0–1, where a value of 0 is the worst and a value of 1 is the best. A relatively high quality was observed at a minimum value of 0.163 and average value of 0.75. For the analysis model, k-epsilon was selected. The inlet condition required for each driving area was referred to as the vehicle test sheet value, and the wall temperature was the measured experimental value. The calculation time reflected the actual driving time of 30 min, and the value of the analyzed result was adopted for the FENSAP-Droplet and ICE. The FENSAP-Droplet and ICE processes are described in our previous published paper [28].

2.4. Tunnel Test with Chassis Dynamometer

Figure 4 illustrates a schematic of the wind tunnel test utilized to obtain data of icing amounts generated in the intake system. The tunnel tests were performed at Denso Korea. The purpose of the wind tunnel test was to evaluate condensate and ice generation accurately by adopting a constant velocity and adjusting the relative humidity and outside air temperature.

Table 3. Specifications of wind tunnel test facility.

Item	Range
Wind speed	0~160 km/h
Temperature	243~328 K
Humidity	30~80%
Dynamometer	4 WD, 300 kW
Evaluation criteria	Driving evaluation of low temperature

represents the parameters adopted in the wind tunnel experiments. The maximum wind velocity was 160 km/h, which was assessed under outside air temperature and sub-zero conditions. The wind speed and engine operating point were kept consistent by setting and testing the wind velocity at the same level as the wind tunnel test point velocity.

The relative humidity was maintained at 65%, the tunnel test operation point was 1750 rpm at 8 bar, and the external suction temperature was tested at 253 K.

BB and low-temperature air, which were released at high temperatures, were mixed in the mixing zone, causing a phase change from gas to liquid and development of ice on the sub-zero wall. The input data from the vehicle test were used to confirm mixing zone icing. These phenomena were compared to CFD results.

3. Results and Discussion

3.1. Validation Result

Figure 5 illustrates the rpm and load conditions for the analysis operation point and amount of icing obtained via comparison of the wind tunnel test and CFD simulation at 1500 rpm, 7 bar and 1750 rpm, and 8 bar conditions. Case 1 was frequently utilized for long-distance driving in the mid–high velocity range. Case 2 was selected to check the effect of inlet air on icing generation because it had a different rpm but similar load points.

The amounts of ice generated when the PTC heater was on or off were compared at 1500 rpm. In addition, at 1750 rpm, a comparison was performed between the quantity of ice created throughout tests during the 30 and 60 min. A normalizing process of the amount of ice was necessary because the findings of the wind tunnel test could not be fully opened due to a non-disclosure agreement (NDA) with the original equipment manufacturer (OEM). When the PTC heater was on at 1500 rpm, the amount of icing decreased, and the icing generated in the mixing zone owing to the heat conduction of the breather hose was resolved. It was determined that icing in the mixing zone was resolved, and the amount of icing had decreased. The quantity of icing more than doubled when the time was doubled, as determined by comparing the amount at 1750 rpm with the test time. According to the plot for modeling reliability in Figure 4, it was established that there were similarities between the trends of the values obtained from the CFD simulation and the results of the wind tunnel test.

The test and simulation results differed at 1500 rpm for 30 min and 1750 rpm for 60 min due to the on and off limit of PTC heater temperature. To be more specific, the PTC heater is equipped with an automatic on/off feature that shuts it off when the temperature hits 393 K.

The temperature was found to have been maintained because the CFD domain to describe the PTC heater is set at a fixed temperature of 393 K. The interpretation showed the variation in the quantity of icing.

3.2. Icing Phenomenon in Case 1 (Four Working Points)

A comparison of the expected quantity of icing at four different points at 1750–2000 rpm based on the ambient temperature and humidity is shown in Figure 6. The Y axis is normalized to the quantity of icing, while the X axis shows 65–85% humidity at 10% intervals. The level of icing in the vehicle test determined the normalization point (N.P.). At location P1 as marked in Figure 5, the ambient air temperature was 253 K with a 65% humidity level. It was confirmed that the higher the rpm load, the lower the total amount of icing; moreover, the higher the humidity, the greater the amount of icing. The effect of temperature on ice formation also decreased.

Although P1 and P2 had the same rpm, it was confirmed that the deviation in the amount of icing due to temperature was significantly reduced. Regarding P1, temperature had a significant effect on icing, i.e., the higher the relative humidity, the greater the amount of icing. However, the amount of icing in P2 decreased at 258 K and 253 K compared with that in P1. This was because of the increased load, amounts of intake air, and BB released, and the BB temperature. Therefore, compared with that in P1, the amount of icing in P2 increased at 268 K, owing to the temperature difference between air and BB. However, it appeared to decrease at 253 K because the inflow velocity increased and icing time was insufficient. When comparing the effects of temperature and relative humidity on the P2 data, the amount of icing increased because of temperature changes rather than relative humidity. For P3 and P4, the amount of icing increased as the load increased at 2000 rpm. The amount of icing increased as the outside air temperature decreased, and the effect of relative humidity was insignificant. Comparing the tendency of the amount of icing owing to the change in rpm, the amount of icing at 2000 rpm decreased compared with that at 1750 rpm. This decrease is attributed to the increase in flow rate owing to the increase in the amount of intake air. There was also insufficient time for phase change between the two gases and icing formation.

Figure 7 illustrates the amount of icing according to the humidity and inlet temperature at each point as a 2D contour map. The numerical values in the contour map are the predicted icing values converted into standardized values. Compared with P1, the overall amount of ice in P2–P4 decreased and the amount of icing increased owing to temperature changes, instead of the effect of relative humidity. In addition, when the inflow air velocity exceeded a certain level, the amount of icing decreased.

3.3. Icing Phenomenon in Case 2 (Three Working Points)

Figure 8 illustrates that the amount of icing altered according to the inlet air velocity. The X axis represents 65–85% humidity at 10% intervals, whereas the Y axis is normalized to the amount of icing.

At location P1 as marked in Figure 5, the ambient air temperature was 253 K with a 65% humidity level. Although the rpm of the four points varied, the load part was similar, and the amount of incoming air increased toward P7. The amounts of air at P1 and P5 were similar. Accordingly, it was confirmed that the tendencies of the amount of icing according to the relative humidity and temperature change were similar. Compared with P5, the amount of air flowing into P6 and P7 increased sharply, and the velocity increased, which tended to reduce the amount of icing. As the velocity increased, the effects of changes in temperature and relative humidity on icing decreased, and the density of icing due to temperature changes increased.

Figure 9 illustrates the amount of icing according to the change in inlet air velocity as a 2D contour map. The variation in the amount of icing owing to the change in temperature and humidity was the largest at P1 and P5. The red region at P5 was more widely distributed than that at P1. These results indicate that the amount of icing increased slightly owing to an increase in the amount of BB released at high temperatures. P6 and P7 were faster than P1 and P5 because of the increased airflow; however, the BB emission was similar to that of P5. Icing generation was phase-changed by the temperature difference between the two gases and generated condensed water, whereas the generated condensate grew through the cold wall. As the mixing velocity between the two gases increased, the amount of icing decreased because of insufficient phase change and time to collide with the wall. Moreover, the amount of BB released decreased slightly, compared with the amount of intake air.

3.4. Comparison of Inlet Air Velocity and BB Velocity

A contour showing a comparison of the intake air and BB velocities at P1, P3, P5, and P7 is shown in Figure 10. The BB passed to the mixing zone via the breather hose and S-pipe. It should be mentioned that the breather hose’s diameter was lower than the S-pipe’s, and the BB’s velocity momentarily increased.

The reference point in Section 3.4 is P5. The following comparison groups were selected to identify factors affecting icing.

-

Comparison 1 (P5–P1): impact on similar input values (identifying icing tendency).

-

Comparison 2 (P5–P3): same rpm but with the same effect on the load (correlation between the amount of intake air and amount of released BB).

-

Comparison 3 (P5–P7): Same load but influence on rpm (emitted BB is the same but influence on intake air volume).

The data comparing the intake air and BB velocities are shown in

Table 4. Inlet air and blow-by gas velocity at each operation points.

Section (Line)	Operation Point Velocity [m/s]
	P1	P3	P5	P7
IAV ()	7.5	14	7.5	12
EBV ()	6	9	6.5	6.5

. The average velocity of the points on the contour is shown in Figure 10. The intake air and BB velocity are shown by the red and black lines, respectively. From Comparison 1, if the input values and icing amount are similar, then the tendency according to the temperature and relative humidity is similarly derived. The amount of the two gases increased in Comparison 2; however, the amount of intake air increased sharply, velocity increased, and amount of icing decreased. As the velocity increased above a certain level, the effects on the temperature and relative humidity decreased. Comparison 3 compared the effects of intake air; if the same BBs were released, the amount of icing decreased. Figure 11 illustrates the velocity of the intake air from the vehicle’s engine RPM load on a two-dimensional (2D) contour map. In the 2D contour map, we confirmed that the main factors affecting icing differed in the low, middle, and high areas.

In the middle area, the inlet air and BB discharge velocities were similar, and the outside air temperature and relative humidity affected icing. In particular, it was significantly affected by the outside air temperature, and the influence of the relative humidity was insignificant, because the sub-zero air has a specific humidity. At an outside air temperature of 253 K, the specific humidity at 65% relative humidity was 0.41 g/kg dry air and that at 85% was 0.54 g/kg dry air. The relative humidity differed by 20%, but the specific humidity was similar to 0.13 g/kg dry air. Therefore, the outside air temperature affected the increase in the amount of icing in the middle area. In the high area, the inlet air velocity was higher than BB discharge velocity, and the inflow rate of air affected icing. More sub-zero air flowed in than the discharged BB, which decreased the icing amount due to the increase in velocity. The increase in inlet air velocity resulted in a large amount of sub-zero air inflow, and the hot and humid BB did not affect the gas mixing of the entire intake system.

4. Conclusions

The input air velocity and the amount of icing that varied in the high-velocity region were predicted by this study. The amount of icing under unfavorable conditions was numerically anticipated in the engine icing analysis: for each operating condition, the outside air temperature was adjusted to 253–268 K at 5 K intervals and the relative humidity was set to 65–85% at 10% intervals.

The four points (P1–P4) in Case 1 are the points that were often applied during high-velocity driving, and the three points (P5–P7) in Case 2 are the points that compared the icing generated by the inflow velocity. The predicted amount of ice was within an acceptable range in the engine intake system, and the amount of ice generated in the vehicle was confirmed for various driving ranges. As a result of performing the analysis at seven operating points, the effects of the outside air temperature and relative humidity were compared as factors affecting icing. However, the amount of icing tended to decrease toward the high-velocity and high-load areas. Therefore, as the inflow velocity increased, the effect of the outside air temperature and relative humidity on icing decreased. In addition, the effect of the inflow velocity was judged to be the largest. In this study, the factors that influence the formation of icing in the intake system of a vehicle were affected in the order of inflow velocity, temperature, and relative humidity.

(1)

As the load increased, the rpm remained the same, but the amount of air flowing in and the amount of BB released increased. It was confirmed that the amount of icing increased to 268–260 K but decreased at 260–253 K, owing to the combined effects of temperature, relative humidity, and inflow rate.

(2)

The values of P2, P3, and P4 inputs were similar, and the 2D contour colors were similar. The influence of intake temperature change and relative humidity decreased as velocity increased.

(3)

A comparison of the amount of icing at the same load point indicated that the amount of icing was affected by the flow rates of the intake air and BB. P1 and P5, which had similar intake air and BB flow rates, had similar 2D contour color distributions. P6 and P7, which had an increased airflow rate, exhibited reduced icing. Despite the difference in the input value, when the velocity according to the inhaled airflow rate exceeded a certain level, the effects of temperature change and relative humidity decreased, and the amount of icing was similar.

(4)

The main variable affecting icing in the low-velocity region was the discharged BB velocity. Because the BB velocity released from the inlet air velocity was greater, the wall was hit, and icing increased. The main variables affecting icing in the middle-velocity region were the intake air temperature and relative humidity. It was demonstrated that similar air and BB velocities improved mixing and reduced the amount of icing. However, the lower the temperature and the higher the relative humidity, the greater the amount of icing. The main variable affecting icing in the high-velocity region was the inlet air velocity. When the inlet air velocity exceeded a certain level, the effects of BB velocity, temperature, and relative humidity decreased, and the amount of icing decreased.