Federal Office of Civil Aviation FOCA

Zurich Universities of Applied Sciences and Arts

Effective density of aircraft engine PM revisited: Effects of engine thrust, engine type, fuel, and sample conditioning

1

Federal Office of Civil Aviation FOCA

Lukas Durdina1, Beni T. Brem2, Miriam Elser3, David Schönenberger3, Julien G. Anet1

1 Centre for Aviation, ZHAW School of Engineering, Winterthur, 8400, Switzerland2 Laboratory for Atmospheric Chemistry, Paul Scherrer Institute, Villigen, 5232, Switzerland3 Empa, Dübendorf, 8600, Switzerland

DMA-CPMA-CPC for mass and sizeclassification in a standardizedsampling and measurement system

It really worked!

How we collected exhaust samples behind the engines


Effective density of non-volatile PM from large turbofan engines is useful for modeling and emissions testing. There is very limited good quality data.

2

• Applications• Estimation of PM number from mass emissions for airport air quality modeling (with

modeled particle size distributions)• Estimation of PM mass emissions from measured size distributions (especially

beneficial at ultra-low concentrations)• Correction for paticle losses in standardized sampling systems for aircraft turbine engine

emissions• Particle morphology information for health effects


Measurements were done on various in-service large turbofans with samples extracted at the engine exit plane and 25 m downstream using the DMA-CPMA-CPC setup

3

Engine Test type Sampling location

Fuel Eff. density measurement method

5 CFM56 variants

3 PW4000 variants

Short test(pass-off test)

Engine exit plane

Jet A-1 Fast - SMPS-CPMA

CFM56-7B Dedicatedtesting

Engine exit plane

Jet A-1 and 32% SAF blend

Fast - SMPS-CPMA

CFM56-7B Dedicated testing

~25 m downstream of the exhaust nozzle

Jet A-1 Slow - scanning CPMA, with and without a Catalytic Stripper

0.002 0.005 0.02 0.05 0.2 0.5 20.01 0.1 10

5000

10000

15000

20000

150 nm

100 nm

50 nm

dN/d

log

mp/c

m3

Particle mass mp [fg]

20 nm

• CPMA step scan of DMA classified particles at a selected mobility size

• Long stabilization and scan time of up to 10 min for sizes <50 nm

← 4 data points took more than 30minutes (very high engine cost!)

Slow method – DMA - scanning CPMA

10 1000

200000

400000

600000

800000dN

/dlo

gDp/c

m3

Dp [nm]

0.003 0.005 0.007 0.01 0.02 0.05 0.1 0.2 0.5 1 2

CPMA mass set point [fg]

Fast method – SMPS-CPMA • SMPS scan with CPMA running

between the DMA and the CPC at a fixed mass set point

• Delay time for SMPS scan must be determined down to 0.05s

← scan time of ~ 1 min per massset point

Particle mass in femtograms (fg) – 10-18 kg

DMA selected mobility sizes

Effective density = mass set point / volume of a sphere with diameter equal to the GMD of the size distribution for the mass set point

Effective density = GMD of the mass distribution / volume of a sphere with diameter equal to the DMA-selected mobility size

DMA-CPMA-CPC setup


Results(1/3): Distinct low thrust and medium-to-high thrustdensity distributions for all engines tested and no effectsof the 32% SAF blend found

4

20 50 20010 100

600

800

1000

1200

1400

1600

1800

20 50 20010 100

600

800

1000

1200

1400

1600

1800

effe

ctiv

e de

nsity

(kg/

m3 )

el. mobility diameter (nm)

4 16 28 40 52 64 76 88 100%max. rated thrust Foo

>30% Foo

<10% Foo

PW4000 engines (Foo ~250-300 kN) CFM56 engines (Foo ~120-150 kN)

effe

ctiv

e de

nsity

(kg/

m3 )


>30% Foo

<10% Foo

PM from CFM56 engines had higher effective density at medium to high thrust at particle sizes >50 nm than PW4000 engines.

Short engine tests with limited data

20 50 20010 100

600

800

1000

1200

1400

1600

1800engine thrust

GI-7% 30% 50% 65% 85% 100%

effe

ctiv

e de

nsity

(kg/

m3 )


symbol fillempty: HEFA blendfill: Jet A-1

Distinct low and medium-to-high thrust density distributions in agreement with those found during short engine tests. Highest eff. density at 65% thrust. At medium to high thrust, effective density >75 nm constant as a function of size. We did not find any fuel composition effects.

Dedicated campaign using a CFM56-7B engine withJet A-1 and 32% SAF blend


Results(2/3): 25 m downstream of the engine, the volatile mass contribution was highest at low thrust and <10%.

5

20 50 20010 100

800

1000

1200

1400

1600

20 50 20010 100

800

1000

1200

1400

1600

20 50 20010 100

800

1000

1200

1400

1600

20 50 20010 100

800

1000

1200

1400

1600

20 50 20010 100

800

1000

1200

1400

1600

0 25 50 75 100 125 150 175 2000.85

0.90

0.95

1.00

1.05

1.10

1.15 (f)(e)(d)

(c)(b) 3-7% no CS 3-7% CS

effe

ctiv

e de

nsity

(kg/

m3 )

(a) 30% no CS 30% CS

65% no CS 65% CS

85% no CS 85% CS

effe

ctiv

e de

nsity

(kg/

m3 )


100% no CS 100% CS


eff

, no

CS

/ ef

f , w

ith C

S


3-7% 30% 65% 85% 100%

0.002 0.005 0.02 0.05 0.2 0.5 20.01 0.1 110

100

1000

10000

mas

s sp

ectra

l den

sity

(dN

/dLo

gMp/

cm3 )

= 0.86= 0.87

= 0.99= 0.95

150 nm

60 nm

particle mass (10-18 kg)

no CS CS

20 nm

= 1.22= 1.19

Results of CPMA step-scans at 100% thrust with and without the catalytic stripper. Very little to no measurable volatile fraction. The red curves with CS have lower concentrations due to losses in the CS.

Effective density distributions with and without CS using the slow DMA-scanning CPMA method. At low thrust, the effective densities without CS were higher than samples after CS due to condensation of volatile PM on the soot.

Dedicated campaign using a CFM56-7B engine with Jet A-1 and sampling 25 m downstream of the engine exhaust nozzle


Results (3/3): Density vectors for low and high thrust (exponential fits) and average density of the integrated size distributions.

6

20 50 20010 100

600

800

1000

1200

1400

1600

20 50 20010 100

600

800

1000

1200

1400

1600

Durdina et al. 20143-5% thrust

effe

ctiv

e de

nsity

(kg/

m3 )


a) b)ef

fect

ive

dens

ity (k

g/m

3 )


3 13 22 32 42 52 61 71 81 90 100% max. rated thrust Foo

Durdina et al. 201450-100% thrust

The data shown include all samples at the engine exit plane and the samples taken 25 m downstream with the catalytic stripper.

0 20 40 60 80 100500

600

700

800

900

1000

1100

1200

Aver

age

PSD

den

sity

< 2

00 n

m (k

g/m

3 ) % max. rated thrust Foo

CFM56 short tests PW4000 short tests dedicated tests CFM56-7B "low" and "high" density vectors, all engines

The average density is the ratio of the integrated SMPS mass using different density vectors to the SMPS volume. The caveat here is the upper limit of ~200 nm. At high thrust, the size distributions are broader and there is a potentially significant mass fraction (~20%) above 200 nm.


Conclusions

7

• We provide new data on PM effective density for large turbofan engines• We confirmed the thrust dependence of effective density distributions for various

engine types• The mass-mobility relationship reported previously underestimated effective densities

at high thrust at sizes > 75 nm. Mass-mobility exponents determined from power lawfits depend on the size range of the data available. Thus, the exponent may bemeaningless if only data in a very narrow size range is available.


Additional information

8


Particle mass vs mobility diameter with Jet A-1 and 32% SAF blend

9

20 50 20010 1000.002

0.005

0.02

0.05

0.2

0.5

2

0.01

0.1

1

20 50 20010 1000.002

0.005

0.02

0.05

0.2

0.5

2

0.01

0.1

1

20 50 20010 1000.002

0.005

0.02

0.05

0.2

0.5

2

0.01

0.1

1

20 50 20010 1000.002

0.005

0.02

0.05

0.2

0.5

2

0.01

0.1

1

3-7% Foo Jet A-1 3-7% Foo HEFA

Mp

[10-1

8 kg]

Dp [nm]

30% Foo Jet A-1 30% Foo HEFA

Mp

[10-1

8 kg]

Dp [nm]


Mp

[10-1

8 kg]

Dp [nm]


Mp

[10-1

8 kg]

Dp [nm]

• Open symbols: Jet A-1• Filled symbols: 32% HEFA blend• No fuel effect observed


Comparison of the “slow” and “fast” methods on the same engine in campaigns one year apart. Slow method was used 25 m downstream of the engine with CS. Fast method was used on samples at the engine exit plane without CS.

10

10 100500

600

700

800

900

1000

1100

1200

1300

1400

1500

1600

10 100500

600

700

800

900

1000

1100

1200

1300

1400

1500

1600

10 100500

600

700

800

900

1000

1100

1200

1300

1400

1500

1600

10 100500

600

700

800

900

1000

1100

1200

1300

1400

1500

1600

10 100500

600

700

800

900

1000

1100

1200

1300

1400

1500

1600 3-7% 3-7%

effe

ctiv

e de

nsity

(kg/

m3 )

open symbols: sampling 25 m downstream after CS (slow method)filled symbols: sampling at the engine exit plane without CS (fast method)

30% 30%

65% 65%


85% 85%

100% 100%


Particle size distributions at the engine exit plane and 25 m downstream (corrected for particle losses and dilution) of a CFM56-7B engine. The gray area is the PM fraction removed by the CS. The apparent nucleation mode in the red curves may be a system loss correction artifact.

11

20 50 20010 100

0.5

1.0

1.5

2.0

20 50 20010 100

0.5

1.0

1.5

2.0

20 50 20010 1000.0

0.5

1.0

1.5

2.0

20 50 20010 100

1

2

3

4

20 50 20010 100

20

40

60

80

100

norm

aliz

ed c

once

ntra

tion

dN/d

logD

p

mobility diameter (nm)

exit plane 25 m downstr. no CS 25 m downstr. CS-removed PM

~100% ~85% ~60% ~30%thrust:

mobility diameter (nm) mobility diameter (nm) mobility diameter (nm)

idle ~3%

mobility diameter (nm)


Comparison of the slow and fast method using miniCASTsoot in the lab. Note the effect of the delay time (td) in the fast method. Here, 6 seconds delay time agreed best with the scanning CPMA method.

12

The SMPS measurements and the postprocessing were doneusing AIM 9.0. In the newerversions of AIM, the delay time cannot be set (AIM 10 allowssetting only the tubing lenght upto 100 cm).

0 20 40 60 80 100 120 140 160200

400

600

800

1000

1200

1400 scanning CPMA power-law fit td=6s td=6.5s 170216 MC1s 170629 MC1s

effe

ctiv

e de

nsity

(kg/

m3 )


Documents

Federal Office of Civil Aviation FOCA