Evaluation of spray-induced turbulence during the induction ...ltces.dem.ist.utl.pt/lxlaser/lxlaser2016/finalworks2016/...Evaluation of spray-induced turbulence during the induction

18th International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics・LISBON | PORTUGAL ・JULY 4 – 7, 2016

Evaluation of spray-induced turbulence during the induction stroke of a four-stroke single-cylinder optical engine

Brian Peterson1,*, Elias Baum2, Carl-Philipp Ding2, Dirk Michaelis3, Andreas Dreizler2, Benjamin Böhm4 1: Dept. of Mechanical Engineering, Institute of Energy Systems, The University of Edinburgh, UK

2: Fachgebiet Reaktive Strömungen und Messtechnik, Technische Universität Darmstadt, Darmstadt, Germany 3: LaVision GmbH, Göttingen, Germany

4: Fachgebiet Energie- und Kraftwerkstechnik, Technische Universität Darmstadt, Darmstadt, Germany * Correspondent author: [email protected]

Keywords: Tomographic Particle Image Velocimetry (TPIV), planar PIV, Internal Combustion Engine, Spray-induced Turbulence

ABSTRACT

Spray-induced turbulence proceeding injection augments mixing, thus playing an important role in engine

performance and stability, while also important for controlling pollutant formation. This work presents an

application of tomographic PIV (TPIV) to resolve the 3-dimensional, 3-component (3D3C) spray-induced turbulent

flow within a spray-guided direct-injection spark-ignition (SG-DISI) optical engine. TPIV measurements were

obtained after a single-injection from a hollow-cone spray when particle densities were suited for accurate TPIV

particle reconstruction. The injection strategy resembled the first-injection from a multi-injection strategy. High-

speed PIV (HS-PIV) measurements (4.8 kHz) were combined with phase-locked TPIV measurements (3.3 Hz) to

provide the time-history of the 2D2C flow preceding TPIV imaging. HS-PIV is also used to validate TPIV

measurements within the z = 0mm plane. TPIV uncertainties of 12% are assessed for non-injection operation. TPIV

was used to spatially resolve spray-induced turbulent kinetic energy (TKE), shear (S), and vorticity (Ω)

distributions. The added 3D3C velocity information is capable of resolving 3D shear layers that produce spatially-

coherent 3D turbulent vortical structures, which are anticipated to augment fuel-air mixing. Measurements spatially

quantify the increase of these parameters from injection and quantity distributions revealed significant differences

to non-injection operation. The isosurface density , defined as the volume percentage for which a flow

parameter exceeds a given value, identified distributions of the largest TKE, S, and Ω magnitudes, which indicated

the highest turbulence levels. Distributions quantify the increase of TKE, S, and Ω from injection and describe the

decay of spray-induced turbulence with time. At values below 10%, fuel injection increases TKE, S, and Ω

magnitudes up to 70% compared to the engine flow without injection. Measurements and analyses provide insight

into spray-induced turbulence phenomena and are anticipated to support predictive model development for engine

sprays.


1. Introduction

Direct-injection (DI) strategies offer the ability to improve fuel economy in spark-ignition (SI)

engines by unthrottled load control. Fuel injection can either take place “early” during the

induction stroke in attempt to provide relatively homogenous fuel-air mixtures or “late” during

the compression stroke to create a locally flammable stratified fuel charge, allowing flame

propagation with an overall lean fuel-air mixture. At present time, early-injection strategies are

more commonly utilized in commercially available highway vehicles than late-injection

strategies. For either DI strategy, proper fuel-air preparation is crucial to obtain reliable ignition

and limited emission levels (e.g. NOX and soot). For late-injection, mixing time is limited due to

the short time between end-of-injection (EOI) and ignition timing. However, even for early-

injection strategies, studies have indicated that mixing can be incompletely such that mixtures

are still heterogeneous near ignition timing (Snyder et al. 2011). Differences of cycle-to-cycle fuel-

air mixture distributions from early-injection strategies can provoke variances in engine

performance and make it difficult to control engine-out emissions (Alger et al. 2004).

The in-cylinder turbulent flow plays a pivotal role for mixture preparation within DI engines.

Injection of liquid fuel in excess of 10 MPa imposes locally high velocities and large velocity

gradients which often modifies the pre-existing flow field. Within the engine and spray

communities there is a need to better understand the spray-induced turbulent flow for proper

mixing and transport. Along these lines, it is important to study the turbulence produced by fuel

injection (termed: spray-induced turbulence) and understand how this turbulence enhances fuel-

air mixing in DI engines.

Laser-based diagnostic measurements and engine-spray simulations have provided our current

understanding of spray-induced flow physics. Experimentally, particle image velocimetry (PIV)

has provided the vast majority of spray-flow physics (Stiehl et al. 2013, Zeng et al. 2014, Peterson

et al. 2014, Zhang et al. 2014, Zeng et al. 2014, Peterson et al. 2015, Zeng et al. 2015, Stiehl et al.

2016). While planar PIV has provided a powerful understanding of spray-flow physics, the

limitations of 2D data suppress the understanding of an inherently 3D phenomenon.

Instantaneous 3-dimensional, 3-component (3D3C) velocity field measurements are required to

fully resolve the spray-induced shear layers that produce spatially-coherent turbulent vortical

flow structures that augment rapid mixing. Such measurements are also highly sought to

develop predictive models for optimizing flow patterns, mixing, and improving injector/engine

compatibility. Holographic PIV has captured 3D spray velocities (Choo and Kang 2003), but is

limited to sparse particle fields. Tomographic PIV (TPIV) and tomographic particle tracking

velocimetry have been applied within engines to capture the complexity of the 3D flow motion


and resolve the complete velocity gradient tensor (Peterson et al. 2012, Baum et al. 2013, Van

Overbrüggen et al. 2015, Zentgraf et al. 2016). TPIV should principally be suited for engine-spray

environments if locally dense particle distributions can be managed.

This work presents TPIV measurements to resolve the 3D3C spray-induced flow field within a

spray-guided DISI optical engine. TPIV measurements are obtained after early-injection when

particle distributions are suitable for accurate TPIV particle reconstruction. TPIV measurements

obtained after late-injection can be found in (Peterson et al. 2016). Planar high-speed PIV (HS-

PIV) measurements at 4.8 kHz are combined with TPIV (3.3 Hz) to provide a time-history of the

fuel-spray and 2D2C flow-field preceding the phase-locked, single-cycle TPIV measurements.

HS-PIV also provides TPIV validation within the central symmetry plane (i.e. z = 0 mm). TPIV is

further used to investigate the 3D turbulent kinetic energy (TKE) and the complete shear (S) and

vorticity (Ω) tensors, otherwise not available with HS-PIV.

2. Experimental

Velocimetry measurements were performed in a 4-stroke single-cylinder SG-DISI optical engine

operating at 800 RPM. Operating conditions are shown in Table 1. The engine is equipped with a

4-valve pent-roof cylinder head, centrally-mounted injector, centrally-mounted spark plug, and

quartz-glass cylinder and flat piston. Further details of the engine are described in (Baum et al.

2014, Freudenhammer et al. 2015). Silicone oil droplets (1µm diameter) were seeded into the

intake air for PIV, but fuel droplets also influence velocimetry measurements. Isooctane was

injected through a centrally-mounted, outwards opening piezo-actuated injector (105o spray

angle) with 18 MPa injection pressure. The injector operated with 500 µs injection duration and

EOI of 277o before top-dead-center (bTDC). The amount of fuel injected was 3.6 mg/cycle. This

injection event mimics a single-injection typically utilized amongst a multi-injection strategy to

avoid wall-wetting.

Table. 1 Engine operating conditions

Engine speed 800RPM

Intake press. / temp. 95 kPa / 300 K

Fuel (C8H18) / EOI 3.6 mg/cycle / 277obTDC

Inj. Press. / Temp. 18 MPa / 333 K

Intake Press. / Temp 95 kPa / 295 K

Charge density at EOI 1.1 kg/m3


Figure 1 shows the experimental setup for the combined TPIV and planar HS-PIV. A dual-cavity

frequency-doubled Nd:YAG (PIV 400, Spectra Physics, 350 mJ/pulse) operating at 3.3 Hz was

used for TPIV. The laser beam passed through a half-wave plate (p-polarized) and two

cylindrical lenses to expand and collimate laser light to specify the laser sheet thickness (5 mm).

The light passed through a polarizing beam-splitter and another set of cylindrical lenses to

expand and collimate the beam to specify the laser sheet width. Laser light was reflected off a 45o

mirror in the crankcase, providing a vertically illuminated volume in the engine. Four interline

transfer sCMOS cameras (LaVision, Imager sCMOS), with identical 100 mm lenses (Tokina) in

Scheimpflug arrangement were arranged circularly around the engine. TPIV camera angles were

chosen to provide the maximum range of camera angles suitable for the field-of-view (FOV). The

large camera angles between cameras 3,4 also accommodated the HS-PIV camera. Each TPIV

camera projection provided independent line-of-sight information of the illuminated volume (50

x 40 x 5 mm3) centered at z = 0 mm (i.e. the cylinder axis).

Fig. 1 Experimental setup of combined HS-PIV / TPIV in the optical engine.

A second dual-cavity, frequency-doubled Nd:YAG laser (Edgewave, INNOSLAB IS4 II DE, 8 mJ

/ pulse) operating at 4.8 kHz was used for planar HS-PIV. The laser beam passed through a

quarter-wave plate (circularly polarized) and a set of focusing optics before being combined with

the TPIV laser at the polarizing beam-splitter. Only the s-polarized light of the HS-PIV laser was

reflected and used for experiments (i.e. 50% of laser energy, 4 mJ / pulse). After the polarizer,

the HS-PIV laser light passed through the same focusing optics as the TPIV system. The HS-PIV

laser sheet of 1 mm thickness was positioned within the center plane of the TPIV volume (i.e. z =

0 mm position). A CMOS camera (Phantom V7.11) was placed between TPIV cameras 3,4 and

imaged onto a 55 x H mm2 FOV (H determined by piston position).


All camera and laser systems were synchronized to the engine at 800 RPM. HS-PIV images were

recorded at crank-angle degree (CAD) resolution from 285o bTDC until the CAD before TPIV

images were acquired, providing the 2D2C flow field evolution and droplet distribution before

TPIV. This was performed for 300 TPIV images acquired at 270o bTDC and at 258o bTDC. HS-PIV

images were not acquired after TPIV because of the sCMOS camera long exposure time (20 ms);

any additional light source (e.g. HS-PIV laser or combustion luminosity) within the second TPIV

exposure negatively biased TPIV measurements. The laser pulse separation (DT) for both the

HS-PIV and TPIV laser systems were 10 µs to resolve the spray-induced flow and high-velocity

intake flow. HS-PIV images were acquired for 288 consecutive cycles, while TPIV images were

recorded every 2nd cycle to acquire 300 phase-locked images at 270o and 258o bTDC. Limited disk

space of the HS-PIV camera (8 GB) prevented the camera from recording more than 288 cycles.

This limited the number of synchronized HS-PIV / TPIV sequences.

Additional phase-locked TPIV images were taken from 274o-270obTDC (100 cycles each CAD).

HS-PIV was not performed for this sequence. These TPIV images were recorded to study the

3D3C spray-induced flow evolution after EOI. Particle distributions were too dense to utilize

TPIV before 274o bTDC. TPIV images were acquired from 269o-258o bTDC, but were performed

with a 2nd fuel injection (400 µs, 269o-266o bTDC). Analysis at these CADs after the 2nd injection is

the focus of future work and is not presented here. TPIV images from 269o-259o bTDC with single

injection were not acquired, but images at 258o bTDC with single injection were acquired and are

sufficient to describe the relevant trends.

TPIV and HS-PIV were processed with DaVis 8.2.1 (LaVision). Images of a spatially defined

target (LaVision) within the engine were used to calibrate images and match viewing planes of

each camera system. A 15 pixel sliding minimum subtraction and local intensity normalization

were applied during TPIV image pre-processing. A volume self-calibration was performed for

100 images without injection. This provided a remaining pixel disparity less than 0.2 pixels. 3D

particle reconstruction was performed using an iterative Multiplication Algebraic Reconstruction

Technique algorithm (FastMART). TPIV was calculated by direct volume correlation with

decreasing volume size (final size: 64 x 64 x 64 pixels) with 75% overlap, providing a 1.2 x 1.2 x

1.2 mm3 spatial resolution (i.e. 80% final window size (Zentgraf et al. 2016)) and 0.375 mm vector

spacing in each direction. HS-PIV images were cross-correlated with decreasing window size,

multi-pass iterations from 64 x 64 to 32 x 32 pixels with 75% overlap, providing a 2.4 x 2.4 x 1.0

mm3 spatial resolution and 0.75 mm vector spacing in the x-y direction.


3. TPIV Assessment and HS-PIV

Mass conservation is applied to ascertain TPIV uncertainty. This is applied for non-injection

operation when density is considered to be spatially uniform. Continuity is assessed at 270o

bTDC:

ρ-1(∂ρ/∂t)+∂ui/∂xi=0 (1)

At 270o bTDC, the first term in equation 1 is neglected. Continuity is assessed similar to (Baum et

al. 2013) for cubic control volumes (CV) of equidistant grid-spacing (0.375 mm) throughout the

entire measurement volume for 300 images. In attempt to quantify the relative deviation from

mass conservation, the velocity difference (∆U =∆u + ∆v + ∆w) is normalized with the averaged

velocity (|V|3D,CV) that enters each CV. The PDF of ∆U / |V|3D,CV is shown in Fig. 2. The distribution

is symmetric around zero. The standard deviation (σ = 12%) represents the deviation of the

velocity flux and is the reported TPIV uncertainty. For reference the ∆U / |V|3D,CV distribution for

injection is also shown in Fig. 2, but this distribution is not used to ascertain mass conservation

because density is spatially variant and not quantified.

Fig. 2 PDF of ∆U / |V|3D,CV (left) and HS-PIV / TPIV velocity differences (right).

HS-PIV measurements are used to assess TPIV for fuel injection. Figure 4 presents instantaneous

HS-PIV and TPIV flow fields for two image sequences: (1) intake flow without fuel injection and

(2) intake flow with fuel injection. Details of the flow fields are described in Sect. 3.1. It is first

important to benchmark TPIV capabilities in comparison with the well-established HS-PIV

technique. To perform this assessment, x- and y-velocity components from HS-PIV at 271o bTDC

are extracted at each point in space and subtracted from TPIV velocity components at 270o bTDC.

This provides a spatially-distributed velocity difference between HS-PIV and TPIV. For

reference, this procedure is also performed for HS-PIV between 272o and 271o bTDC. These

operations are performed for 144 cycles (i.e. maximum number of synchronized HS-PIV / TPIV

datasets) for operation with and without fuel injection. Differences are evaluated within the z = 0


mm TPIV domain at the HS-PIV spatial resolution. Figure 2 shows PDFs of these velocity

differences. Velocity differences are not expected to always equal zero because data is extracted

at different CADs and velocity changes with CAD. All PDFs are centered at zero and show

similar distributions, demonstrating the HS-PIV and TPIV are in agreement. y-velocity

components have a slightly broader distributions than x-velocity components under all

conditions. This demonstrates that the vertical velocity components (predominant in the “intake-

jet” and “droplet-jet” (see Sect. 3.1)) exhibit a slightly larger spatially variation with CAD than

horizontal velocities. Figure 2 also shows that TPIV vs. HS-PIV differences from 270o-271o bTDC

are smaller than HS-PIV differences from 272o-271o bTDC. This aspect might result from

decreasing in-cylinder velocity magnitudes due to the gradual deceleration of the piston or

intake flow phasing at the particular imaging time-frame. Findings indicate that TPIV is as

reliable as the HS-PIV measurements and validates TPIV for the early-injection engine

environment performed within this study.

Fuel droplets, acting as tracer particles, influence velocimetry findings. In order to not disturb

gas-flow measurements, fuel droplets should behave similarly to oil droplets and accurately

follow the gas flow. A PTV algorithm (LaVision) was applied to the HS-PIV dataset to identify

individual fuel and oil droplets of diameter (d) and intensity (I) for injection and non-injection

operation (288 cycles). Non-injection findings identified oil droplets with average diameter

= 1 µm and average intensity = 580 counts. Individual droplet diameters for injection

operation were estimated by:

ddroplet = x (Idroplet / )1/2 (2)

The maximum particle intensity was calculated to be 3861 counts, which is less than the camera

saturation level of 4096 counts. The fuel droplet response time is calculated as:

to = ρfueldfuel2/18µair (3)

where ρfuel = 690 kg/m3 and µair = 1.85 e-5 Ns/m2 (evaluated for Tair = 300 K) (Green and Perry 2008).

Figure 3 shows to vs. ddiameter at 274o and 271o bTDC for clarity. Particle response times range from to =

2.0-14.0 µs, corresponding to droplet diameters (fuel or oil) of ddroplet = 0.9-2.6 µm. Droplet

diameters cover the same range for both CADs. Response times remain below 15 µs (i.e. greater

than 67 kHz frequency response), indicating that fuel droplets accurately follow the gas-flow

and do not disturb velocimetry measurements. This is not expected since images occur 0.8-4.2 ms

after injection in an expired spray plume.


3.1 Planar flow-field evolution from HS-PIV and TPIV

Figure 4 presents the flow evolution from two image sets to describe the instantaneous engine

flow without fuel injection (top) and with fuel injection (bottom). HS-PIV images are shown for

selected CADs from 278o-271o bTDC, while the image on the far right of each sequence represents

the two-component velocity field captured by TPIV (z = 0 mm) at 270o bTDC. The top row for

each image set shows the Mie scattering images, while the bottom row shows the corresponding

2D2C velocity field (z = 0 mm) represented by streamlines. The TPIV Mie scattering images are

taken from camera 2, which viewed nearly perpendicular to the imaging volume. Mie images in

Fig. 4 are normalized by the maximum intensity for better visualization since HS-PIV and TPIV

images were acquired with different light sources.

Fig. 3 Droplet response time vs. diameter for injection operation (288 cycles).

Images without injection show that velocity distributions are characterized by high velocities

entering the cylinder and the downward piston motion. Velocity magnitudes are highest near

the intake valves where the annular flow from each intake port impinges on each other, creating

a strong jet-like flow into the cylinder. This high-velocity, jet-like flow is referred to as “intake

jet” (Voisine et al. 2011, Freudenhammer et al. 2014). As the flow extends beyond the intake jet,

the flow is recirculated by the cylinder wall and the piston top, forming a clockwise tumble

motion within the symmetry plane. The HS-PIV image sequence shows the formation and

evolution of the recirculated tumble vortex below the intake jet. The TPIV image (far right)

shows the flow in a smaller FOV and qualitatively shows good agreement with the HS-PIV at the

preceding CAD (271o bTDC).


Fig. 4 Image sequences showing instantaneous Mie and velocimetry images for operation without injection (top)

and with injection (bottom). HS-PIV 278o-271o bTDC, TPIV (z = 0 mm) 270o bTDC.

Mie scattering and velocity images with injection show the distribution of fuel droplets and the

spray-induced flow field. The hollow-cone spray geometry is greatly distorted as the fuel

impacts the dual intake valves located directly beneath the centrally-located injector. At 278o

bTDC, liquid fuel is shown to penetrate through the dual intake valves (left-side), while liquid

fuel impacts the spark plug and scatters fuels on the right-side of the image. Light intensities

from multiple scattering saturate the 12-bit HS-PIV camera (4096 counts) in the fuel spray during

injection, but are under the saturation limit after injection. At 278o bTDC, liquid spray regions are


masked, but the remaining flow field shows the pre-existing tumble flow formation. Liquid fuel

penetrating between the dual intake valves quickly progresses downwards in the z = 0 mm

plane on the left-side of the image with velocities exceeding 25 m/s from 278o-270o bTDC. For

simplicity, this high-velocity fuel droplet region that penetrated between the intake valves and

progresses through the FOV on the left will be referred to as “droplet-jet”. After injection, fuel

droplets are dispersed within upper-half and left side of the FOV, while previously induced

fresh air, now redirected by the cylinder wall / piston top, is located on the lower-right of the

FOV.

The spray-induced flow after injection shows similar aspects to the induction flow without

injection; namely after injection, (1) the presence of the high-velocity intake-jet exists in both

cases and (2) the intake-jet flow impinges onto the flow redirected by the piston top, creating a

stagnation-zone comprised of several identifiable flow vortices. The main flow difference

observed within the example image sequences is the strong-downward droplet-jet motion on the

left-side induced from fuel droplets penetrating between the intake valves. This droplet-jet

quickly progresses downwards, creating a roll-up vortex at the edge of high-velocity region. The

presence of the droplet-jet reduces the area of the flow stagnation-zone and competes against

tumbling flow motion redirected flow from the piston top near the bottom of the images.

4. TPIV 3D3C velocity field results and discussion

TPIV is applied to spatially resolve 3D-TKE and instantaneous S and Ω distributions within the

engine spray environment. Measurements spatially quantify the increase of TKE, S, and Ω from

injection and distributions are compared against non-injection operation. Analysis describes how

the turbulent-infused fuel-cloud spatially evolves within the FOV and statistically reveals the

local decay of spray-induced turbulence with time due to molecular diffusion and dissipation.

Quantifying spray-induced turbulence for early-injection is important to understand rapid fuel-

air mixing to provide repeatable fuel-air mixtures at ignition timing. Such measurements are also

suitable for the development and validation of numerical engine spray models.

4.1 3D Turbulent Kinetic Energy (TKE)

Turbulence is first assessed by 3D-TKE defined as: TKE = ½(), where ui’ represents the

fluctuating velocity component in the ith direction and represents the ensemble-average.

TKE is calculated by Reynolds Decomposition (100 cycles). Figure 5 shows isosurfaces of 3D-

TKE at selected CADs after injection. The ensemble-average velocity (shown in the z = 0 mm

plane, every 29th vector shown) reveals flow motion and identifies regions such as the (1) intake-

jet, (2) stagnation-zone, (3) redirected tumble flow, and (4) droplet-jet.


For non-injection opertion, TKE is largest within the intake-jet region with values reaching as

high as 100 m2/s2. TKE values decrease with distance away from the intake-jet. With injection,

TKE values exceed 100 m2/s2 and are largest within the droplet-jet region and within a clockwise

formed vortical flow on the right-side of the FOV. As CAD progresses, the droplet-jet quickly

moves towards the bottom of the FOV, while TKE values decrease. At 270o bTDC, the droplet-jet

and clockwise tumble flow that remain in the FOV exhibit TKE values less than 100 m2/s2. As

time progresses to 258o bTDC, TKE distributions are similar between injection and non-injection

operation, but the flow field with injection exhibits a more-pronounced clockwise tumble flow at

the bottom of the FOV. The combination of the tumble flow and typcial spray-induced toroidal

vortical flows formed from the shear layers of the hollow-cone spray likely contribute to the

more-pronoucned clockwise tumble motion.

Fig. 5 3D-TKE isosurfaces and ensemble-average velocity (100 cycles, every 29th vector shown in the z = 0 mm) at

selected CADs.

Shortly after injection from 274o-270o bTDC, it is interesting to see that TKE values in the intake-

jet region are lower for injection than without injection, while the ensemble-average velocity

direction remains similar. Apparently, fuel injection organizes the flow within the intake-jet such

that lower TKE values (i.e. ui’) are calculated in comparison to the intake-jet not manipulated by

injection. Velocity fluctuations can occur from turbulent flow behavior or from cycle-to-cycle

flow variations (CCV). Zentgraf et al. have extensively analyzed the turbulent- and cyclic flow


behavior of the intake-jet for non-injeciton operation (Zentgraf et al. 2016). The intake-jet without

spray often exhibits high TKE values originating from the unsteady turbulent behavior (e.g. flow

separation, vortex shedding) as well as cyclic variances of the location and direction of mean

flow features (i.e. CCV). TKE assessement with larger sample-sizes is the focus on ongoing work

to further investigage the intake jet behavior with and without injection.

Fig. 6 3D isosurfaces of instantaneous ||S|| (left), ||Ω|| (middle), and threshold-based ||Ω|| with 3C velocity

field (z = 0 mm, every 21st vector shown) for non-injection operation. Each CAD represents different cycles.

4.2 Instantaneous shear (S) and vorticity (Ω) distributions

Analysis of the complete S and Ω tensors from 274o-270o and 258o bTDC are further used to study

instantaneous spray-induced turbulence and compare it to turbulence for operation without

injection. Unlike TKE, assessment of S and Ω does not require Reynolds decomposition; thus

they are not directly biased by CCV. Access to the complete S and Ω tensors enable quantitative

measurements of spatially coherent 3D vortical flows produced from the 3D spray-induced

shear layers. These aspects emphasize the advantage of TPIV measurements to characterize the

in-cylinder turbulence.


Fig. 7 3D isosurfaces of instantaneous ||S|| (left), ||Ω|| (middle), and threshold-based ||Ω|| with 3C velocity

field (z = 0 mm, every 21st vector shown) for injection operation. Each CAD represents different cycles.

Sequences of single-cycle images of the intake flow are presented in Figs. 6 (non-injection) and 7

(injection) at selected CADs. For brevity, only 270o and 258o bTDC are shown for non-injection

operation because the description of the intake flow without injection does not drastically

change from 274o-270o bTDC. Each CAD in Figs. 6 and 7 represent a different cycle. The

Frobenius norm (||...||) is used to represent the S- and Ω-magnitudes, which is used within the

remainder of the discussion. 3D ||S|| and ||Ω|| isosurfaces are shown in the left and middle

columns, while the 3C velocity (z = 0 mm) and threshold-based ||Ω|| isosurfaces (based on

average isosurface density, (see Fig. 8)) are shown in the right columns of Fig. 6 and 7.


Isosurface density is defined as the percentage of voxels exceeding a threshold (V thresh) to the total

number of voxels in the FOV (Vtotal) (Zentgraf et al. 2016). This identifies the distribution of largest

Ω-magnitudes occupying a percentage of the volumetric domain. Red isosurfaces represent the

largest spray-induced Ω-magnitudes occupying 4% volume at the CAD imaged closest to end-of-

injection (i.e. = 4%, ||Ω|| ≥ 19400 s-1), while blue isosurfaces represent the largest Ω-

magnitudes occupying 4% volume at the CAD of interest. For an instantaneous flow field image,

the red/blue isosurfaces provide a local comparison between large, spray-induced Ω-

magnitudes typically observed immediately after injection to the largest Ω-magnitudes observed

at the CAD and operation of interest.

At 270º bTDC, the individual non-injection cycle (Fig. 6) shows a high-velocity intake-jet

penetrating onto the recirculating flow. The largest ||S||,||Ω|| values are primarily contained

within the vicinity of the intake jet. At 258o bTDC, the individual cycle shows a less pronounced

intake-jet and the large ||S||,||Ω|| magnitudes are located further into the volume domain.

At both CADs, largest ||S||,||Ω|| magnitudes exist as individual spherically- or

cylindrically-shaped pockets and are more predominant near the intake-jet for the cycle shown

at 270o bTDC. Flows with ||Ω|| ≥ 19400 s-1 are seldom in the cycle at 270o bTDC and almost do

not exist for the cycle at 258o bTDC.

Representative individual cycles with injection (Fig. 7) show drastic differences in ||S||,||Ω||

distributions compared to non-injection operation. At 273o bTDC, the spray induces a larger

distribution of ||S||,||Ω|| magnitudes exceeding 19400 s-1. Flows exceeding these limits are

located within the intake-jet, but are more predominant in the droplet-jet (lower-left) and within

a roll-up vortex region shown on the right. Such roll-up vortices were more often observed for

operation with injection as opposed to operation without injection (see also Fig. 9). As CAD

progresses, the spray-induced ||S||,||Ω|| flow progresses further through the volume

domain, while pockets of high ||S||,||Ω|| decrease in size and magnitude. The largest

||S||,||Ω|| magnitudes remain near the droplet-jet and the spray-induced roll-up vortex, but

also exist within the intake-jet. At 258o bTDC, ||S||,||Ω|| magnitudes have significantly

decreased and are more reminiscent of those seen without injection. The droplet-jet does not

appear within the volume domain at 258o bTDC, but a strong clockwise vortex appears within

the lower-right of the image. This vortical flow (see also Fig. 9), could be a combination of the

spray-induced roll-up vortex and the tumble flow formation. The majority of the large

||S||,||Ω|| magnitudes exist within the intake-jet at 258o bTDC, but also exist near the

clockwise vortex edges, which was not present for non-injection operation.


4.3 Decay of spray-induced turbulence using distributions

Figure 8a-c describes the evolution of spray-induced TKE, S, and Ω distributions for all imaged

CADs. The figure shows 0-10% distributions vs. threshold for TKE, ||S||, and ||Ω|| (100

cycles). This analysis describes the distributions of highest magnitude for each variable to

indicate the highest turbulence levels and to quantify the evolution of spray-induced turbulence

with CAD. All curves exhibit a decaying trend with CAD, indicating a decrease in

parameter magnitude and turbulence level with CAD. Curves for injection show a monotonic

decrease with CAD; a larger decrease exists from 274o-272o bTDC, but is lessened from 272o-270o

bTDC. A large decrease exists between datasets at 270o and 258o bTDC, which is not unexpected

since more time progresses between images.

Curves for non-injection operation are shown in gray. There is little deviation in values from

274o-270o, while curves at 258o bTDC consistently exhibit lower values. This might be expected as

flow induction begins to decelerate past 270o bTDC as the piston speed decreases. Spray-induced

TKE and ||Ω|| values approach the non-injection values as time progresses, while ||S||

values continue to remain greater than non-injection values. For the engine and spray operation

employed, it appears that the spray-induced shear can remain within the cylinder up to 19 CAD

after EOI (~4.0 ms).

Fig. 8 (a-c) Average TKE, ||S||, and ||Ω|| isosurface density with CAD. Curves without injection are shown in

gray and indicated by arrows. CAD arrow indicates monotonic decrease for spray-induced quantities.

Figure 9 shows the spatially resolved decay of ||Ω|| with CAD in the volume domain. 3D

isosurfaces corresponding to the probability that the local flow exceeds ||Ω|| ≥ 19400 s-1 are

shown at selected CADs. This Ω-threshold represents the largest Ω occupying 4% volume at 274o

bTDC (i.e. Ω-threshold at = 4%). The image sequence shows the most-probable

locations of large ||Ω|| and the reduced tendency of high vorticity occuring as CAD

progresses. The ensemble-average flow is represented by streamlines. For non-injection

operation (top-row), ||Ω|| > 19400 s-1 only occur in the intake-jet region. Streamlines show the

general formation of the clockwise tumble motion in time.


For injection operation, the strong downwards droplet-jet (left) and roll-up vortex (right) are

evident by streamlines. The combination of the spray-induced roll-up vortex and tumble flow

appear to combine as one, which modifies the large-scale clockwise tumble motion compared to

operation without injection. Probabilities for ||Ω|| ≥ 19400 s-1 exceed 70% at 274o bTDC near the

droplet-jet and roll-up/tumble vortex regions. Values decrease as time progresses, but the

droplet-region and roll-up/tumble vortex continue to exhibit the highest probabilities of ||Ω||

≥ 19400 s-1. At 258º bTDC, probability distributions are similar to non-injection operation, but also

show that ||Ω|| ≥ 19400 s-1 flows can exist near the roll-up/tumble vortex, which is not the case

for non-injection operation.

Fig. 9 Probability isosurfaces for ||Ω|| ≥ 19400 s-1 at selected CADs for non-injection (top) and injection (bottom)

operation. Images for injection operation reveal location of highest turbulence and the decay of spray-induced

turbulence with CAD. Ensemble-average flow is represented by streamlines in the z = 0 mm plane (100 cycles).

5 Conclusions

Spray-induced turbulence proceeding injection augments mixing, thus playing an important role

in engine performance/stability and controlling pollutant formation. TPIV was applied to

resolve the 3D3C spray-induced flow within a spray-guided direct-injection spark-ignition

optical engine. TPIV measurements were obtained after a single-injection from a hollow-cone

spray when particle densities were suited for accurate TPIV particle reconstruction. The injection

strategy utilized mimics the first-injection from a multi-injection strategy. HS-PIV measurements


(4.8 kHz) were combined with TPIV (3.3 Hz) to provide the time-history of the 2D2C flow field

preceding TPIV images. TPIV uncertainties were assessed (12%) and measurements were

validated with HS-PIV in the z = 0 mm plane.

TPIV was used to spatially resolve 3D TKE, ||S|| and ||Ω|| distributions, otherwise not

available with planar PIV. Measurements quantify the increase of these parameters from

injection and distributions are compared against non-injection operation. Analysis revealed the

progression of the spray-induced droplet-jet and roll-up vortex within the FOV and quantified

the turbulence decay with CAD. For the engine operation, parameter magnitudes increased up

to 70% from injection. High shearing flows remained within the cylinder 19 CADs after injection.

Measurements and analyses provide insight into the spray-induced turbulence phenomena

during intake and are appropriate to support predictive model development for engine sprays.

Further analysis of this work will analyze probability distributions for shear and additionally

quantity PDFs within selected regions (e.g. intake-jet, droplet-jet, roll-up/tumble vortex) to

quantiatively study spray-induced turbulent flows. Operation with double-injection was also

utilized and will be the focus of future work.

6. Acknowledgements

Financial support by Deutsche Forschungsgemeinschaft (DFG) through grants PE 2068 and the

Gottfried Wilhelm Leibniz program are gratefully acknowledged. Financial support by

Sonderforschungsbereich Transregio (SFB/TRR) 150 is also gratefully acknowledged. The

authors are especially grateful to LaVision for loan of equipment.

7. References

Alger TF, McGee JM, Wooldridge ST (2004) Stratified-charge fuel preparation influence on the

misfire rate of a DISI engine. SAE Paper 2004-01-0549.

Baum E, Peterson B, Surmann C, Michaelis D, Böhm B, Dreizler A (2013) Investigation of the 3D

flow field in an IC engine using tomographic PIV. Proc. Combust. Inst. 34, 2901-2910.

Baum E, Peterson B, Böhm B, Dreizler A (2014) On the validation of LES applied to internal

combustion engine flows: part I: Comprehensive experimental database. Flow, Turbul.

Combust. 92, 269-297.

Choo Y-J, Kang B-S (2003) Measurements of three-dimensional velocities of spray droplets using

holographic velocimetry system. KSME Int. J. 17, 1095-1103.

Green DW, Perry RH (2008) Perry’s Chemical Engineers’ Handbook, 8th edn. McGraw-Hill Inc.,

New York.


Freudenhammer D, Baum E, Peterson B, Böhm B, Jung B, Grundmann S (2014) Volumetric

intake flow measurements of an IC engine using magnetic resonance velocimetry.

Freudenhammer D, Peterson B, Ding C-P, Böhm B, Grundmann S (2015) The influence of

cylinder head geometry variations on the volumetric intake flow captured by magnetic

image velocimetry. SAE Int. J. Engines 8(4), 549-556.

Peterson K, Regaard B, Heinemann S, Sick V (2012) Single-camera, three-dimensional particle

tracking velocimetry. Optics Express 20(8), 9031-9037.

Peterson B, Reuss DL, Sick V (2014) On the ignition and flame development in a spray-guided

direct-injection spark-ignition engine. Combust. Flame 161, 240-255.

Peterson B, Baum E, Böhm B, Sick V, Dreizler A (2015) Spray-induced temperature stratification

dynamics in a gasoline direct-injection engine. Proc. Combust. Inst. 35, 2923-2931.

Peterson B, Baum E, Ding C-P, Michaelis D, Dreizler A, Böhm B (2016) Assessment and

application of tomographic PIV for the spray-induced flow in an IC engine. Proc. Combust.

Inst. (in review).

Snyer J, Dronniou N, Dec J, Hanson R (2011) PLIF measurements of thermal stratification in an

HCCI engine under fired option. SAE International Journal of Engines 4(1), 1669-1688.

Stiehl R, Schorr J, Krüger C, Dreizler A, Böhm B (2013) In-cylinder flow and fuel spray

interactions in a stratified spray-guided gasoline engine investigated by high-speed laser

imaging techniques. Flow, Turbul. Combust. 91, 431-450.

Stiehl R, Bode J, Schorr J, Krüger C, Dreizler A, Böhm B (2016) Influence of intake geometry

variations on in-cylinder flow and flow-spray interactions in a stratified DISI engine

captured by time-resolved PIV. Int. J. Engine Res. doi: 10.1177/1468087416633541.

Van Overbrüggen, Klaas M, Bahl B, Schröder W (2015) Tomographic particle-image velocimetry

analysis of in-cylinder flows. SAE Int. J. Engines 8(3), 1447-1461.

Voisine M, Thomas L, Borée J, Rey P (2011) Spatio-temporal structure and cycle to cycle

variations of an in-cylinder tumbling flow. Exp. Fluids 50, 1393-1407.

Zeng W, Sjöberg M, Reuss DL (2014) Using PIV measurements to determine the role of the in-

cylinder flow field for stratified DISI engine combustion. SAE Int. J Engines 7(2), 615-632.

Zeng W, Sjöberg M, Reuss DL (2014) PIV examination of spray-enhanced swirl flow for

combustion stabilization in a spray-guided stratified-charge direct-injection spark-ignition

engine. Int. J. Engine. Res, 1-17.

Zeng W, Sjöberg M, Reuss DL (2015) Combined effects of flow/spray interactions and EGR on

combustion variability for a stratified DISI engine. Proc. Combust. Inst. 35, 2907-2914.


Zentgraf F, Baum E, Böhm B, Dreizler A, Peterson B (2016) On the turbulent flow in piston

engines: coupling of statistical theory quantities and instantaneous turbulence. Physics of

Fluids (1994-present), 28(4):045108.

Zhang M, Xu M, Hung DLS (2014) Simultaneous two-phase flow measurement of spray mixing

process by means of high-speed two-color PIV. Meas. Sci. Technol. 25 095204.

Documents

Evaluation of spray-induced turbulence during the induction ...ltces.dem.ist.utl.pt/lxlaser/lxlaser2016/finalworks2016/...Evaluation of spray-induced turbulence during the induction