SIMULTANEOUS INCREASING OF THERMAL CONVERSION

EFFICIENCY AND BMEP WHILE REDUCING EMISSIONS

Victor Gheorghiu, Prof PhD ME Hamburg University of Applied Sciences, Germany

AVL AST 2012

23 – 24 Oct 2012 Heidelberg

Content Introduction Concept regarding strict Implementation of Atkinson cycles =

Ultra-Downsizing Goals of this Investigation BOOST Simulation Tool and Model & Setting of the Simulations 1st Goal: Investigation Case of Optimal Ratio between Internal

and External Expansions & Compressions 2nd Goal: Atkinson cycles for part & full load even with

stoichiometric AFR (λ = 1), without throttling, intensive EGR, mixture stratifying, HCCI etc.

3rd Goal: Evaluation of the maximum improving potential of Ultra-Downsizing performances by means of an ideal V,p,T-Model for avoiding high optimizing effort of all BOOST model parameters

Conclusion 2 23-24.10.2012

23-24.10.2012 3

Introduction Thermodynamic Ways for Improving TCE* of Engine Cycles and therefore for CO2 Emission Reduction:

Ways 1. Increasing effective compression

ratio 2. Shorten eff. compression stroke

(e.g. delaying intake valve closing) 3. Completing eff. expansion stroke

(e.g. delaying exh. valve opening) 4. Turbo-charging for a concurrent

increase in TCE & BMEP

Schematic Pressure-Volume diagram of a four stroke Seiliger cycle, where the Heat Release is modeled by constant Volume (2-3v), Pressure (3v-3p) and Temperature (3p-3). For this reason this Ideal Cycle can be referred as Seiliger V,p,T-cycle.

Effective compression and expansion strokes of classical ICE cycles (usual named as Seiliger cycle) are almost identical.

* TCE = Thermal Conversion Efficiency

23-24.10.2012 4

Introduction Thermodynamic Ways for Improving TCE* of Engine Cycles and therefore for CO2 Emission Reduction:

Ways 1. Increasing effective compression

ratio 2. Shorten eff. compression stroke

(e.g. delaying intake valve closing) 3. Completing eff. expansion stroke

(e.g. delaying exh. valve opening) 4. Turbo-charging for a concurrent

increase in TCE & BMEP

Schematic Pressure-Volume diagram of a four stroke Seiliger cycle, where the Heat Release is modeled by constant Volume (2-3v), Pressure (3v-3p) and Temperature (3p-3). For this reason this Ideal Cycle can be referred as Seiliger V,p,T-cycle.

Effective compression and expansion strokes of classical ICE cycles (usual named as Seiliger cycle) are almost identical.

* TCE = Thermal Conversion Efficiency

1.

23-24.10.2012 5

Introduction Thermodynamic Ways for Improving TCE* of Engine Cycles and therefore for CO2 Emission Reduction:

Ways 1. Increasing effective compression

ratio 2. Shorten eff. compression stroke

(e.g. delaying intake valve closing) 3. Completing eff. expansion stroke

(e.g. delaying exh. valve opening) 4. Turbo-charging for a concurrent

increase in TCE & BMEP

Schematic Pressure-Volume diagram of a four stroke Seiliger cycle, where the Heat Release is modeled by constant Volume (2-3v), Pressure (3v-3p) and Temperature (3p-3). For this reason this Ideal Cycle can be referred as Seiliger V,p,T-cycle.

Effective compression and expansion strokes of classical ICE cycles (usual named as Seiliger cycle) are almost identical.

* TCE = Thermal Conversion Efficiency

1.

2.

23-24.10.2012 6

Introduction Thermodynamic Ways for Improving TCE* of Engine Cycles and therefore for CO2 Emission Reduction:

Ways 1. Increasing effective compression

ratio 2. Shorten eff. compression stroke

(e.g. delaying intake valve closing) 3. Completing eff. expansion stroke

(e.g. delaying exh. valve opening) 4. Turbo-charging for a concurrent

increase in TCE & BMEP

Schematic Pressure-Volume diagram of a four stroke Seiliger cycle, where the Heat Release is modeled by constant Volume (2-3v), Pressure (3v-3p) and Temperature (3p-3). For this reason this Ideal Cycle can be referred as Seiliger V,p,T-cycle.

Effective compression and expansion strokes of classical ICE cycles (usual named as Seiliger cycle) are almost identical.

* TCE = Thermal Conversion Efficiency

1.

2.

3.

23-24.10.2012 7

Introduction Thermodynamic Ways for Improving TCE* of Engine Cycles and therefore for CO2 Emission Reduction:

Ways 1. Increasing effective compression

ratio 2. Shorten eff. compression stroke

(e.g. delaying intake valve closing) 3. Completing eff. expansion stroke

(e.g. delaying exh. valve opening) 4. Turbo-charging for a concurrent

increase in TCE & BMEP

Schematic Pressure-Volume diagram of a four stroke Seiliger cycle, where the Heat Release is modeled by constant Volume (2-3v), Pressure (3v-3p) and Temperature (3p-3).

* TCE = Thermal Conversion Efficiency * * BMEP = Brake Mean Effective Pressure

1.

2.

3. 4.

8

Introduction

Ways 1. Increasing effective

compression ratio 2. Shorten eff. compression

stroke (e.g. delaying intake valve closing)

3. Completing eff. expansion stroke (e.g. delaying exh. valve opening)

4. Turbo-charging for a concurrent increase in TCE & BMEP

Conclusion: These four Thermodynamic Ways for improving TCE lead from Seiliger to Atkinson cycle, i.e. to a cycle with shorted effective compression stroke!

Schematic Pressure-Volume diagrams of classical four stroke Seiliger (left) and Atkinson V,p,T-cycles

9

Introduction

Consequences & Restrictions of their Implementation:

Thermodynamic Ways

1. Increasing effective volumetric compression ratio (VCR)

2. Shorten effective compression stroke

3. Completing effective expansion stroke

4. Turbocharging for a concurrent increase in TCE & BMEP

Consequences & Restrictions

Exceeding pmax , Tmax , NOx (Diesel & SI) limits, knocking occurrence (SI)

Decreased aspirated fluid mass lower BMEP & lower TCE improvement

Large displacement of the engine heavy engine, lower TCE improvement

Exceeding pmax , Tmax , NOx (Diesel & SI) limits, knocking occurrence (SI)

23-24.10.2012

Introduction

23-24.10.2012

1. Atkinson cycles have been implemented so far mostly with symmetrical crank mechanisms, where intake valves are closed very late on cycle. Thus, a part of charge sucked into cylinder is pushed back to intake pipes, and effective compression stroke is in this way decreased. This quasi implementation shows no noticeable improvements of IFCE and, hence, it will not be discussed here in detail.

2. Real Atkinson cycles can be implemented only with help of asymmetrical crank mechanisms. This strict Atkinson cycle implementations allow to use concurrently very high boost pressures (to increase IMEP) and higher VCR (to enhance IFCE) and to set them much more independently of each other compared to Seiliger cycles.

Possible Ways for Atkinson Cycle Implementation:

10

Introduction

1. Quasi-Atkinson cycles with symmetrical crank mechanisms by means of very late intake valves closing

Possible Ways for Atkinson Cycle Implementation:

11 Schutting, E, Neureiter, A, Fuchs, Ch., Schwarzenberger, T, Klell, M, Eichlseder, H, Kammerdiener, T: Miller- und Atkinson-Zyklus am aufgeladenen Dieselmotor, MTZ 06 / 2007

Introduction

23-24.10.2012

1. Atkinson cycles have been implemented so far mostly with symmetrical crank mechanisms, where intake valves are closed very late on cycle. Thus, a part of charge sucked into cylinder is pushed back to intake pipes, and effective compression stroke is in this way decreased. This quasi implementation shows no noticeable improvements of IFCE and, hence, it will not be discussed here.

2. Real Atkinson cycles can be implemented only with help of asymmetrical crank mechanisms. This strict Atkinson cycle implementations allow to use concurrently very high boost pressures (to increase IMEP) and higher VCR (to enhance IFCE) and to set them much more independently of each other compared to Seiliger cycles.

Possible Ways for Atkinson Cycle Implementation:

12

13 23-24.10.2012

A possible Realisation of such Asymmetrical Crank Mechanism

14 23-24.10.2012

A possible Realisation of such Asymmetrical Crank Mechanism

15 23-24.10.2012

Requirements: To raise IFCE and IMEP

simultaneously: The engine must be

highly turbocharged Compression stroke must

be much shorter as expansion stroke and VCR accordingly adapted.

Most of compression of working gas should occur outside of cylinder and Most of expansion within cylinder.

Advantages: As an important part of

compression takes place beyond cylinder, this high compressed fresh charge can be cooled intensively before it is sucked in.

Following moderate compression within cylinder leads to lower temperature peaks during combustion process and, consequently, to less NOx emissions.

Concept regarding strict Implementation of Atkinson cycles = Ultra-Downsizing

16 23-24.10.2012

Requirements: To raise IFCE and IMEP, to

limit pressure & temp. peaks during combustion even for stoichiometric mixture, simultaneously: Very high turbocharged

engine Optimized Ratios

between internal and external compression and expansion parts.

VCR must be varied accordingly.

Advantages: Lower temperature peaks

during combustion and thus less NOx emissions.

Realize real Atkinson cycles for part and full loads even with stoichiometric AFR and without throttling, intensive external EGR, mixture stratifying, HCCI…

Only 3-way catalysts are sufficient for after-treatment in this case.

Concept regarding strict Implementation of Atkinson cycles = Ultra-Downsizing

17 23-24.10.2012

1. To look for optimum ratio between internal (i.e. within cylinder) and external (within turbines) expansions of working gas, which leads simultaneously to maximizing IFCE and enabling sufficiently high values of IMEP.

2. To make possible the implementation of Atkinson cycles for part and full loads even with stoichiometric AFR and without throttling and/or intensive EGR.

3. To evaluate the maximum improving potential of Ultra-Downsizing performances, however avoid the high optimizing effort of all BOOST model parameter.

Goals of this Investigation Simulation

Tool:

BOOST*

BOOST*

Self made analytical

model

* AVL BOOST www.avl.com/boost1

18 23-24.10.2012

BOOST Simulation Tool and Model BOOST*-Model considers true

geometrical dimensions of engine components and losses caused by friction and heat transfer.

Power balances of all three turbochargers (TC) determine actual boost pressure level.

Expansion processes in turbines (Tx) are described by means of their discharge coefficients (µTx). When boost pressure required for preserving pressure limit on cycle is low, superfluous TC are kept for simplicity and comparability in use (i.e. are not bypassed here).

Inta

ke s

ide

Exh

aust

sid

e

19 23-24.10.2012

Setting of Simulations

Most parameters of BOOST model are selected for a hypothetical engine and are kept unchanged for all simulations, e.g.: All geometrical dimensions (with exception their of crank mechanism) Valve timing Wall temperatures, heat transfer coefficients, efficiencies and pressure

losses of intercoolers (target efficiency = 0.75, target pressure drop = 5 kPa), friction coefficient in pipes (0.019), blow by gap size of cylinder, frictional characteristic curve of engine etc.

Efficiency of turbochargers (compressor efficiency = 0.75, turbocharger overall efficiency = 0.5)

AFR, engine speed Combustion parameters:

Simple Vibe function (for modeling of heat release) Different positions of TDC on Atkinson cycles are compensated by

choosing a suitable start of combustion (SOC), so that combustion begins in all cycles uniformly at 15°CA before TDC.

20 23-24.10.2012

1st Goal: Investigation Case of Optimal Ratio between Internal and External Expansions

In Investigation Case A (IC A) VCR is varied and VER is kept steady Expansion stroke is kept

unchanged and compression stroke is varied significantly to allow modification of ratio between internal and external expansions.

Atkinson (Atk) cycles are implemented by varying eccentric radius exx of crank mechanism used.

Seiliger cycle is realized with zero eccentric radius.

TDC

Seiliger

Rel

ativ

e Pi

ston

Pos

ition

[-]

Crank Angle [°]

Exp

ansi

on S

troke

Com

pres

sion

Stro

ke

Inta

ke S

troke

Exh

aust

Stro

ke

Table shows: VER VCR, µTx, n (engine speed), AFR, SOC, CD (combustion duration), mVibe (exponent of Vibe heat release function), IFCE, IMEP, max(p) and max(T) (maximum cylinder pressure and temperature), pMP8 and TMP8 (mean boost pressure and temperature; i.e. at measuring point MP8) and pMP12 and TMP12 (mean exhaust back pressure and temperature; i.e. at MP12) for cylinder 1 (C1). 21

Parameter and Simulation Results for IC A

MP8

MP12

C1

Inta

ke s

ide

Exh

aust

sid

e

MP8 MP12 C1

Turbine discharge coefficients µTx are tuned in all cycles for reaching max(p) ≈ 230 bar. For reaching approximately same expansion rate in all three turbines, their µTx are set at the same level and compensated with cross sections ratios of turbine output pipes. Hence, only µT3 is adapted for each cycle to meet max(p) ≈ 230 bar. 22

Parameter and Simulation Results for IC A

MP8

MP12

C1

Inta

ke s

ide

Exh

aust

sid

e

23

Parameter and Simulation Results for IC A

Cylinder pressure (logarithmic) - Displacement volume diagram with valves timing for all cycles

IFCE - Crank angle (right axis) and Displacement volume – Crank angle diagram for all cycles

eo denotes exhaust open, ec exhaust closed, io intake open, ic intake closed

Atk e62

TDC

IFC

E [-

]

-33%

Forced exhaust

24

Parameter and Simulation Results for IC A

In IC A for all Atkinson cycles the aspirated gas mass changes only slightly

Seiliger (VCR=7)

Seiliger (VCR=15)

eo denotes exhaust open, ec exhaust closed, io intake open, ic intake closed

Atk e62

Gas Mass - Displacement Volume diagram with valve timing for all cycles

Cylinder Pressure (logarithmic) - Displacement Volume diagram with valves timing of all cycles

25 23-24.10.2012

Parameter and Simulation Results for IC A Trends arise from analysis of table values: All Atkinson show better IFCE values than Seiliger cycles. Seiliger cycles reach higher IMEP values because of longer

intake stroke and thus of more aspirated gas mass. IMEP follows IFCE variation and is mostly independent of boost

pressure variation in all Atkinson cycles. Highest IFCE value of Atkinson cycles is not reached in variant

with highest VCR, but in variant where VCR is ≈50% of VER.

26 23-24.10.2012

In Investigation Case B (IC B) VER and VCR are simultaneously varied

Exp

ansi

on S

troke

Com

pres

sion

S

troke

All four strokes are simultaneously varied by setting of parameter g while eccentric radius is kept steady to e32. Dashed curve shows

null position (g = 0) where a) expansion & exhaust and b) intake & compression strokes are identical (like IC A).

Inta

ke

Stro

ke

Exh

aust

S

troke

Dis

plac

emen

t Vol

ume

[lite

r]

Crank Angle [°] VIR = Volumetric Intake Ratio VXR = Volumetric Exhaust Ratio

2nd Goal: Atkinson cycles for part & full load even with λ = 1, without throttling and/or intensive EGR

27 23-24.10.2012

IC B = simultaneously variation of VER & VCR by steady eccentric radius (e32)

VIR = Volumetric Intake Ratio VXR = Volumetric Exhaust Ratio

Turbine discharge coefficients µTx should be set appropriately (while parameter g is varied) for fulfilling restriction for maximal cylinder pressure max(p) ≈ 230 bar.

All four volumetric ratios are varied simultaneously by setting parameter g.

null position

≈ 230 bar

max(p) µT1

µT3 µT2

Parameter g Parameter g

28 23-24.10.2012

IFCE varies in all OPs only within a 6% wide band while: Residual gas resides < 7% IMEP varies between 8.5

and 42 bar, max(T) varies between

1800 and 2300 K by unchanged heat release! max(p) within cylinder is

kept steady at ≈230 bar

IC B: Parameter & Performance at Full & Part Loads (stoichiometric AFR, without Throttling & EGR)

6%

IFCE Residual Gas

IMEP max(T)

29 23-24.10.2012

IFCE varies in all OPs only within a 6% wide band while: … Boost pressure varies

between 2.5 and 12 bar Boost temperature does

not exceed 360 K Maximal cylinder back

pressure reaches ≈11 bar Max. temperature before

T3 achieves only ≈1000 K

IC B: Parameter & Performance at Full & Part Loads (stoichiometric AFR, without Throttling & EGR)

6%

IFCE Residual Gas

IMEP max(T)

Boost pressure

Boost temp.

Back pressure

Temp. before T3

Parameter g Parameter g

pMP8 TMP8

pMP12 TMP12

30

IC B: Parameter & Performance at Full & Part Loads (stoichiometric AFR, without Throttling & EGR)

6%

IFCE Residual Gas

IMEP max(T)

Boost pressure

Boost temp.

Back pressure

Temp. before T3

Parameter g Parameter g

pMP8 TMP8

pMP12 TMP12

Comments: TMP12 achieves only 1000 K

because of extended expansion within cylinder. As benefit turbine wheel must

not be protected against higher exhaust gas temperature. As disadvantage a higher

cylinder back pressure pMP12 is required for achieving desired boost pressure pMP8.

31

Comments: Required cylinder back

pressure pMP12 diminishes the level of IFCE by ca. 25% because of consumed work for forced exhaust. The load independence of

these IFCE losses is quite unexpected, but …

IC B: Parameter & Performance at Full & Part Loads (stoichiometric AFR, without Throttling & EGR)

-25%

Forced exhaust

IFC

E

Dis

plac

emen

t Vol

ume

[lite

r]

Crank Angle [°]

32

Comments: The load independence

of these IFCE losses is quite unexpected, but if the difference between cylinder pressure at eo and cylinder back pressure is noted, the positive effect of free exhaust becomes evident.

IC B: Parameter & Performance at Full & Part Loads (stoichiometric AFR, without Throttling & EGR)

Forced exhaust Cyl

inde

r Pre

ssur

e (lo

garit

hmic

) [ba

r]

Volume Displacement [liter]

Free

exh

aust

33

IC B: Parameter & Performance at Full & Part Loads (stoichiometric AFR, without Throttling & EGR)

6%

IFCE Residual Gas

IMEP max(T)

Boost pressure

Boost temp.

Back pressure

Temp. before T3

Parameter g Parameter g

pMP8 TMP8

pMP12 TMP12

Big Question: How much can IFCE be improved, if all the BOOST model parameter (like heat release, valve times, turbocharging etc.) would be optimized?

23-24.10.2012

34 23-24.10.2012

3rd Goal: Evaluation of the maximum improving potential of Ultra-Downsizing performances by

means of V,p,T-Model for avoiding the high optimizing effort of all BOOST model parameters

In the case of supercharged engines, the number of parameters which influence the IFCE and BMEP is very high.

As a consequence, the effort to achieve combinations of parameters which maximize the performances of the real (by BOOST) cycle becomes difficult and very time expensive.

For these reasons, the V,p,T analytical model of ideal open cycles have been developed for this purpose (see Appendix).

In ideal V,p,T-cycle the heat is partially released isochorically (2 – 3v), isobarically (3v – 3p) and isothermally (3p – 3).

35 23-24.10.2012

… In ideal V,p,T-cycle the heat

is partially released • isochorically (2 – 3v), • isobarically (3v – 3p), • isothermally (3p – 3).

The amounts of heat released isochorically and isobarically depend on the targets for maximum pressure and temperature on the cycle (i.e. isochorically up to pmax, isobarically up to Tmax and the rest of the heat is released isothermally).

Cyl

inde

r Pr

essu

re [b

ar] (

loga

rithm

ic)

Displacement Volume [liter]

Some Details of V,p,T-Model for Open Cycles V,p,T (dashed curves)

with optimized valve timing heat release turbocharging …

BOOST (solid curves)

eo, ec = exhaust open / closed io , ic = intake open / closed

36 23-24.10.2012

In the analytical V,p,T model the thermal properties (i.e. isentropic exponents κc for unburned and κe for burned parts) of the working fluid are kept constant throughout the cycle. The entire fuel mass is added to the cylinder gas mass in the

state 3v of the cycle. The mass contribution of the exhaust rest gas part is also taken

into consideration. The available heat (from fuel combustion) is decreased by the

amount of heat transferred to cylinder wall. In this case, compression, combustion and expansion can be treated adiabatically. The backpressure pT behind the cylinder (equivalent of MP12

from the BOOST model) is computed by means of energy balance at the turbocharger for the desired boost pressure.

Some Details of V,p,T-Model for Open Cycles

Maximum improving potential of Ultra-Downsizing performances

37

38

Maximum improving potential of Ultra-Downsizing performances

Comparison between simulations doing for IC B by BOOST and V,p,T-Model

Optimized heat release of the V,p,T-cycle : • isochorically (2 – 3v) • isobarically (3v – 3p) • isothermally (3p – 3)

eo exhaust open, ec exhaust closed, io intake open, ic intake closed

V,p,T (dashed curves) with optimized valve timing heat release turbocharging …

Cyl

inde

r Pr

essu

re [b

ar] (

loga

rithm

ic)

Cyl

inde

r Pr

essu

re [b

ar]

Displacement Volume [liter] Displacement Volume [liter]

39 23-24.10.2012

Maximum improving potential of Ultra-Downsizing performances

V,p,T (dashed curves)

Comparison between simulations doing for IC B by BOOST and V,p,T-Model (max(T) limited by 2100 K)

Cyl

inde

r G

as M

ass

[gra

m]

Cyl

inde

r Te

mpe

ratu

re [K

]

Displacement Volume [liter] Displacement Volume [liter]

1 2

3v 4

5

6

7

Maximum improving potential of Ultra-Downsizing performances

40

41 23-24.10.2012

Maximum improving potential of Ultra-Downsizing performances

Parameter g Parameter g Parameter g Parameter g

IFCE

IMEP

pMP8 TMP8

pMP12 TMP12

Residual Gas

κc κe

ψ

λ

1 − ψ − θ

θ

ma

Hea

t Rel

ease

Rat

es

Ther

mal

Pr

oper

ties

42 23-24.10.2012

Maximum improving potential of Ultra-Downsizing performances

Parameter g Parameter g Parameter g Parameter g

IFCE

IMEP

pMP12 TMP12

Residual Gas

κc κe

ψ

λ

1 − ψ − θ

θ

ma

VCR VER

Hea

t Rel

ease

Rat

es

Ther

mal

Pr

oper

ties

43 23-24.10.2012

CONCLUSION

The optimum ratio between internal (i.e. within the cylinder) and external (i.e. within turbines) expansions of the work gases which maximize IFCE is reached when the VCR is close to ca. 50% of VER.

An asymmetrical crank mechanism which permits in addition to vary the VCR makes possible to realize Atkinson cycles for part and full load even with stoichiometric AFR and without throttling and/or intensive EGR.

The presented comparisons between V,p,T and BOOST simulations show that this analytical V,p,T model of ideal open cycles can simulate a real cycle relative accurate and predict correctly the upper limit of cycle performances under given engine operating conditions.

Thank you for your attention!

44 23-24.10.2012

Contact Information

Victor GHEORGHIU Prof. PhD ME HAW Hamburg University of Applied Sciences Faculty TI, Engineering and Informatics Dpt. MP, Mechanical Engineering Berliner Tor 21 20099 Hamburg, Germany Tel.: + 49 40 42875-8636 Fax: + 49 40 42875-8799 grg@rzbt.haw-hamburg.de victor.gheorghiu@haw-hamburg.de www.haw-hamburg.de/pers/Gheorghiu www.victor-gheorghiu.de

45 23-24.10.2012

Cyl

inde

r Pr

essu

re [b

ar]

Specific Volume [m3/kg]

q = Specific Heat [J/kg] n = Polytropic Exponent κ = Isentropic Exponent T = Temperature [K] cv = Isochoric Spec. Heat [J/kgK]