Active Control of Cylinder Charge Motion Using Vortex Generating Jets (VGJs) on Generic Intake Port Geometries.
Modern engine combustion systems need a perfectly adjusted charge motion to achieve optimal performance. An increased swirl flow at low engine speed can reduce the exhaust smoke opacity of diesel engines by improving the fuel-air mixing quality . An increased swirl can enhance the burning rate and thus improve the engine efficiency on gasoline and diesel engines. However, too strong a swirl leads to rising charge exchange- and wall heat losses and the so-called overswirl-effect may occur on diesel engines . A control of swirl ratio according to engine operating conditions is therefore necessary for reducing exhaust gas emissions and improving engine performance.
Swirl flows are generally created on engines within such intake ports that are characterized by a helical shape or tangential position on the cylinder head . The helical port forces the intake airflow to rotate around the valve axis in the intake ports so that the air streams into cylinder in the direction of desired swirl. The tangential port guides the airflow at a flat angle tangentially into the cylinder and imparts a rotational motion to the cylinder mass. It is also possible to generate swirl using asymmetrical valve seat maskings which enable high swirl ratios at low valve lifts. In addition, swirl-flaps have been featured on some engines to enhance the swirl intensities by closing one of the intake ports. However, such features introduce extra pressure losses to the engine air path, constricting the flow rate and raising pumping losses as well as fuel consumption.
In the present article a further possibility to generate and regulate the swirl flow is studied. The angular momentum of the intake air is not as usual given by the helical or tangential intake port geometries but produced from air jets or the so-called Vortex Generating Jets (VGJs) that inject air at a certain angle directly into the intake ports and create air vortices by this means in the ports or near the valve seat area. The idea of such active flow controls derives initially from the aircraft sectors and goes back to some 90 years ago . The primary use of VGJs for aircraft applications is to improve the aerodynamic performance by generating longitudinal streamwise vortices in the boundary layer preventing flow separation on the aircraft wings . In the last decades the issue about active flow control has been investigated intensively. It has been reported that VGJs generates vortices which suppress flow separations and gain impressive lift . In the present contribution the idea of controlling airflow using VGJs was transferred into the automotive sector with the aim of controlling air motion in cylinder actively while reducing charge exchange losses and thus fuel consumption. The effect of air injection through VGJs was studied by means of experimental investigations on two sets of generically built intake ports in view of swirl generation and airflow optimization. It is noted that simulation of VGJs with e.g. steady-state RANS-simulations, is not necessarily straightforward [7, 8]. Therefore, in the present study, we focus on experimental results.
Flow Test Bench and Characteristic Parameters
A stationary flow test bench following the concept of Tippelmann  was employed to execute flow measurements. The test bench is established in the combustion process development as a standard method to assess the flow rate and swirl characteristics of intake ports. Figure 1 shows the design of the test bench with measurement accuracies listed in Table 1. The test intake ports are mounted on the measuring cylinder using an adapter. Air is sucked by a blower from the ambient through the intake ports and measuring cylinder under a given differential pressure of 50 mbar. A micrometer screw is used to set the desired valve lift manually. The resulting mass flow rate is measured by an impact pressure flow meter. The environmental pressure and temperature are also recorded during the measurement. Basing on the measured values the flow characteristic of the ports can be characterized by the effective flow area [A.sub.eff] defined as the flow area of an ideal nozzle with adiabatic flow using the same pressure head and giving the same mass flow rate:
[mathematical expression not reproducible]
where [p.sub.1] is the pressure before entering the nozzle (total pressure), [p.sub.2] is the pressure behind the nozzle (static pressure), [[rho].sub.1] is the density of gas entering the nozzle, m is the mass flow and [gamma] is the specific heat ratio.
Furthermore the discharge coefficient is introduced to evaluate the flow rate characteristics of different investigated intake ports depending on valve lift and defined as ratio of effective area and geometrical valve seat area:
[C.sub.d] = [[A.sub.eff]*100%/[d.sup.2.sub.valve]*[[pi]/4]]
where [d.sub.valve] is the valve seat diameter.
A flow straightener is installed downstream of the test intake ports in the measuring cylinder with a given distance to the cylinder head. Since this distance has been observed as a sensitive influencing parameter to the results , it is kept constant at 73 mm for all cases to ensure the comparability between diverse test intake ports. The flow rectifier turns the rotating flow into the direction of the cylinder axis and measures the angular momentum acting on it with a torque cell. Under the assumption that the swirling motion in cylinder has a solid body rotation, the strength of swirl induced by intake ports is assessed by a dimensionless swirl number introduced by Tippelmann :
where M is the measured torque around the axis of the cylinder liner, [R.sub.cyl] is the radius of the cylinder, [[rho].sub.z] is the density of the gas inside cylinder and m is, again, the measured mass flow through the inlet valves.
The swirl numbers measured by a paddle wheel anemometer and by such a flow rectifiers are in principle equivalent and can be derived from each other according to equation:
[[c.sub.u]/[c.sub.ax]]=2* [pi]* D
where [c.sub.u] is the angular velocity and [c.sub.ax] is the axial velocity.
In addition the test bench is equipped with extra air injection devices as shown in Figure 2 that comprised VGJs assembled in the intake ports with Jet-diameters from 1.8 to 3.4 mm, PUN-compressed air tubing with Festo-adapters to simplify the installation of different air injection variants, a Sensysflow-air mass flowmeter from ABB with measuring range from 0 to 60 kg/h ([+ or -]1 %), temperature ([+ or -]0.1 [degrees]C) and pressure sensors up to 2.5 bar ([+ or -] 0.0375 bar), a pressure regulator as well as a central compressed air supply up to 10 bar.
Parameter Study to Determine VGJs Configuration
Injection Air Flow Rate
At first, the influence of air injection using VGJs was studied on a set of low-swirl intake ports derived from a preliminary study . The ports consist of two simple intake geometries defined by two smooth center splines and conical surfaces (see Figure 3 left). Two VGJs are equipped on each of the upper and lower sides of the ports respectively, as shown in Figure 3 right, at a reference position perpendicular to the airflow direction and at a flat angle of 25[degrees] to the circumference of the port cross section, which has been proved in the preliminary study as an efficient injection angle to generate swirl. A medium jet-diameter of 2.6 mm is used in this case as a compromise between adequate injecting airflow and acceptable pressure loss on the jets. A test program is carried out, varying the valve lift of both inlet valves from 2 to 10 mm simultaneously and the injection airflow from 0 % up to ca. 15 % of the total intake airflow with all four VGJs blowing.
Figure 4 top shows the swirl numbers as a function of valve lift and injection air proportion (defined as ratio of injection airflow and total intake airflow). The effect of air injection through VGJs can be seen here clearly: the swirl number grows generally with increasing injection air proportion and reaches its maximum at an air proportion of approx. 8 % depending on valve lift. The absolute maximum appears at a valve lift between 4 and 5 mm. At higher valve lifts the maximal swirl number drops slightly but is, nevertheless, still kept on a related high level. This approach demonstrates the possibility to use air injection as an extra variability to control swirl. It must be mentioned that the energy used to compress air has to be supplied by an external power source, e.g. by the engine power on board. As shown in Figure 4 bottom, the pressure loss at the VGJs increases, as expected, with rising injection airflow and increasing valve lift. This energy requirement as well as the resulting fuel consumption for compressing air will be considered in a following section more closely using engine simulation models.
VGJ Position on Intake Ports
In the tests above, the VGJs were located at the reference position at a relative long distance away from the cylinder inlet, in accordance with the idea, that the long distance and more mixing time could benefit the formation of stable swirling motion in the ports. However, it involves also rising air drag that would have a negative effect on the swirl formation and energy efficiency. In order to determine an efficient injection position, tests have also been carried out with VGJs close to the valve seats, as shown in Figure 5.
Figure 6 shows the results of a test, in which only one valve was activated and its lift was varied from 4 to 8 mm with one VGJ injecting close to the valve seat. The other valve was kept closed during the test, so that the interaction between the two inlet flows is avoided and the effect of swirl formation can be illustrated clearly.
The VGJ-position near valve seat shows greater potential, as an almost monotonous rising swirl growth, attributed to the air injection close to valve seat, can be observed with increasing injection airflow at the valve lifts of 6 and 8 mm, i.e. the swirl generated by a VGJ close to the valve seat is nearly twice as much as that from ref-position in high air injection area. In low air injection area (<ca. 3 %), both positions show similar swirl growth, while at the valve lift of 4 mm the ones with valve seat near position are still larger than the ones from ref-position.
Furthermore, it is worth to note, that the air injection close to the valve seat leads to flow rate improvement in middle or high air injection areas, whereas the ref-position rather impairs the flow rate behavior slightly. In this sense, the VGJ-position close to valve seat was implemented for further investigations.
VGJ Diameter (Efficiency Analysis)
The desired angular momentum that is contributed by the VGJs is considered to be dependent on the product of injection flow velocity and mass flow. A larger jet-diameter may reduce the flow velocity but raise the mass flow rate and vice versa. As shown in Figure 4, the pressure losses on the VGJs increase with higher valve lifts and stronger air injection. The energy expenditure is lastly determined by the jet characteristics and the effectiveness with which the momentum of injection air is converted into angular momentum of the intake airflow. In the present case, various sizes of VGJs with diameters from 2.2 to 3.4 mm were assessed in terms of energy requirement for swirl generation based on the above determined VGJ position and under the same test conditions with one valve activated. The injection airflow was varied up to ca. 15 % of the total intake airflow for each VGJ variation at the representative valve lifts of 4 mm, 6 mm as well as 8 mm.
For a clear illustration of the effectiveness of the tested VGJs, the required injection airflow rate and pressure ratio are calculated for the given target swirl growth of [DELTA][c.sub.u]/[c.sub.ax] = 0.3, 0.5, 0.8 and 1.2 respectively (Figure 7). The resulting power requirements of the compression air are calculated as follows :
[P.sub.c] = [??]*[DELTA][h.sub.s]*[1/[[eta].sub.c]]
[DELTA][h.sub.s] = [c.sub.p]*[T.sub.total,in]*[[([P.sub.R]).sup.[[gamma]-1/[gamma]]-1]
where [P.sub.c] is the compression power, [DELTA][h.sub.s] is the isentropic enthalpy change, [[eta].sub.c] is the compressor isentropic efficiency, which is assumed here constant as 70 %, [T.sub.total,in] is the inlet total temperature, which is assumed here as the ambient air temperature of ca. 23 [degrees]C and [P.sub.R] is the pressure ratio. Figure 7 shows that the VGJs with larger diameters demand generally higher injection airflows (see Figure 7 left) but lower pressure ratios (see Figure 7 middle) and thus lower isentropic enthalpy changes to achieve the given target swirl growths. These both effects almost compensate each other at the valve lifts of 4 and 6 mm, so that the energy consumptions for compression air, as shown in the right part of Figure 7, seem not to be much different from each other up to 6 mm valve lift. However, at the valve lift of 8 mm, a considerable advantage of large jet diameter can be observed in view of energy consumption, as the small ones due to the significant rising pressure ratios require almost double the amount of compressing power as the large ones. It should also be noted, that a strong swirl growth of e.g. [DELTA][c.sub.u]/[c.sub.ax] = 0.8 or 1.2 is no more possible with the small ones (indicated by the missing data in the diagrams), since the required injection air pressures exceed the system limit. In sum, the diameters 3.1 and 3.4 mm are shown to be appropriate in terms of swirl efficiency and have therefore been used for further investigations.
Methodology for the Generation of Generic Intake Ports
The approaches above have proven the feasibility of generating swirls by means of air injection via VGJs. However, the effectiveness of air injection, especially in comparison to common swirl intake ports (with no VGJs), is yet to be determined. This issue will be discussed in the following section by analyzing air injection based on realistic intake port geometries. In doing this, a set of such intake ports was of interest, which represents the typical features of its kind, serves as the reference case for comparisons and can be used as a basis for the layout of the so-called jet-ports, on which the VGJ investigations were carried out later. Some reference ports can be found in the literatures; however the detailed geometries are mostly proprietary. Furthermore, it would be an advantage for the comparison purpose, if diverse desired intake port characteristics can be achieved by adjusting single relevant layout parameters. In this sense, a universal methodology is developed which is generally applicable to construct helical and tangential intake ports with usual forms. The methodology depicts a generic port with about 20 parameters and, more importantly, it enables the derivation of comparable intake ports with desired characteristics.
The shape and position of the ports such as helical ports, filling ports or tangential ports determine the swirl intensity of intake ports essentially and can be mainly characterized by the following features, as shown in Figure 8:
* Scroll angle of the valve seat (Figure 8 upper left), which turns the inflow direction of intake air and forces the air flow rotating around the valve stem and entering in cylinder with desired swirl flow field, however, accompanied by noticeable pressure losses.
* Offset of port inlet (Figure 8 bottom left), which is used on some filling ports, leading the intake airflow further in direction towards the center of cylinder head and thus compensate the undesirable counter-rotating flow motion via asymmetric inflow, accompanied by marginal pressure losses.
* Rotation of the valve pattern (Figure 8 upper right) to the longitudinal axis of the engine, which enhances the charge motion around the vertical cylinder axis by increased asymmetric and tangential inflow into cylinder.
* Port slope (Figure 8 bottom right), that affects the outflow angle and thus the tangential part of intake airflow. A flat slope angle generally increases the swirl and reduces flow rate and vice versa.
In order to realize the features above, following rules are laid down as a universal method to conceive the generic port:
* A port is composed of four edges, i.e. two upper edges (shown as solid lines in Figure 9) and two bottom edges (shown as the dashed lines). The outer and inner edges in top view are marked with red and blue respectively for better illustration.
* The contour of the port is defined with the 4 support points distributed around the valve seat projection at 0[degrees], 90[degrees], 180[degrees], 270[degrees] and the two points at port inlet, as shown in Figure 10 top. An extra point (in grey) can be settled on each edge as required to fine-tune the curvature.
* Cubic splines are used to join the support points into the port edges, i.e. the red and blue curves shown in Figure 10, ensuring the desired smooth curvature changing. Tensions at support points are set at a default value (of 1).
* The scroll angle of the straight port shown in Figures 9 and 10 top is defined as 0[degrees]. Rotating the scroll angle for example from 0[degrees] to 45[degrees] in this case, the red and blue support points move in clockwise direction to the positions shown in Figure 10 bottom. The port edges change correspondingly.
* The same rules are complied with for both upper and bottom edges in case of slight or moderate scroll angles.
Figure 11 shows the case of a strong scroll angle on port bottom edges. The scroll angle is rotated here from 0[degrees] to 270[degrees] as an example (see Figure 11 top). The red and blue support points are rotated by 270[degrees] to the positions as shown in Figure 11 middle. This brings about changes especially on the bottom inner edge (dashed blue). As a result the cross-section of the port is constricted and the port bottom contour shows a more strongly helical form (Figure 11 bottom).
It should be noted here, that the spline (thin dashed blue curve in Figure 11 middle), determined by the inner inlet point and the valve guide radius, is defined as the limit for minimum cross-section to ensure a realistic port form. Laying the bottom inner edge after rotation above it, this spline will be considered as the port edge instead.
Some intake ports deflect the airflow smoothly away from the valve guides to reduce the flow disturbance and thus energy losses. Considering the influence of the valve guide on the shape of realistic intake ports, the construction rules are supplemented for the upper contour as follows:
* The flow diversion in valve guide area is defined by the valve guide radius and an additional support point (marked in green in Figure 12). Figure 12 top shows the resulting gradual flow diversion area around the valve guide in case of a straight port, i.e. scroll angle 0[degrees].
* Rotating the valve seat through more than 180[degrees] in clockwise direction, the inner upper edge (in solid blue) is, as compared to the bottom one (in dashed blue), rotates through the support point at 0[degrees], as shown in Figure 12 middle. The inner half of the port (blue hatched area) overlaps in this case virtually in the other part (red hatched area) after rotation. The contour is now shaped mainly by the form of the outer half and the cross-section is further constricted. As a result, a strong helical formed port upper contour is generated as shown in Figure 12 bottom.
The effects of port inlet offset as well as Rotation of valve pattern can be taken into account by shifting the two port inlet points and the valve seat center, as illustrated in Figure 13. Figure 14 shows the definition of port slope on port upper edges. It is assumed, that both upper and bottom edges in unfolded state have constant slopes to simplify the structure and reduce the number of parameters. The transition from port to valve seat is rounded to avoid flow separation due to sharp borders.
Other parameters to define the port geometry such as length, width and height of the port inlet, valve seat diameter etc. as well as a detailed sketch of an intake port are given in the Appendix as an example.
Construction of Generic Intake Ports (Ref. Port and Jet-Port)
By using the methodology above, one set of generic intake ports (referred to as "ref. ports" in following) was designed, composed of a swirl- and a filling port. The geometrical parameters are laid down with reference to existing production cylinder heads [12, 13, 14] and some references [1, 3, 15] to conceive a cylinder head possessing a swirl character of state of the art (cu/ca in the range of 2 ~ 3). To deal with this requirement on swirl, both ports are constructed with helical forms in the valve areas with a scroll angle of 270[degrees] and 30[degrees] respectively. In addition to its strong helix shape, the swirl intensity of the swirl port was further enhanced by being positioned tangentially to the cylinder circumference. On the filling port, the inlet was offset by 25 mm to compensate the undesired counter-rotating flow motion, resulting in a neutral or slight positive swirl performance. Furthermore, the valve pattern was twisted by 10[degrees] in the swirl direction benefiting the swirl generation of both ports. Detailed parameters can be found in the Appendix.
Figure 15 shows the layout of the ref-ports with significant helical shapes that indicate strong swirl ratios but impaired flow performance. Since the swirl requirement is supposed to be met by air injection via VGJs in the following VGJ investigations, the requirement on the swirl ability of the port itself is reduced, which makes it possible to layout the intake ports with focus on better flow performance. In this context, a second set of intake ports (referred to as "jet ports") was derived from the geometry of the ref ports using the layout methodology mentioned above. The main features of the ref-and jet-ports are summarized in Table 2.
As illustrated in Figure 16, the single geometric differences between the ref-and jet-ports are the scroll angles of the valve seats, whereby the ones of jet-ports, by contrast with the ref ports, are rotated from 270[degrees] and 30[degrees] back to the 0[degrees] positions. All other port parameters remain unchanged. Figure 17 shows the cross sections on slice planes perpendicular to the port centerline along the port length. The minimal cross sections are enlarged especially on the swirl port from 295.30 to 588.60 [mm.sup.2], leading to improved flow properties. However, reduced swirl intensities are to be expected, according to the well-known trade-off relationship.
The intake ports are conceived for a cylinder head with 81 mm bore based on a 2.0 l class diesel engine. Both sets of ports, shown in Figure 16, are subsequently verified and optimized in 3D-Simulation to avoid high local pressure drop caused by abrupt cross section changes and to reduce the flow disturbance and associated energy losses in the valve guide areas. The intake port flow models were generated by additive manufacturing using the polyjet-technique (printer of type Objet Eden 350, nominal accuracy [[sigma].sub.x] = [[sigma].sub.y] [approximately equal to] 16 [micro]m, [[sigma].sub.z] [approximately equal to] 42 [micro]m). After printing, the models were scrubbed from support materials and hardened in special caustic solution. The valve stem guides were finely prepared and assembled with two inlet valves of 25 mm diameter as well as valve springs.
Both models are measured on the flow test bench for a valve lift range from 1 to 10 mm. Measuring the ports individually, the swirl port of the ref-ports generates, as expected, strong swirl intensities (cu/ca) in the entire valve lift range with a maximum value of about 10, which is much higher than the ones of the jet-ports, however, at the cost of flow rates, as shown in Figure 18 top. The maximum discharge coefficient reduces from 0.75 to 0.47. The filling ports of both sets don't differ much from each other, since both ports are just slightly scrolled and possess similar minimum cross sections (see Figure 17 bottom). Testing with both inlet vales operating simultaneously, the results show that the requirements on port features listed in Table 2 are fulfilled. The flow rates ofjet-ports have been improved from 0.54 to 0.66 by reason of the scroll angles rotated to 0[degrees]. By contrast, the jet-ports show declines in swirl intensity of up to [DELTA][c.sub.u]/[c.sub.a] = 2, compared to the ref-ones. In the following section, attempts will be described to raise the swirl intensities of the jet-ports to the level of ref-ones by using air injection via VGJs. The efficiency of air injection and its influence on engine operations will also be analyzed.
Test with Air Injection via VGJs
Basing on the geometries shown in Figure 16 right, VGJ are installed on the jet ports in the valve seat areas injecting compressed air directly into the ports with the aim of raising the swirl intensity onto the level of the ref-ports. The VGJs were set, as in the previous approaches, perpendicular to the airflow direction and at a flat angle of 25[degrees] to the port circumference. A jet diameter of 3.4 mm was implemented based on the experience gained in previous tests (see section VGJ diameter).
Three injection concepts are tested to determine the optimal arrangement of VGJs. As shown in Figure 19, the tests are carried out with air injecting in swirl port, filling port and both ports respectively, under the condition that both inlet valves were operated simultaneously. It is noted that the first concept with air injection in both ports blows highest injection airflow through VGJs; however, the effect on swirl intensity is not increasing to the same extent. One possible reason could be that the two rotating air streams from the inlet valves interact with each other when entering the cylinder, counteracting the effects of the VGJs. The second concept, i.e. injecting air only in the tangentially arranged swirl port, has proven to be the most efficient way in terms of swirl generation.
Figure 20 shows the results of the second concept in the case of 4 to 10 mm valve lifts. The measurements are carried out with injection air proportion varying from 0 % up to approx. 12 %, which is effectively limited by the given maximum rel. Injection pressure of 2 bar on the VGJs. A significant growth in swirl intensity of more than [DELTA][c.sub.u]/[c.sub.[alpha]] = 2 has been observed with rising injection air proportion in range from 0 % to approx. 5 %, while the air flow rates remain almost constant or show slight growth. The maxima are reached by all tested valve lifts with an injection air proportion between 5 % and 6 %. Raising the proportion further above, both swirl and flow characters begin to deteriorate continuously. As the approach shows quite a good potential of swirl generation via air injection on jet-ports, controlling swirl by VGJ as an extra variability as well as raising the swirl intensity onto the level of ref-ports is therefore feasible.
The efficiency of air injection and its influence on engine operations is analyzed by means of a 1-D engine simulation model in the commercial software GT-SUITE. The model depicts the correlations between engine operating parameters and the corresponding effects on engine behaviors. A 2.0 l class passenger car diesel engine with four cylinders and exhaust-gas turbocharging was taken as baseline engine. The burn rates used in this approach were derived from real engine measurements on a comparable research diesel engine for the purpose of predicting engine behavior properly. The turbo-charger was removed from the engine model to neglect the influence of inlet port characteristics on the turbocharging system for a clear comparison. The boost- and exhaust gas pressures were given in the simulation manually and held constant for each case. Valve lift curves of serial product are implemented for inlet and outlet valves to ensure realistic charge exchange procedures. The swirl numbers measured in the previous section were converted into GT-swirl-coefficients [C.sub.s] according to following equation :
[C.sub.s] = [M/[??]*[U.sub.is]*[R.sub.cyl]]
[mathematical expression not reproducible]
where M is, again, the measured torque around axis of cylinder liner, m is the measured mass flow through the inlet valve, [R.sub.cyl] is the radius of the cylinder, [U.sub.is] is the isentropic valve velocity, [P.sub.R] is the absolute pressure ratio (static outlet pressure/total inlet pressure), R is the gas constant, [gamma] is the specific heat ratio and [T.sub.0] is the upstream stagnation temperature.
For simulation with ref-ports, the data of the ports were added into the engine model in lookup tables as a function of valve lifts. The model calculates and integrates the angular momentum generated by the intake air entering the cylinder using data taken from the lookup tables. The angular momentum is applied to the cylinder charge and the swirl velocity is calculated for each time-step in simulation based on combustion chamber geometry, intake airflow and piston speed. As a result, the swirl development can be predicted over the whole cycle. It should be noted, that the detailed geometry of the combustion chamber was not considered for this work, thus the predicted swirl number may not be equivalent to the absolute values; however, it is adequate for the illustration and comparison of the effects of port features and air injections on swirl development. Figure 21 shows the predicted swirl number in case of using ref-ports at 2000 rpm with 7 bar net indicated mean effective pressure (IME[P.sub.n]) over a cycle. In the intake stroke, a high swirl numbers is generated through intake ports. It declines in a later phase by reason of the piston speed slowing down. The point marked in green indicates the swirl intensity at inlet valve closing (IVC) and is considered in following as target swirl value for the swirl generation via VGJ on jet-ports.
Coupling between Measured Data and Engine Model
For simulations with jet-ports equipped with VGJs, 3D-maps were generated for discharge- and GT-swirl-coefficients [C.sub.d] and [C.sub.s] from flow rate and swirl intensity measurements as a function of valve lifts and air injection proportions. Figure 22 illustrates the coupling of the air injection in the engine model. The air injection devices shown in Figure 2 incl. VGJs are built into the model physically according to the dimensions on test bench. Four VGJs are modelled by orifices placed on the intake ports and activated in the inlet valve opening phase. The mass flows through them are calculated in the simulation by giving upstream temperatures and pressures. The air injection proportion, calculated from simulated inlet and jet mass flows, is determined at each time step and the corresponding discharge and swirl coefficients are looked up from the 3D-maps. These values are entered back into the engine model and used to calculate the swirl development over the cycle, as shown in Figure 21. The swirl number at inlet closing of each cycle was led into a PI-controller as actual value and regulated to the given target value, i.e. the swirl number of ref-ports under same operating condition, by adjusting the air injection pressures. As result, the required air injection pressure, mass flow for raising the swirl intensities as well as the influence on engine performance such as brake specific fuel consumption (BSFC) and charge exchange work can be determined.
Evaluation of Energy Demand for Air Injection
The energy consumption in case of using jet- and ref-ports was evaluated in this work for engine operation at 2000 rpm and 3000 rpm with 4 bar to 18 bar IME[P.sub.n] respectively. The jet-ports possess a better flow rate performance due to their straighter port forms, as discussed above, which benefits the charge exchange procedure and thus reduces the fuel consumption for the same indicated power. The blue columns in Figure 23 shows the benefit on fuel consumption by using jet-ports compared with using ref-ports. However, since the swirl in case of the jet-ports is no more generated by the intake port itself, but rather imposed by air injections via VGJs, the extra energy involved for compressing the injection air has to be taken into consideration to evaluate the overall system efficiency. Under the assumption that the compressor has an efficiency of 0.7, the power as well as the fuel consumption required by the compressor can be evaluated. As shown in Figure 24, the engine has to raise its indicated power output by approx.46 W at the 4 bar IME[P.sub.n] operating point to drive the compressor. At 18 bar IME[P.sub.n] up to 106 W are required. This leads to an additional fuel consumption of up to l g/kWh that partially compensate the benefit gained by lower charge exchange (shown as the red columns in Figure 23). The difference between both columns indicates the saving on fuel consumptions by using jet-ports of about 0.7 g/kWh at 4 bar IME[P.sub.n].
Figure 25 shows the energy demands of compressor and engine concerning air injection at a higher engine speed of 3000 rpm. The demands on extra engine power output for compressing air increase to approx. 150 W at partial load and 330 W at high load. However, the better flow feature of the jet-ports contributes to a greater saving of fuel consumption with maximum 4.5 g/kWh at 4 bar IME[P.sub.n] and 1.3 g/kWh at 18 bar IME[P.sub.n] (see Figure 26). As a consequence, the concept of jet-ports with VGJs shows an overall benefit of up to 1.8 g/kWh in comparison with ref-ports.
Direct air injection via VGJs in intake ports has been introduced in this article as a further possibility and variability to control the swirl ratio of the cylinder charge. Feasibility of the approach has been demonstrated based on experiments. Clear dependence of swirl number on air injection can be determined in the test results. The influence of position and diameter of the VGJs on the effectiveness of air injection has also been studied. A VGJ placed in the valve seat area with a diameter of 3.1-3.4 mm turned out to be appropriate in the current case.
As realistic intake ports with typical swirl features are requested for the evaluation and for reasons of comparability, a universal methodology for constructing generic intake ports was conceived. The shape of a port is specified by the layout of its four edges (two upper- and two bottom edges). Rules are laid down for each edge to perform the features of the port, i.e. the scroll angle, the offset of port inlet, the rotation of valve pattern as well as the port slope. This Methodology specifies realistic intake port geometries with about 20 parameters and enables the derivation of comparable intake ports of desirable characteristics.
The implementability of methodology was demonstrated by constructing two sets of generic intake ports. One set of them, i.e. the ref-ports, consisting of a swirl- and a filling port, represents typical swirl features and serves as reference case for comparisons. The other set, i.e. the jet-ports, was derived from geometries of the ref-ports by rotating the scroll angles to 0[degrees], which led to higher flow numbers but reduced swirl numbers.
The jet-ports were equipped with VGJs in the valve seat area. A measuring program has been carried out in dependence of valve lift and injection air proportion. A growth of swirl number greater than [DELTA][C.sub.u]/[C.sub.[alpha]] = 2 has been observed for the tested valve lifts at an injection air proportion in range of 5% and 6%. The feasibility of raising swirl intensity onto the level of ref-ports by using VGJs has been demonstrated.
An engine model was coupled with the measured data to predict the development of swirl intensity over a cycle and to assess the effect of different ports on engine performance. The effectiveness of VGJs was investigated by means of simulations. The results show that the greater flow numbers of the jet-ports reduce the charge exchange work and thus benefit the fuel consumption, however, if raising the swirl number onto the level of ref-ports, the extra power for compressing air partially compensates the fuel consumption saving. With the geometries used herein, an overall saving of up to 0.7 g/kWh and 1.8 g/kWh can be expected at 2000 rpm and 3000 rpm respectively.
In the current study a straightforward port design was used to demonstrate that VGJs in the intake ports can generate the desired swirl and a positive net BSFC-balance. Both, ref-ports and jet-ports could be subject to optimization for a real engine and it is yet to be studied, if the benefit in BSFC can be further increased and how the combustion would be affected by swirls produced in such a different way as well as by different turbulent characteristic on real engines. At this point it is also of interest to establish a CFD-method for further investigation to study the building of swirls generated in such a way and the effect on cylinder internal flow. Nevertheless, the concept introduces an increased variability, since it enables a quick and individual controlling of in-cylinder motions to match the changing engine operating points.
Technische Universitat Braunschweig
Institut fur Verbrennungskraftmaschinen ivb Hermann-Blenk-Str. 42 38108 Braunschweig
VGJ - Vortex Generating Jet
[A.sub.eff] - Effective flow area
[p.sub.1] - Total pressure before entering the nozzle
[p.sub.2] - Static pressure behind the nozzle
[[rho].sub.1] - Density of gas entering the nozzle
[??] - Mass flow
[gamma] - Specific heat ratio
[C.sub.d] - Discharge coefficient
[d.sub.valve] - Valve seat diameter
D - Swirl number introduced by Tippelmann
M - Measured torque around cylinder axis
[R.sub.cyl] - Cylinder radius
[[rho].sub.z] - Density of the gas inside cylinder
[c.sub.u] - Angular velocity
[c.sub.ax] - Axial velocity
[C.sub.s] - Swirl-coefficients defined in GT-SUITE
[U.sub.is] - Isentropic valve velocity
[P.sub.R] - Absolute pressure ratio (static outlet pressure/total inlet pressure)
R - Gas constant
[T.sub.0] - Upstream stagnation temperature
IME[P.sub.n] - Net indicated mean effective pressure
BSFC - Brake specific fuel consumption
IVC - Inlet valve closing
[P.sub.c] - compression power
[DELTA][h.sub.s] - isentropic enthalpy change
[[eta].sub.c] - compressor isentropic efficiency
[1.] Kunstscher, V. and Hoffmann, W., Kraftfahrzeugmotoren (Zwickau: Vogel Buchverlag, 2014).
[2.] Endres, H., Wurms, R., and Neu[beta]er, H.J., Verbrauchs- und Emissionsminderungbei Mehrventil-Ottomotoren durch Anwendung von Drall und Tumble (Aachen: Aachener Kolloquium Fahrzeug- und Motorentechnik, 1991).
[3.] Heywood, J.B., Internal Combustion Engine Fundamentals (New York: McGraw-Hill Book Co, 1988).
[4.] Prandtl, L., "The Generation of Vortices in Fluids of Small Viscosity," Aeronautical Reprints, The Royal Aeronautical Society No. 20, London, 1927.
[5.] Radespiel, R., Scholz, P., and Burnazzi, M., "Active Flow Control for High Lift with Steady Blowing," The Aeronautical Journal 120:1223, January 2016.
[6.] Johnston, J.P. and Nishi, M., "Vortex Generator Jets-Means for Flow Separation Control," AIAA Journal 28:989-994, 1990.
[7.] Mahmood, S., Scholz, P., and Radespiel, R., "Numerical Design of Leading Edge Flow Control over Swept High-Lift Airfoil," Aerotecnica Missili & Spazio, The Journal of Aerospace Science, Technology and Systems 92(1/2):3-16, 2013.
[8.] Mahmood, S. and Radespiel, R., "Detached-Eddy Simulation of Vortex Generator Jet Using Chimera Grids," International Journal of Mechanical and Mechatronics Engineering 5(7), 2011, doi:10.1999/1307-6892/14808.
[9.] Tippelmann, G., "A New Method of Investigation of Swirl Ports," SAE Technical Paper 770404, 1977, doi:10.4271/770404.
[10.] Uzkan, T., Borgnakke, C., and Morel, T., "Characterization of Flow Produced by a High-Swirl Inlet Port," SAE Technical Paper 830266, 1983, doi:10.4271/830266.
[11.] Scholz, P., Francois, D.G., Haubold, S., Sun, S. et al., "WM-LES-Simulation of a Generic Intake Port Geometry," SAE Journal of Engines 11(3), 2018, doi:10.4271/03-11-03-0023.
[12.] Audi AG, "Selbststudienprogramm 420: Audi 2,0l-TDI-Motor mit Common-Rail-Einspritzsystem," Ingolstadt, 2007.
[13.] Volkswagen AG, "Selbststudienprogramm 316: Der 2,0l TDI-Motor," Wolfsburg, 2003.
[14.] Kawashima, J.-I., Ogawa, H., and Tsuru, Y., "Research on a Variable Swirl Intake Port for 4-Valve High-Speed DI Diesel Engines," SAE Technical Paper 982680, 1998, doi:10.4271/982680.
[15.] Frank, W., "Beschreibung von Einlabdrallstromungen fur 4-Takt-Hubkolbenmotoren auf Grundlage stationarer Durchstromversuche," Dissertation, RTWH Aachen, 1985.
[16.] Gamma Technologies, "Engine Performance Application Manual," 2016.
Appendix: Parameter of Ref-Ports Geometry
Name Index Swirl port Filling port Scroll angle [alpha] 270[degrees] 30[degrees] Offset port inlet S 0 mm 25 mm Rotation of valve pattern [beta] 10[degrees] 10[degrees] Port slop upper side [theta] 25[degrees] 30[degrees] Port slop lower side ([phi]) 17.5[degrees] 15[degrees] Port length L 95 mm 85 mm Port width B 25 mm 25 mm Port height H 40 mm 40 mm Valve seat diameter [d.sub.v] 25 mm 25 mm Valve chamber upper diameter [d.sub.c] 28 mm None (optional) Valve chamber eccentricity e 4 mm None (optional) Valve chamber height h 21 mm 21 mm Valve item diameter 14 14 Distance between valve center 1 25 mm 15 mm and point for flow diversion Height of the point for flow j None 44 mm diversion (optional) Distance from upper slope to m 20 mm 20 mm port inlet Distance from bottom slope to n 20 mm 20 mm port inlet Radius for transition between r 15 mm 15 mm port and valve seat Additional point to tune x 25 mm None curvature of upper contour (optional) Additional point to tune y 20 mm None curvature of lower contour (optional)
Shaowei Sun, Peter Eilts, Peter Scholz, and Stefanie Haubold, Technical University of Braunschweig
Sun, S., Eilts, P., Scholz, P., and Haubold, S., "Active Control of Cylinder Charge Motion Using Vortex Generating Jets (VGJs) on Generic Intake Port Geometries," SAE Int. J. Engines 11(4):475-489, 2018, doi:10.4271/03-11-04-0032.
Received: 22 Feb 2018
Revised: 22 May 2018
Accepted: 08 Jun 2018
e-Available: 08 Aug 2018
TABLE 1 Measurement accuracies of flow test bench. Measuring channel Measuring range Accuracy Torque - 0.2 ~ +0.2 [Nm] [+ or -]0.2% FS Mass flow rate 30 ~ 750 [kg/h] [+ or -]1% FS Differential pressure 30 ~ 70 [mbar] [+ or -]0.2% FS Temperature 0 ~ 70 [degrees]C [+ or -]0.1 [degrees]C TABLE 2 Main features of ref-ports and jet-ports. Ref-ports: Jet-ports: * Focus on swirl features * Focus on optimized flow with scroll angles of * Features with scroll angles of 0 270[degrees] and 30[degrees] [degrees] * With VGJ in valve seat area * Without VGJ ([??] 3.1 mm) * Strong swirl number * Low swirl number but high but low flow number flow number * Swirl induced by * Swirl controlled by air injection port geometry via VGJs * No demand on * Extra power required for extra energy compressing air
|Printer friendly Cite/link Email Feedback|
|Author:||Sun, Shaowei; Eilts, Peter; Scholz, Peter; Haubold, Stefanie|
|Publication:||SAE International Journal of Engines|
|Article Type:||Technical report|
|Date:||Oct 1, 2018|
|Previous Article:||Fracture-Splitting Processing Performance Study and Comparison of the C70S6 and 36MnVS4 Connecting Rods.|
|Next Article:||Conditioning Turbocharger Compressor Map Data for Use in Engine Performance Simulation.|