Diesel Fuel injector Nozzle
Design of the diesel fuel injector nozzle is critical to the performance and emissions of modern diesel engines. Some of the important injector nozzle design parameters include details of the injector seat, the injector sac and nozzle hole size and shape. These features not only affect the combustion characteristics of the diesel engine, they can also affect the stability of the emissions and performance over the lifetime of the engine and the mechanical durability of the injector.
All nozzles must produce a fuel spray that meets the requirements of the performance and emissions goals of the market for which the engine is produced regardless of details of the fuel system design (i.e., regardless if the fuel system is of the common rail, unit injector, unit pump or pump-line-nozzle type). Additionally, specific requirements for injection nozzles can also depend on the fuel system type [Potz 2000]:
· Common rail—nozzle operates under more demanding tribological conditions and must be better designed to prevent leakage.
· Unit injector/unit pump—pressure pulsing conditions create more demanding fatigue strength requirements.
· Pump-line-nozzle—hydraulic dead volume must be minimized.
Figure 1. Basic diesel injector nozzle with single cone seat
Figure 1 shows a basic overview of the major components of a diesel fuel injector nozzle [Parrish 1976]. Several of these components are discussed in detail in the following sections. Readers should also review the introduction to injector nozzles that was given under Fuel Injection System Components.
The nozzle seat is the first important element of the injector nozzle. Some requirements of the seat are that it must seal well, maintain a consistent and predictable seating line and help maintain accurate fuel metering over the life of the engine.
Controlling the effects of seat wear on the long term stability of injection quantity and thus emissions over the life of a vehicle is a critical aspect of nozzle seat design. One of the parameters that must be controlled over the life of the vehicle in order to maintain injection quantity is nozzle opening pressure. Depending on how the seating line of the nozzle valve wears and whether it travels towards the tip or away from the tip, nozzle opening pressure can increase or decrease and produce unwanted changes in injection duration and injection quantity. This was explained when discussing injection system components.
While this is an important consideration in spring biased nozzles that are completely actuated by fuel pressure acting on the nozzle seat, it can also be a critical consideration in many common rail systems that use electro hydraulic servo controlled nozzles. Common rail injectors often employ multiple injection strategies where individual injection events such as pilot injections must inject very small quantities of fuel [Catania 2008]. Any variation in these small quantities of fuel can severely compromise engine emissions and performance. In cases where injection quantity is sufficiently small, wear of the nozzle seat can cause the injection event to disappear altogether [Lambert 2009][Mattes 2006]. Tolerances in nozzle opening pressure for common rail injectors for some applications can be as tight as 1% [Farth 2005].
Figure 1 and Figure 2 show several examples of nozzle seat designs that have been used in closed nozzle diesel fuel injectors [Parrish 1976][Schlaf 1998][Lambert 2009]. The nozzle seat comprises the needle tip and the mating surface in the injector body. A simple nozzle seat design (no longer used in modern diesel engines) is a single cone on the end of the needle valve that mates with a conical surface on the injector body, Figure 1. The included angle of the needle valve is larger than that of the mating surface of the body. This forms a seating line at the largest diameter of the needle valve’s conical tip. The problem with this design is that wear and carbon deposits between the conical surfaces can cause the seating line to move toward the tip and increase the effective area acted on by injection pressure. Wear can also cause the spring force holding the needle to its seat to drop. These factors can all contribute to a decrease in nozzle opening pressure that can affect injection duration and for spring biased nozzles, allow hot combustion gas to enter the injector under high load conditions and lead to rapid injector failure [Parrish 1976]. Another disadvantage of the simple conical nozzle is that irregularities in either the needle or body surfaces can cause the seating line to form along a random line of contact that may not be at a constant diameter at the seat edge and lead to a nozzle opening pressure that can vary with the rotational position of the needle valve.
Figure 2. Examples of nozzle seat designs
(a) Double cone; (b) Double cone, grooved
Another nozzle seat design, Figure 2a, consists of a needle valve having two conical surfaces with different included angles (angles B and C). The adjoining peripheral line forms the initial seating surface. If the differential angle of the upper cone is smaller than that of the lower cone (A-B < C-A), wear will cause the valve seating line to move upward and is referred to as a “negative” or “inverted” seat region. This will decrease the effective area acted on by the injection pressure and increase nozzle opening pressure [Parrish 1976][Lambert 2009]. A disadvantage of double cone nozzle seat with an inverted seat region is that sac volume will increase as the injector wears leading to an increased crevice volume for unburned hydrocarbons [Stevens 1999].
If the differential angles are such that that of the lower cone is smaller than that of the upper cone, the seating line will migrate towards the tip of the needle valve and is referred to as a “positive” or “non-inverted” seating region [Lambert 2009]. Since this would lower nozzle opening pressure, the placement of a groove on the lower cone such as in Bosch’s ZHI seat, Figure 2b, can be used to limit the migration toward the tip. This will stabilize the seating line and injection quantities and in spring-biased nozzles, ensure that the nozzle remains closed at all cylinder pressure conditions [Parrish 1979][Stevens 1999][Potz 2000]. A grooved needle can also account for irregularities and ensure a reliable seating line [Schlaf 1998].
Figure 2c shows another nozzle seat design—featuring three conical surfaces with different angles—that is intended to provide a highly accurate and repeatable seating line diameter to be achieved during manufacture. It would be suitable for unit injectors, injectors for unit pump systems and common rail injectors [Limmer 2002].
Seat design can also be used to compensate for the effects of wear on injection quantity in other parts of the system. For example, the seat can be designed to wear in a manner that would decrease injection quantity to compensate for wear in other parts of the injector that may increase injection quantity [Lambert 2009]. Nozzle design features can also be used to prevent cavitation damage due to pressure pulses that occur in some unit injector and unit pump systems [Lambert 2010].
As noted elsewhere, the injector sac can have a significant effect on emissions, nozzle strength and hole-to-hole fuel distribution. The injector nozzle can be a sac-type (blind hole) or use a valve covered orifice (VCO) design, Figure 3 [Potz 2000]. Sac-type nozzles can have either a cylindrical or a conical sac. Injectors for modern diesel engines typically use a conical design. Depending in the dead volume downstream of the nozzle seat, sac-type nozzles can be characterized as mini-sac or micro-sac type. A typical light duty injector with a mini-sac can have a sac volume (including injector holes) of about 0.315 mm3, a micro-sac a volume of about 0.230 mm3, while a typical VCO nozzle has a dead volume of about 0.145 mm3 [Seneschal 2009].
Figure 3. Comparison of mini-sac, micro-sac and VCO type nozzles
The sac-type nozzle has the advantage that the holes can be made to communicate directly with the sac. This ensures the entrance condition at the hole inlet is essentially the same for each hole and the spray distribution among the holes is relatively uniform. As the sac volume in a conical sac nozzle decreases, the outer surface of the needle valve will approach the entrance of the nozzle holes and have an increasingly significant effect on the entrance to the hole. When this occurs, small misalignments in the needle can have increasingly significant impacts on the hole-to-hole flow distribution of the nozzle. Of the three nozzles illustrated in Figure 3, the flow distribution is most uniform in the mini-sac nozzle, and least uniform in the VCO nozzle. An example of the degraded flow distribution from a VCO nozzle is illustrated in Figure 4. Significant improvements in spray distribution from VCO nozzles can be achieved through the use of a double-guided needle [Bae 2002][Dingle 2005].
Figure 4. Initial development of spray from two-spring VCO nozzle
Single-guided needle under atmospheric condition, 0.1 ms after start of injection.
Video 1 illustrates the hole-to-hole variation in nozzle flow from a common rail injector. For this example, x-ray tomographic imaging was used to verify that variations in nozzle hole entrance radius and hole diameter were minimum, Figure 5. Thus the variation shown is due entirely to variations in internal nozzle flow [Busch 2017].
Video 1. Fuel spray from a diesel injector with a pilot/main injection strategy
Pre-production solenoid actuated servo controlled injector with pressure balanced control valve, 800 bar injection pressure, 1.5 mg pilot. 7 holes, 139 µm, 1.5 concentricity, 149° included angle, nozzle volume 0.23 mm3.
Figure 5. Nozzle hole characteristics of injector in Video 1
The effect of sac volume on emissions is illustrated in Figure 6. It is apparent that most of the fuel that dribbles out of the nozzle sac during the expansion stroke contributes to unburned hydrocarbons. However, a significant amount ends up in the soluble organic fraction of diesel PM and some also contributes to soot emissions. The relative contribution to emissions can be greatest at low load where the sac volume is relatively large compared to the injected fuel quantity. However, at low load, nozzle temperatures are lower and all the fuel in the sac may not evaporate and escape from the sac [Eastwood 2008].
Figure 6. Effect of sac volume on hydrocarbon and particulate emissions
In the example of Figure 6, the contribution to diesel PM is minimum at a sac volume of about 0.4 mm3. One possible explanation for this optimum point is the trade-off between hole-to-hole fuel distribution effects and post-injection dribble, Figure 3. At very low sac volumes, poor hole-to-hole distribution of fuel leads to more over-rich regions in the combustion chamber and increased soot production. While hole-to-hole distribution of fuel improves with increased sac volume, leading to reduced soot, the PM contribution from fuel dribble increases.
The injector hole is the final element of the fuel injection system before the fuel enters the combustion chamber. As such, the size and shape of the nozzle hole can play a critical role in the emissions and performance of the diesel engine. Two important objectives of nozzle hole design are ensuring sufficient fuel flow at the available injection pressure to meet maximum torque and power conditions and ensuring sufficient cylinder gas entrainment (air + EGR) into the fuel jet to ensure emission goals are met. Since nozzle hole flow area in modern production engine injector nozzles is fixed, the first objective essentially fixes the nozzle hole flow area (individual hole effective area × number of holes). The second objective considers the effect of hole design on combustion characteristics, specifically the flame lift-off length and the fuel-air ratio at the lift-off length of the fuel jet.
Since fuel injection pressure and rated torque or power determines the required nozzle hole flow area, the next parameter to be chosen is how the effective flow area is distributed. In modern direct injection diesel engines, flow area distribution is usually determined by the number of symmetrically spaced holes. Obviously, for a given effective flow area, the smaller the hole size the higher the number of holes required. One guiding principle in selecting the number of holes is to choose the maximum number of holes with a size that can be economically produced while preventing any overlap of the fuel jets and while meeting nozzle flow requirements. Any jet overlap will form regions with insufficient oxygen and can lead to unburned fuel and/or soot. This explains why injectors for combustion chambers with high swirl tend to have fewer holes than those for quiescent chambers. The tangential air flow in a high swirl chamber requires a greater angular separation of fuel jets to prevent their overlap [Benajes 2006].
The effect of nozzle hole design on the combustion characteristics of the fuel jet can be seen from the equations for lift-off length and the average fuel-air equivalence ratio at the lift-off length. Two important variables in these equations that can be influenced by nozzle design are injector nozzle hole diameter (Do) and how well the momentum of the fuel jet is used to maximize the effective jet velocity (U) at the exit of the nozzle hole (i.e., the momentum efficiency).
Nozzle Hole Size
Injector hole size is perhaps the most obvious injector nozzle design variable that can be used to influence the combustion characteristics of the jet. It can be shown that the average fuel-air equivalence ratio at any axial location well downstream of the injector, φ(x), is proportional to Do. This reflects the variation of the volume to surface area ratio of the jet in the region near the nozzle, ~[π (Do/2)2]/[π Do]. As the hole size is decreased, more surface area for air entrainment relative to jet volume is available and φ(x) becomes leaner. It should be noted that the lift-off length, H, also decreases with Do but by a relatively weak power of 0.34. The net effect is that φ(H) becomes more fuel lean as nozzle hole size decreases, Figure 7 [Pickett 2005].
Figure 7. Effect of orifice diameter on lift-off length and percent of entrained stoichiometric air
ζst = 100/φ(H)
The higher levels of air entrainment prior to the lift-off length (relative to the amount of fuel injected) for smaller orifice diameters leads to a reduction in the total soot production as reflected by the total soot incandescence data of Figure 8 [Pickett 2005]. Figure 8 can be further generalized to show that total soot production can become negligible when φ(H) < ~2. It should be noted that while total soot incandescence is an indication of total soot production, engine out soot levels are also affected by soot oxidation and trends in total soot production may not always be reflective of trends in engine out soot. Conditions leading to little or no soot production, however, will also produce little or no engine out soot.
Figure 8. Effect of orifice diameter on soot incandescence
Fuel flow normalized total soot incandescence of the lift-off length.
While Figure 8 suggests that reducing nozzle hole size to about 50 µm could almost eliminate soot production under some conditions, there are some significant challenges with implementing such small holes in production diesel fuel injectors. Some of these challenges include compensating for a reduced flow area per hole, reduced jet penetration and the difficulty of reliably and economically mass producing small holes.
As hole size decreases, something must be done to maintain the maximum fuel injection rate to ensure maximum engine torque and power objectives and engine efficiency are met. This can be illustrated by an example of what would be required to maintain injection rate when reducing hole size from 170 µm to 100 µm. If the same number of holes is maintained in both nozzles, injection pressure in the 100 µm hole nozzle would have to increase by more than a factor of 8 to maintain the same fuel injection rate. On the other hand, if the injection pressure is maintained, the number of nozzle holes would have to increase by almost a factor of 3 to maintain the same nozzle flow area. Increasing injection pressure by even a more modest 10 or 20% comes at considerable cost. Increasing the number of holes may compromise nozzle integrity and increase the potential for overlapping of the burning zones of individual fuel sprays—thus defeating the reason for making holes smaller in the first place. A third option that can be used to maintain injection rate is to increase the efficiency of the nozzle holes by increasing the discharge coefficient, Cd (defined below). For a given injection pressure and nozzle hole exit size, this approach can improve fuel flow rate by up to about ~20%. This option is discussed in more detail below. Thus, none of the options considered is able to easily compensate for the loss in fuel flow rate with the relatively large reduction in hole size considered in this example. By combining a more modest increase in injection pressure, modest increase in the number of nozzle holes and improvements in Cd, it is possible to realize some reduction in injector hole size but reductions to sizes well below 100 µm where soot production could almost be eliminated remain a considerable challenge for fixed geometry nozzles that must also use the same flow area to achieve rated torque and power.
An additional challenge with smaller holes is that flame length decreases because the jet is decelerated more rapidly by the transfer of momentum from the high velocity fuel exiting the nozzle to a relatively larger amount of air entrained into the jet by the higher jet surface area/volume ratio. A shorter flame length can lead to higher soot emissions at high load due to poor air utilization in the combustion chamber. Figure 9 shows the flame length, HF, normalized by orifice diameter [Pickett 2005]. While there is some increase is relative flame length with smaller nozzle orifice diameters, the absolute length of the flame decreases. Relative flame length for momentum drive jets can be correlated with the following equation:
HF/Do = CH · √(ρf/ρa) · (1 + (A/F)st) · √((Ma · Tad)/(Mprod · Ta))(1)
CH is the flame length coefficient. CH varies approximately linearly with nozzle hole size from ~6 for 50 µm holes to ~4 for 175 µm holes.
Ma and Mprod are the ambient and product gas molecular weights respectively.
With the exception of CH which varies with nozzle hole size as noted above, the terms in Equation (1) are independent of orifice geometry.
Figure 9. Effect of orifice diameter on relative flame length of burning fuel jet
The loss of jet penetration can make it a challenge to utilize all the air in the combustion chamber for a fixed bore diameter without additional measures being taken. One step that can be taken to maintain air utilization is a combustion chamber design change, such as a smaller bowl diameter, to better ensure adequate utilization of the available oxygen during combustion. Injection pressure increases can also be used to offset some of the loss of jet penetration with a smaller hole size. However, the pressure increases required to offset lower jet penetration even for a modest decrease in nozzle hole size can be significant. For example, a 10% reduction in nozzle hole size can require a 23% increase in injection pressure just to maintain jet penetration. Improving nozzle efficiency by increasing Cd can also help improve jet penetration. Group hole nozzles, discussed later, are an interesting development and can also help maintain jet penetration from smaller nozzle holes while providing enhanced air entrainment. As was the case with maintaining injection rate with smaller hole size, penetration decreases with nozzle hole size reductions are not easy to offset. However, by utilizing all of the options available, it is possible to offset potential decreases in jet penetration for modest decreases in nozzle hole size.
In summary, improvements in the combustion characteristics of a burning fuel jet in a diesel engine by reducing nozzle hole size are limited by factors such as the need to maintain engine power and torque, injection pressure limitations, engine bore size and by the need to maintain effective utilization of available oxygen in the combustion process. However, by combining hole size reductions with other changes such as increases in injection pressure, changes in combustion chamber design, improvements in nozzle hole flow performance and more nozzle holes, important improvements in combustion characteristics can be achieved.
Nozzle Hole Momentum Efficiency
The second way to affect the combustion characteristics of the jet via nozzle design is to change the effective jet velocity (U). While injection pressure is perhaps the most straight forward means available to change U, attention to nozzle hole details and a reduction of flow losses through the nozzle hole can also be used to increase U without the need to increase injection pressure. Increases in U of up to about 10-15% are possible by a reduction in losses through the injector nozzle hole. This is a significant increase as injection pressure increases in the range of 20-30% would be required to achieve a similar increase in U if no changes were made to the nozzle hole.
The effect of U on combustion characteristics is relatively straightforward. Increasing U will increase the lift-off length, H, of the jet and provide a longer distance over which air can be entrained into the jet. While the equation for φ(x) is relatively unaffected, φ(H) will be leaner because of the longer value of H. Hole design features that can be utilized to provide improvements in jet velocity are those that increase Cd and include honing the hole to achieve a rounded inlet and conical holes that have a larger entry diameter than exit.
Before discussing these nozzle design features in more detail, it is necessary to define some basic parameters and provide some background material on cavitation in injector nozzle holes first.
Definitions. The definition of U in the equation of lift-off length, H, shows the means for achieving increased U via nozzle design is to affect the ratio Cd/Ca. To understand this term better, it is useful to consider Figure 10 which shows a simplified sketch of fuel flowing through a round sharp edged injector nozzle hole [Desantes 2003].
Figure 10. Schematic of straight edged inlet injector nozzle hole
In a fuel injector nozzle, fuel entering the nozzle hole is driven by a very high pressure differential and may often need to make strong changes in flow direction. As a consequence, there can be a very low pressure at the nozzle inlet and the boundary layer can separate from the nozzle hole surface. The separation forms a “vena contracta”. The static pressure at the throat of the vena contracta will be lowest—in some cases lower than the nozzle hole exit pressure. In cases where this static pressure drops below the vapor pressure of the fuel, cavitation will appear in the nozzle hole. It should be noted that in real fuel injectors, the fuel may enter the nozzle hole asymmetrically, for example all flow lines entering the hole may originate from above the hole. As a result, the vena contracta may be asymmetrical as well.
Some of the coefficients used to characterize the performance of a nozzle hole include, area contraction coefficient (Ca), discharge coefficient (Cd), velocity coefficient (Cv) and momentum efficiency (ηM) [Benajes 2004][Desantes 2003][Siebers 1999][Greeves 2010].
The effective area of the nozzle hole can be less than the actual geometric area because of the vena contracta and/or cavitation effects. The ratio of this area to the geometric area of the hole is the area contraction coefficient:
Ca = A/Ageo(2)
It is also known as the contraction coefficient, Cc.
Due to this area contraction and friction in the nozzle hole, the mass flow for a given pressure drop will be less than the theoretically possible maximum flow rate. The ratio of the actual mass flow of fuel (mf) to the theoretically possible maximum flow rate defines the discharge coefficient:
Cd = mf/(Ageo √(2 · ρf · ΔP))(3)
A third useful coefficient is the velocity coefficient and is the ratio of the effective jet velocity and the maximum theoretical velocity:
Cv = U/Uth(4)
where Uth = √((2 · ΔP)/ρf).
It can be shown that:
Cd = Cv · Ca(5)
It should be noted that while it is common to characterize nozzle holes with the discharge coefficient, Cd, an accurate determination of Cd requires knowledge of the nozzle diameter. For some nozzle hole designs, such as those that have a diameter that varies along the length of the hole, it can be difficult to accurately quantify the diameter. It is thus useful to have a coefficient that can be used to characterize the nozzle hole without the need for accurate knowledge of the nozzle hole geometry. A coefficient that can be used to achieve this objective is the momentum efficiency (ηM). It is the ratio of the actual axial momentum in each fuel jet over the maximum theoretically possible axial momentum for the actual or measured fuel mass flow of the jet [Greeves 2010]:
ηM = M/Mth = (mf · U)/(mf · Uth) = Cv(6)
A typical straight nozzle hole with a sharp edged inlet used as a baseline for many nozzle hole performance comparisons has a Cd ~0.7 and a ηM ~0.82. The key to increasing U and realizing an increase in lift-off length by utilizing nozzle hole design is to increase the value of these coefficients. The potential benefits of doing so are illustrated in Figure 11. In this example, an improvement in nozzle hole momentum efficiency is shown to correlate with a reduction in the net soot emissions as indicated by filter smoke number [Greeves 2010].
Figure 11. Impact of nozzle momentum efficiency on engine-out smoke number
The coefficient of discharge for flow through a non-cavitating nozzle hole can be expressed as [Jung 2008]:
Cd = 1/√(1 + f·L/Do + Kent + Kaux)(7)
f = friction factor
L = nozzle hole length
Kent = entrance loss coefficient
Kaux = adjust Cd for any auxiliary effect through the fuel passage of the injector
This expression is convenient in that it shows which aspects of nozzle design can be used to reduce losses through the nozzle hole. Two important aspects that can be addressed are the losses occurring at the entrance to the hole (Kent) and friction of the hole (f·L/Do). However, before considering these terms in more detail, it must be recognized that cavitation can play an important role in determining the performance of an injector nozzle hole that is not reflected in Equation (7). This is considered next.
Cavitation. One of the main factors affecting nozzle discharge coefficient and the momentum of the injected fuel jet is cavitation in the nozzle hole. This is especially the case in straight round holes with a sharp edged entrance. Geometric-induced cavitation occurs when the local pressure drops below the vapor pressure and a vapor phase forms. When cavitation occurs in the nozzle hole the flow becomes choked.
Figure 12 illustrates what happens to the flow from the fuel injector under cavitating and non-cavitating conditions [Payri 2004]. A nozzle experiencing cavitation is illustrated by the red data points while a nozzle not experiencing cavitation is illustrated by the blue data points. Both nozzles were tested at injection pressures of 10, 20 50 and 80 MPa. For each injection pressure, pressure at the nozzle outlet was varied. With the nozzle experiencing cavitation, it can be seen that at each injection pressure, there is a value of outlet pressure below which no further increases in mass flow occur (horizontal red lines). This is the point at which cavitation starts and flow in the nozzle hole becomes choked. For the non-cavitating nozzle, no such value of outlet pressure is reached and no choking occurs. Under non-cavitating conditions, the flow through the injector nozzle increases in proportion to the square root of the pressure drop across the nozzle. When cavitation occurs and flow becomes choked, the flow increases in proportion to the square root of the nozzle hole inlet pressure.
Figure 12. Total mass flow for fuel injector under cavitating (red) and non-cavitating conditions (blue)
ΔP - pressure drop across the nozzle
Based on the observation that cavitation will occur when the local static pressure of the fuel drops below the fuel’s vapor pressure, it is possible to define a cavitation index [Som 2010]:
KCI = (P - Pb)/(Pb - Pv)(8)
P is the local pressure at any point in the nozzle hole
Pb is the pressure at the nozzle outlet (also referred to as back pressure)
Pv is the vapor pressure of the fuel
When KCI < -1, cavitation will occur. However, making use of Equation (8) is complicated by the fact that it is based on local pressure which is difficult to determine in practice. Application of Equation (8) is thus limited to numerical modeling work where it is possible to calculate the local pressure at every location in the nozzle hole. It should also be noted that in order to accurately identify locations where cavitation may occur, Equation (8) should also include a term to account for stress arising from dynamic and turbulent viscosity [Som 2010].
As a result of the difficulty of applying Equation (8), it is more common to use a cavitation number such as:
K = (Pi - Pv)/(Pi - Pb)(9)
where Pi is the pressure at the inlet of the nozzle hole.
Other definitions of cavitation number are also used [Gavaises 2008][Som 2010][Gavaises 2009]. Cavitation occurs when the cavitation number defined by Equation (9) is below a critical value, Kcrit, that depends on pressure and nozzle details. While Equation (9) is easier to use in that all the pressures can be either measured or estimated, Kcrit must be determined experimentally or numerically.
Figure 13 shows an example of how Kcrit (red) varies with injection pressure for one particular fuel injector design. Also shown is the actual value of K for a nozzle outlet pressure of 6 MPa. In this example, cavitation occurs when the nozzle hole inlet pressure exceeds a relatively modest 45 MPa [Payri 2004].
Figure 13. Critical (Kcrit) and actual (K) cavitation numbers at different nozzle hole inlet pressures
Nozzle outlet pressure of 6 MPa. Fuel injector with 130 µm cylindrical nozzle holes with sharp edged inlets.
Since modern fuel injections typically use injection pressures well above those shown in Figure 13, cavitation will occur at most conditions in fuel injection systems utilizing straight nozzle holes with a sharp edged entrance. For this type of nozzle hole, the Cd under cavitating conditions is of more interest than that from Equation (7). Figure 14 shows the variation of Cd as a function of the square root of the cavitation number for three different nozzle hole inlet pressures. Cavitation occurs when the cavitation number is less than Kcrit. Under these conditions, Cd is seen to vary with √K. At K~1, Cd reaches a minimum value that is equal to Ca. In summary, under cavitating conditions:
Cd = Ca √K(10)
Figure 14. Discharge coefficient as a function of the square root of the cavitation number
Cylindrical nozzle with a sharp edged entry
Since cavitation has a major influence on the discharge coefficient of the baseline straight nozzle with a sharp edged inlet, it is worth considering what can be done to reduce or even eliminate cavitation in the nozzle and what effect that would have on the performance of the nozzle hole. Extrapolation of the curves right of Kcrit in Figure 14 to lower values of K suggests that reducing the effects of cavitation in an injector nozzle hole (i.e., reducing Kcrit) could potentially yield a Cd value of about 0.8 or higher.
Nozzle Hole Geometry. The principle approach used to minimize cavitation in nozzle holes is through changes to nozzle geometry from a cylindrical design with a sharp-edged inlet to something that has improved pressure distribution in the nozzle hole and is less prone to cavitation. Two techniques commonly used to modify nozzle hole geometry are:
abrasive flow machining (honing the hole by flowing an abrasive laden fluid through it) is used to radius the entrance to and smooth the inner surface of a cylindrical hole that has been produced with another technique such as electrical discharge machining (EDM) and
machining the hole to have a conical profile with a larger inlet diameter than exit.
Abrasive flow machining of a cylindrical hole provides numerous advantages including:
improved discharge coefficient of the nozzle. By minimizing or even eliminating the vena contracta, the low pressure at the nozzle entrance can be avoided,
narrower flow band tolerance for the nozzle due to the ability to hone the nozzle holes until a given flow rate is achieved,
pre-aging of the nozzle to ensure a more stable flow rate from the nozzle over the life of the engine,
improved nozzle durability due to the reduction of stress concentration at nozzle hole surface roughness features.
An example of a nozzle hole before and after abrasive flow machining is illustrated in Figure 15 [Jung 2008a]. The rounding of the inlet and the smoothing of the surface is apparent in the scanning electron microscope images.
Figure 15. Abrasive flow machining (AFM)
Left: Scanning electron
microscope images of an injector nozzle hole before and after AFM.
Right: Effect of total amount of AFM fluid passed through the nozzle on diameter gain, inlet chamfer radius and surface roughness.
The abrasive particle laden fluid used for abrasive flow machining can be either a relatively low viscosity fluid that has a viscosity similar to diesel fuel or it can be much more viscous—a paste. Abrasive flow machining with a low viscosity fluid is often referred to as hydro erosive grinding while that using a viscous paste is referred to as paste honing. Hydro erosive grinding has the advantage that honing and nozzle calibration can be carried out simultaneously due to the similar flow properties between the low viscosity grinding fluid and diesel fuel. It has the disadvantage that if relatively large amounts of material need to be removed, uncontrolled changes to nozzle geometry can result. Paste honing requires periodic checks of the nozzle calibration during the honing process using a different fluid from that used for honing but provides a more controlled material removal process [Diver 2007].
Rounding the inlet of the injector nozzle hole is a critical aspect of abrasive flow machining that can be utilized to reduce the effects of cavitation on Cd. The impact of nozzle hole inlet rounding on nozzle performance is illustrated in Figure 16 where the effect of different nozzle hole inlet chamfer radius (rc) is simulated [Jung 2008]. In this example, the baseline case with no rounding of the nozzle inlet shows a Cd of about 0.75 at low pressure (i.e., low flow rate) and about 0.64 at 100 MPa. As the inlet chamfer radius increases, two things happen, Cd increases at all pressures and the pressure (i.e., flow rate) at which cavitation occurs increases. The general upward shift in Cd is due to an a reduction in the nozzle hole entrance loss (Kent in Equation (7)). The effect of nozzle hole chamfer radius on Kent is shown in Figure 17. In real nozzle holes, an additional improvement in Cd would result from a reduction in the friction factor (f in Equation (7)) due to smoothing of the hole by abrasive flow machining. With sufficient inlet chamfer rounding, the increase in the pressure at which cavitation occurs lowers Kcrit enough that, at least for the range of pressures shown, the effect of cavitation on nozzle performance becomes almost negligible. As a result, Cd is able to remain high even for lower values of K where a cylindrical hole with a sharp edged inlet would experience significant cavitation and a drop in Cd as shown in Figure 14. Thus, the relatively high values of Cd as suggested by the extrapolation of the curves of the right side of Kcrit in Figure 14 appear to be possible with the reduction of cavitation effects.
Figure 16. Effect of nozzle hole inlet chamfer radius on Cd
Figure 17. Effect of inlet chamfer radius on loss coefficient (Kent)
In reality, large increases in Cd may be challenging to achieve with the rounding of the inlet chamfer of a cylindrical hole. If a relatively large amount of material is removed with hydro erosive grinding, uncontrolled shape changes to the hole can result. For example, a keyhole shape is one common deformity. This is produced by the removal of material from the lower side of the nozzle hole inlet due to the sharp change in flow direction. While paste honing could be considered an alternative, it is time consuming and would lead to increased production costs.
Another alternative is to machine the hole to have a larger diameter inlet than outlet. Some EDM machines have this capability. The conical shape can eliminate cavitation in the hole without the need for the removal of relatively large amounts of material by hydro erosive grinding to produce a large inlet chamfer radius. A nozzle with such a hole shape is sometimes referred to as a ks nozzle. Examples of the amounts of diameter difference required for this hole shape are shown in Figure 18 [Posalux 2007].
Figure 18. Cylindrical nozzle hole and conical holes of different amounts of taper
For conical holes, it is common to describe the hole in terms of the diameter difference between the inlet and outlet using the factor:
ks = (Dentry - Dexit)/10(11)
where Dentry and Dexit are in µm. In the literature, it is common to call this the “K factor” but ks is used here to avoid confusion with the cavitation number. Others define a similar parameter that normalizes the diameter difference between the inlet and outlet by the nozzle hole length [Benajes 2004].
The primary reason for the improvement in Cd with this hole shape is that there is a significant increase in the pressure in the inlet region of the nozzle hole and pressures below the vapor pressure of the fuel can be more easily avoided, Figure 19 [Dorri 2009].
Figure 19. Effect of conical hole shape on pressure distribution in nozzle hole
Position is distance from nozzle hole exit
While the conical shape provides much of the benefit in Cd for this nozzle shape, a light hydro erosive grinding can still be applied to impart the benefits of abrasive flow machining listed above [Dingle 2005]. When a lightly hydro erosive ground conical nozzle hole with a Cd of ~0.85 is compared to a cylindrical hole with sharp-edged inlet and Cd ~0.65, the conical shape provides about ~75% of the improvement in Cd while inlet rounding and honing effects contribute ~25%.
Figure 20, which is similar to Figure 14, compares Cd for a cylindrical and conical hole shape as a function of √K. The cylindrical hole behaves in a manner similar to that shown in Figure 14. The dramatic improvement in Cd with the conical hole shape is readily apparent—especially as K approaches 1 [Benajes 2004].
Figure 20. Comparison of performance of cylindrical and conical nozzle hole shapes at different nozzle hole inlet pressures
The effect of improved nozzle performance is illustrated in Figure 21 for three different nozzles with different Cd values. Each nozzle was designed to provide the same flow rate. The improvement in soot emissions is apparent with increasing values of Cd and is consistent with lower soot production with a longer lift-off length. It is also interesting to note that there is no clear NOx/PM trade-off with increased Cd. In this example, emissions of NOx appear to correlate much better with mean rate of heat release which could be controlled by other means [Benajes 2008].
Figure 21. Effect of nozzle hole Cd on ζst and dry soot emissions
NOx emissions and mean rate of heat release are shown to illustrate that there is no clear correlation of these two parameters with Cd.
Figure 16 and Figure 20 clearly show that nozzle hole geometry modifications such as inlet rounding and conical hole shape can provide a significant improvement in nozzle performance as indicated by increased Cd. This improved performance can provide an increase in the bulk discharge velocity of the fuel jet exiting the nozzle hole, U. Increased U will increase lift-off length and the amount of air entrained into the jet upstream of the lift-off length. This can be exploited to decrease combustion generated soot without necessitating increased injection pressure and/or major reductions in nozzle hole sizes. However, increased nozzle hole Cd can be combined with increased injection pressure and/or smaller nozzle hole size to realize further improvements.
The improvement in Cd by the minimization or even elimination of cavitation in the nozzle hole has other consequences as well. These include:
· Cavitation aids in the break-up and atomization of the fuel spray exiting the injector nozzle hole [Cameron 2008]. However, the effect of any degradation in break-up and atomization characteristics of sprays due to elimination of cavitation effects in injectors in modern diesel engines can be considered to be negligible under most conditions. Evaporation of liquid fuel in the fuel sprays in these injectors is typically controlled by the bulk entrainment of gas into the fuel spray and not by atomization.
· While eliminating cavitation increases nozzle efficiency and eliminates the choking condition, it can complicate fuel metering. Under cavitation conditions, nozzle flow rate depends only on upstream pressure and fuel metering can be achieved by adjusting pulse width and injection pressure. However, when cavitation is eliminated, fuel flow rate will additionally depend on pressure at the nozzle exit. To maintain accurate fuel metering, this nozzle exit pressure may need to be taken into account.
· Cavitation also serves to keep injector nozzle holes clean. Non-cavitating nozzle are more prone to carbon build-up inside the nozzle and may require elevated levels of fuel detergents to ensure that they remain clean [Tang 2009][Birgel 2008].
· Although geometric cavitation can be suppressed with changes to nozzle hole geometry, string cavitation can still occur under some conditions. String cavitation can occur even if pressure does not drop below the fuel’s vapor pressure. Cavitation strings appear in areas where large-scale vortices develop and originate either from pre-existing geometric-induced cavitation sites or even from gas or vapor drawn into the fuel jet from the nozzle exit. These vortices can form upstream of the injection holes as a result of nonuniform flow distribution and can persist inside the injection hole. Cavitation strings have been frequently observed to link adjacent holes. The nozzle flow and cavitation strings have been found to be sensitive to small variations in the needle shape and needle eccentricity. String cavitation is most likely to occur at low needle lifts [Gavaises 2009].
Variable Injection Nozzle Geometry
Variable geometry nozzles are another development that potentially overcomes some of the challenges of a fixed geometry injector nozzle. One important motive for pursuing variable geometry nozzles is that they can allow a variable fuel injector flow area. This could be beneficial for example at low load conditions where a flow area designed to provide the required full load fuel flow may be too large. A smaller flow area could allow injection over a longer time period and at higher velocity but with a lower penetration at low load to improve the combustion characteristics of the fuel jet and enable lower emissions. Less fuel injected during the ignition delay period would also reduce noise levels at lower loads. However, a variable flow area is not the only benefit of variable geometry nozzles. Some designs allow different spray patterns that could potentially enable different combustion modes to be used over the engine operating map.
Options for variable geometry nozzles include complete hole nozzles where different holes are opened at different load conditions and throttled holes where the nozzle opening is throttled at low load to reduce the effective nozzle flow area [Soteriou 2000][Busch 2004].
Figure 22 shows an example of a variable flow area nozzle by Bosch. One application of this Vario nozzle, also referred to as KVD (Koaxial Vario Düse), allows for a reduced spay hole area at part load [Mahr 2002][Busch 2004]. Additionally, this nozzle design also could allow using two different spray cone angles for the two rows of holes. If multiple injections are required, it could also allow a hydraulic separation down to zero between two injection events by using both nozzle rows [Engström 2008]. The Vario nozzle was one of the design concepts that Bosch had for their fourth generation light vehicle common rail system unveiled in 2006 for the US market [Bosch 2003].
Figure 22. Bosch Vario nozzle
Some development work around this nozzle for heavy-duty applications was carried out as part of the European-Integrated Project “GREEN” [Engström 2008]. However, after the GREEN project ended, Bosch stopped further development activities of fuel injection systems with coaxial-vario nozzles. No commercial benefits were expected from such a complex fuel injection system [Holleis 2011].
Nozzle hole entry and exit throttling, Figure 23, has also been investigated as a means to achieve a variable geometry injector nozzle. Both approaches create various spray types as the nozzle opening progresses. Also, the spray angle and targeting vary considerably as the nozzle opening changes. These characteristics can make it challenging to apply this design to most diesel engine combustion systems [Soteriou 2000].
Figure 23. Nozzle entry (left) and exit (right) hole throttling
Injector nozzles that use separate coaxial needles to open and close different sets of injector holes are a common approach to achieve a variable geometry nozzle, Figure 24 [Hergemöller 2012]. Such designs have been pursued by many injector manufacturers including Cummins [Perr 2003], Bosch [Hofmann 1983][Winter 2005] and Delphi [Hergemöller 2012].
Figure 24. Nozzle using coaxial needles to control separate sets of injector holes
A design analogous to the coaxial design of Figure 24 is an injector that has two separately controlled nozzles mounted side-by-side. Figure 25 shows such a design by Caterpillar. Each nozzle can have different size and number of holes as well as different spray angles that can be better optimized to a narrower range of engine operating conditions than a more conventional fixed injector design [Hergart 2011].
Figure 25. Injector having two separate nozzles
Mixed mode injectors are injectors with an injector nozzle that can provide different spray patterns to enable different combustion modes to be used in the same engine. A prototype mixed mode injector developed by Caterpillar that allows two different included spray angles is illustrated in Figure 26. With a narrow spray angle, fuel impingement on the cylinder wall can be avoided when fuel is injected early in the compression stroke (during LTC mode operation) or late in the expansion stroke (to raise exhaust temperature for DPF regeneration). The wider spray angle is intended for more conventional diesel injection near TDC of the compression stroke. This injector allows mixed mode operation of the engine where different combustion regimes are employed over different parts of the engine map [Duffy 2004].
Figure 26. Caterpillar mixed mode injector
Figure 27 illustrates a different mixed mode injector by QuantLogic [Hou 2006][Hou 2014][Hou 2009]. At low lifts, a narrow angle cone is produced by a pintle that disperses fuel more evenly and is intended to promote fuel and air premixing and minimize wall wetting. At full lift, fuel exits more conventional nozzle holes for more conventional diesel combustion. At mid lifts, fuel exits both the holes and the pintle opening. A similar design is provided by Delphi [Dingle 2010].
Figure 27. Mixed mode injector by QuantLogic
Group Hole Nozzles
In the group hole nozzle (or cluster nozzle) concept, a single larger hole is replaced by several smaller holes that provide the same flow area as the single larger hole are spaced closely together. By clustering several smaller holes together, the momentum of the jets from the individual holes is combined and spray penetration can be significantly increased from what would occur if the smaller holes were spaced farther apart. This is advantageous, especially at high load conditions where the spray penetration from a smaller hole may not otherwise be sufficient to make use of all the available air in the combustion chamber. While jet overlap in the sprays from the group of smaller holes would lead to lower air entrainment than if the smaller holes were farther apart, it is still considerably higher than what would be achieved with the single larger hole. Figure 28 illustrates some experimental results comparing a group of two 0.10 mm nozzle holes with a single 0.14 mm nozzle hole. The enhanced air entrainment in the group hole nozzle leads to enhanced liquid evaporation and a larger vapor region. A leaner fuel-air equivalence ratio at the lift-off length, φ(H), should also be possible [Miyaki 2006][Moon 2010][Nishida 2006][Sasaki 2007][Won 2009].
Figure 28. Effect of group hole nozzle on fuel spray
Laser Induced Exciplex Fluorescence. Injection pressure 80 MPa, 25 mm3/st. Ambient: 5 MPa, 600°C.
Another area of development for fuel injector nozzles are new materials that can ensure better nozzle durability. As fuel injection pressures continue to rise, it is imperative that the materials surrounding the nozzle spray holes resist high-pressure fatigue. Key challenges and questions for nozzle material selection that are being considered include [Blau 2010]:
· Holes must maintain dimensional tolerances and flow characteristics for tens of millions of pressure cycles.
· Nozzle materials must resist changes in shape, and allow holes to remain clear and open.
· What are the effects of residual stress, hole characteristics and metallurgy on the high-cycle fatigue response of nozzle tips?
· Will existing injector tip materials withstand the new design requirements, and if not, what alternative materials may be suitable?
0users like this.