BerkeleyUNIVERSITY OF CALIFORNIA /

© 2

011

LM

AS

c

onta

ct e

mai

l: Manufacturing Process Precision Effects on Life Cycle Impacts of Automotive Drivetrain

Mon

eer H

elu

mhe

lu@

berk

eley

.edu

■  Manufacturers have become increasingly responsible for the environmental impact of their products throughout all life cycle stages

■  Manufacturing decisions can have a direct effect on product use (e.g. operational efficiency, service life)

■  Therefore, it is important to evaluate effect of manufacturing decision on life cycle environmental impacts

■  Products with environmental impacts dominated by their use phase may be especially affected by manufacturing decisions

Funding Sources: DTL/Mori Seiki and Industrial Affiliates of LMAS

Motivation

■  Efficiency of gear systems strongly influenced by manufacturing process: ■  Surface roughness of mating surfaces ■  Assembly errors (e.g. shaft misalignments) ■  Form errors

Introduction

General Vehicle Model Manufacturing Phase Analysis

Use Phase Analysis Results

Conclusion Future Work

■  Vehicle based on Honda Civic: ■  Vehicle mass, M = 1193kg ■  Frontal area, Ar = 2m2

■  Drag, Cd = 0.30 ■  Rolling resistance, Crr = 0.013

■  Functional load: ■  1.2 passengers each weighing

71.2kg and carrying 7kg of luggage

■  Fuel tank assumed 55% filled (13gal tank)

■  Functional life 130000km

■  Empirical surface roughness relationship for grinding (from Malkin & Guo 2008):

■  Representative process: ■  Process rate ~10-2cm3/s ■  Specific energy ~200000J/cm3

■  Volume of modeled gear pair ■  Surface area ~21055mm2

■  DOC = 1mm

■  Change in fuel consumption dependent on change in power that must be delivered by fuel:

■  Neglects power for accessories and idling powertrain

■  U.S. EPA FTP-75 used as standard driving cycle ■  Deceleration events removed from calculations

■  Fuel was regular, unleaded gasoline ■  Engine assumed to fully combust fuel

■  Comparing both results indicates that improving manufacturing precision of the final drive reduction can provide a substantial reduction in life cycle impacts

■  Since final drive reduction is one of several gears in automobile, impact of manufacturing precision on use phase could be much greater

■  Increased precision may also increase service life, which would further reduce life cycle impact of automotive drivetrain

■  Relationship exists between the manufacturing process precision of a product and its environmental impacts over its entire life cycle

■  In the case of automotive drivetrain components, increased process precision can reduce life cycle environmental impacts

■  Manufacturing process precision should be improved if resources required for improvement are less than potential benefit of improvement

■  Refinement of data and analysis

■  Explore causal effects

■  Extend to other manufacturing considerations and environmental impacts

■  Extend to other products

■  Extend traditional LCA methodology to evaluate the impact of process precision and other manufacturing considerations on the functional performance of a product during use

combined influence of the other two design parameters consid-ered, involute contact ratio !" and face contact ratio !#, on $̄ areillustrated in Fig. 11!b". Here the value of !" is varied from 1.25to 2 by changing the outside diameter of the pinion while the gearremained unchanged. Similarly, !# is varied from 1 to 2.5 bychanging the face width of both the pinion and the gear. Theinfluence !# is minimal perhaps since the load applied is kept thesame at Lin=200 Nm. This agrees with the experimental data #43$.Meanwhile, when the pinion addendum is increased to obtain ahigher !", larger sliding velocities are introduced with more sig-nificant efficiency losses. The difference in $̄ of gear pairs having!"=1.25 and 2.0 is nearly 0.1%, the latter one being less efficient.An increase in !" is known to reduce vibration amplitudes andnoise of the gear set. Figure 11!b" suggests that any noise im-provements through increased !" could reduce the efficiency ofthe gear pair.

The influence of two key tooth surface modification parameterson $̄ is shown in Fig. 12!a". Here varying amounts of linearprofile modification !tip relief" of magnitude %" !starting at theoperating pitch point" is applied to the tips of both the pinion andthe gear while a parabolic lead crown %# is applied to the piniononly. Here %# appears to have a limited effect on while increasing%" reduces the effective involute contact ratio by removing mate-rial from the areas of the larger sliding velocities, and hence,causes $̄ to increase quite significantly. A gear pair having a tiprelief of %"=15 &m is nearly 0.2% more efficient than its unmodi-fied counterpart. Finally, in Fig. 12!b", the combined influence of

pinion shaft misalignment error ' and center distance shift error (are demonstrated for Lin=200 Nm, Np=2000 rpm, Toil=40°C,and S=0.4 &m. Here, a positive ( corresponds to a larger centerdistance and causes $̄ to increase since the effective involute con-tact ratio is reduced. A misalignment slope is used to define thepinion shaft misalignment ', which is the ratio of the misalign-ment magnitude to the pinion face width. The gear shaft is as-sumed to be perfectly aligned. With a pinion shaft misalignment,the size and location of the contact zone tends to reduce and shifttoward one side of the flank along the lead direction, causing thecontact pressure to increase to certain amount depending on theamount of misalignment. In Fig. 12, it seems that $̄ is minimum atperfectly aligned condition, while $̄ increases very slightly withan increase of the amount of misalignment.

5 ConclusionsA model was proposed in this paper for prediction of friction-

related mechanical efficiency losses of parallel-axis gear pairs.The model combines a gear load distribution model, a frictioncoefficient model, and a mechanical efficiency formulation to pre-dict the instantaneous mechanical efficiency of a gear pair undertypical operating, surface and lubrication conditions. A new fric-tion coefficient formula was obtained by performing a multiplelinear regression analysis to a large number of EHL model simu-lations representing various combinations of all key parametersinfluencing the friction coefficient. The new friction coefficientformula was shown to agree with the measured traction data. Inaddition the gear pair efficiency predictions were compared to a

Fig. 10 Influence of „a… Lin on !̄ for various Np values for S=0.4 "m and Toil=100°C, and „b… S on !̄ for various Toil valuesfor Lin=200 Nm and Np=2000 rpm

Fig. 11 Influence of „a… #n on !̄ for various $n values, and „b…%# on !̄ for various %& values. S=0.4 "m, Lin=200 Nm, Np=2000 rpm, and Toil=100°C.

Journal of Mechanical Design JANUARY 2007, Vol. 129 / 67

Relationship between gear mesh efficiency, η, and RMS surface roughness, S, for different inlet lubrication temperatures, Toil, for a gear set modeled after an automotive final drive reduction (Xu, et al. 2007).

■  Objective was to study impact of increased manufacturing precision through higher surface finish of final drive reduction on life cycle environmental impacts of an automotive drivetrain

Slide 6 of 15 ! January 19, 2010

© LMAS UC Berkeley

Powertrain and drivetrain model

•! All figures based on

analysis from Wang

(2008)

•! All axles are assumed to

be stiff, solid

components so that no

losses occur during

torque transmission

•! Functional unit was

final drive reduction

unit

Engine

!e = 30%

Transmission

!t = 95%

Final Drive

Reduction

!f = 97.47%

Differential

!d = 97.47%

Drive Axle

Powertrain

Drivetrain

Simplified powertrain and drivetrain

Specific energy requirements of various manufacturing processes varied by process rate (Gutowski, et al. 2006)

Slide 10 of 15 ! January 19, 2010

© LMAS UC Berkeley

Use phase analysis

•! Fuel consumption dependent on power needed to:

–! Meet commanded acceleration

–! Run any accessories (e.g. air conditioning, radio, etc.)

–! Overcome losses in powertrain and drivetrain

•! Change in fuel power due to change in drivetrain efficiency:

–! Neglects power for accessories and losses in powertrain

!

"Pfuel =Ptractive

#e#t#d

1

# f 2$1

# f1

%

& ' '

(

) * *

!

Ptractive = v ma +mgsin" +mgCrr +(1

2#airv

2ArCd

$

% &

Where:

v velocity of vehicle

m total mass of vehicle and load

a commanded acceleration of vehicle

g acceleration due to gravity

! road grade

"air = 1.225kg/m3 at 15°C and 101.32kPa

! !"# !"$ !"% !"& !"' !"(!'"'

!'

!&"'

!&

!%"'

!%

!$"'

!$

!#"'

)*+,-./0123

456278+.2+9:+;8-62<+,==>?@3

+

+

?1+A+&!

14

?1+A+#!!

14

Manufacturing Phase: Use Phase:

Slide 7 of 15 ! January 19, 2010

© LMAS UC Berkeley

Manufacturing phase analysis

•! Only gear finishing considered

•! General grinding processes applied for gear finishing

•! Malkin and Guo (2008) provide empirical surface roughness

relationship based on correlation to undeformed chip thickness:

!

Ra= R1Q'w

vs

"

# $

%

& '

x

!

Qw2 =Qw1

Ra2

Ra1

"

# $

%

& '

1x

Where:

Ra avg. height surface roughness

Q’w specific volumetric removal rate

Qw volumetric removal rate

vs grinding wheel speed

R1, x experimentally determined constant

(0.15 < x < 0.60)

Ra average height surface roughness Qw volumetric removal rate x experimentally defined constant

Slide 8 of 15 ! January 19, 2010

© LMAS UC Berkeley

Manufacturing phase analysis

•! Representative grinding

process:

–! Process rate ~ 10-2cm3/s

–! Specific energy ~ 200000J/

cm3

•! Volume of modeled

helical gear pair:

–! Surface area ~ 21055mm2

–! DOC = 1mm

•! Michigan energy mix:

–! 7015.2Btu/kWh

–! 0.7131kg CO2-eq/kWh

!"#$%&'()%'*#+,*-#./*-0%&/*1+,+*2+%/*%!"#$%&'()$%*+,"+$$-"+,3%4+56+*3%7-8%"!

9#%:%;5*+%<

*=3%<>>?%

!

;/@%ABC8CDE%A?F?GE

7-2$.*.*@%A?HFIGE

&/JK5#+,%-*=%L-*9%AHFMGE

4/-=

&/*9#-*#%

A,5*%#.J+E%

A<>F<GE

N-,.-O0+%

A?HFIGE

P//0%&$-*@+%A"F"GE

QK.*=0+%AMFMGE

&/*9#-*#%

A9#-,#5KE%

A!"F<GE

&-,/59+0%A>FRGE

S*0/-=+=%7/#/,9%A<F>GE

QK.*=0+%T+8%A<F>GE

&//0-*#%)5JK%A<F>GE

Q+,6/9%A!F"GE

;/@%ABC8CDE%A?F?GE

7-2$.*.*@%A?HFIGE

&/JK5#+,%-*=%L-*9%AHFMGE

4/-=

&/*9#-*#%

A,5*%#.J+E%

A<>F<GE

N-,.-O0+%

A?HFIGE

P//0%&$-*@+%A"F"GE

QK.*=0+%AMFMGE

&/*9#-*#%

A9#-,#5KE%

A!"F<GE

&-,/59+0%A>FRGE

S*0/-=+=%7/#/,9%A<F>GE

QK.*=0+%T+8%A<F>GE

&//0-*#%)5JK%A<F>GE

Q+,6/9%A!F"GE

%

L.@5,+%"U%%V*+,@8%59+=%-9%-%15*2#./*%/1%J-#+,.-0%,+J/6-0%,-#+%1/,%-%"W-B.9%&X&%J.00.*@%J-2$.*+%YIZF%

%

>

!

<

"

R

H

?

[

I

> H> !>> !H> <>>

P$,/5@$K5#%A\@C$,E

])%<H])%H>])%?>])%[H])%!>>4/^%V*#$-0K8%W%%(-.9+%(+9.*%#/%'*_F%P+JK%W%)N&].@$%V*#$-0K8%W%(-.9+%(+9.*%#/%'*_F%P+JK%W%]̀ )V

N-,.-O0+%)5JK%]8=,-50.2%'*_+2#./*%7/0=.*@%7-2$.*+9F

QV&%A7;C\@E

%

L.@5,+%RU%%V*+,@8%59+=%1/,%6-,./59%$8=,-50.2%.*_+2#./*%J/0=.*@%J-2$.*+9%-9%-%15*2#./*%/1%#$,/5@$K5#%YMZF%

%

a*+% .JK/,#-*#% 2/*2059./*% 1,/J% #$.9% @+*+,-0.D-#./*% .9% #$-#%#$+%9K+2.1.2%+B+,@8%/1%-%J-*51-2#5,.*@%K,/2+99%.9%-%9#,/*@%15*2#./*%/1% #$,/5@$K5#F% %S*0.\+% #$+%+9#.J-#+9%/1#+*%J-=+%.*%4&b%9/1#^-,+3%9K+2.1.2%K,/2+99%+*+,@8%A+B+,@8E%.9%*/#%-%2/*9#-*#3%-*=%2-*%6-,8%95O9#-*#.-008c%.*%1-2#3%.#%.9%.*1.*.#+%-#%

.=0+% "!!! F% % d$.0+% .=0.*@% +e5.KJ+*#% .9% /1% 2/5,9+%

+6+*#5-008%9$5#%=/^*3%/K+,-#.*@%-#% 0+99% #$-*% 1500%2-K-2.#8%.9%*/#%5*2/JJ/*%-#%-00F%

%

b%9+2/*=%/O9+,6-#./*%.9%#$-#%/6+,%#$+%6+,8%O,/-=%,-*@+%/1%J-*51-2#5,.*@%K,/2+99+93%#$+%#/#-0%K/^+,%,+e5.,+J+*#%1/,%-*8%@.6+*%K,/2+99%^.00% 6-,8%/*08%O8%/*+%/,% #^/%/,=+,9%/1%J-@*.#5=+3% ^$.0+% #$+% ,-*@+% .*% J-#+,.-0% #$,/5@$K5#% 2-*%6-,8%O8%!>%/,=+,9%/1%J-@*.#5=+%/,%J/,+F%%P$-#%.93%+0+2#,.2-0%9#-*=-,=9% -*=% 9$.KK.*@% 2/*6+*.+*2+% $-6+% 2/*9#,-.*+=%J/9#%J-*51-2#5,.*@%+e5.KJ+*#%#/%#$+%,-*@+%/1%H%#/%H>%\dF%%a*% #$+% /#$+,% $-*=3% *+^% #+2$*/0/@.+9% A-*=% 0/^% +*+,@8%K,.2+9E%$-6+%J/6+=%J-*51-2#5,.*@%K,/2+99+9% .*#/%$.@$+,%K,+2.9./*%-*=%9J-00+,%9.D+%92-0+9F%%P$+9+%*+^%K,/2+99+93%^$.2$% /1#+*% /K+,-#+% .*% #$+% 6-K/,% K$-9+3% $-6+% J52$%9J-00+,% #$,/5@$K5#9% .*% #+,J9% /1% #$+% 5*.#% /1% J-#+,.-0%

/K+,-#+=%/*%.*%-%5*.#%/1%#.J+F%%P$+%,+950#%.9%-%,-#$+,%/,=+,08%K,/@,+99./*% /1% J-*51-2#5,.*@% K,/2+99+9% .*#/% 0/^+,% -*=%0/^+,% K,/2+99.*@% ,-#+9% -*=% $.@$+,% -*=% $.@$+,% 9K+2.1.2%+0+2#,.2.#8%,+e5.,+J+*#9F%%P$.9%#,+*=%.9%=.9K0-8+=%.*%L.@5,+%H% 1/,% "?% +B-JK0+9% 1,/J% !>% =.11+,+*#% J-*51-2#5,.*@%K,/2+99+9F% %P$.9%1.@5,+%^-9%2/*9#,52#+=% 1,/J%-%6-,.+#8%/1%9/5,2+9%0.9#+=%.*%P-O0+%!F%%X/#+%#$-#%-*%.*=.6.=5-0%K,/2+99%2-*%J/6+% 5K% .*% +0+2#,.2.#8% ,+e5.,+J+*#% O8% /K+,-#.*@% -#% -%0/^+,% K,/2+99% ,-#+F% % P$.9% $-KK+*93% 1/,% +B-JK0+3%^$+*% -%J.00.*@%J-2$.*+%.9%59+=%1/,%1.*.9$%J-2$.*.*@%6+,959%,/5@$%J-2$.*.*@3%/,%^$+*%-%&N`%K,/2+99%/K+,-#+9%/*%!%^-1+,%6+,959%<H>%^-1+,9%-#%-%#.J+F%

%

X/#+% #$-#% #$+% =-#-% .*% L.@5,+% H% J-8% ,+e5.,+% 15,#$+,%J/=.1.2-#./*% .*% /,=+,% #/% -@,++% ^.#$% #8K.2-0% +9#.J-#+9% /1%+*+,@8%A+B+,@8E%2/*95JK#./*%O8%J-*51-2#5,.*@%K,/2+99+9%@.6+*%.*%#$+%0.#+,-#5,+F% %L/,%+B-JK0+3%#$+%=-#-% .*%L.@5,+%H%1/,% .*_+2#./*% J/0=.*@3% @.6+*% O8% P$.,.+D3% -6+,-@+9% -O/5#%"%\;C2J

"F% %b995J.*@%-%K/08J+,%=+*9.#8%/1%!%@C2J

"%-*=%-%

@,.=%+11.2.+*28%/1%""G3% #$+%8.+0=9%-%9K+2.1.2%+*+,@8%6-05+%/1%!>%7;C\@F%%]/^+6+,3%J/9#%.*_+2#./*%J/0=.*@%/K+,-#./*9%.*205=+% -% 6-,.+#8% /1% -==.#./*-0% 95OWK,/2+99+9% 952$% -9%+B#,59./*3% 2/JK/5*=.*@3% -*=% =,8.*@3% -00% /1% ^$.2$% -==%95O9#-*#.-008% #/% #$+% +B+,@8% #/#-09F% % '1% #$+9+% -==.#./*-0%K.+2+9%/1%+e5.KJ+*#%-,+%-09/%.*205=+=%#$+8%,+950#%-%6-05+%1/,%.*_+2#./*%J/0=.*@%/1%-O/5#%<>%7;C\@%^$.2$%-@,++9%^.#$%#$+%0.#+,-#5,+%YM3%!>3%!!ZF%%%

%

!FVf><

!FVf>"

!FVf>R

!FVf>H

!FVf>?

!FVf>[

!FVf>I

!FVf>M

!FVf!>

!FVf!!

!FVf!<

!FVW>[ !FVW>H !FVW>" !FVW>! !FVf>! !FVf>"

),/2+99%(-#+%Y2J"C9Z

'*_+2#./*%7/0=.*@ 7-2$.*.*@ L.*.9$%7-2$.*.*@

&N` QK5##+,.*@ g,.*=.*@

bO,-9.6+%d-#+,_+# d.,+%V`7 `,.00%V`7

4-9+,%`7` aB.=-#./* SKK+,%h/5*=

4/^+,%h/5*=

V0+2#,.2.#8%(+e5.,+J+*#9%Y;C2J"%% Z

%

L.@5,+%HU%QK+2.1.2%+0+2#,.2.#8%,+e5.,+J+*#9%1/,%6-,./59%J-*51-2#5,.*@%K,/2+99+9%-9%-%15*2#./*%/1%#$+%,-#+%/1%

J-#+,.-0%K,/2+99+=F%

%

.% &!/0123%&/44*250%

L.@5,+% H% -*=% +e5-#./*% "% K,/6.=+% -% 9.JK0+% 2/*2+K#5-0%J/=+0%1/,%+9#.J-#.*@%#$+%+0+2#,.2-0%+*+,@8%,+e5.,+J+*#9%1/,%-%J-*51-2#5,.*@%K,/2+99F% %P$.9% .*1/,J-#./*% .9%*++=+=% 1/,%4.1+% &820+% b*-089.9% -*=% /#$+,% +*6.,/*J+*#-0% -22/5*#.*@%K,/O0+J9F% % P$+% =-#-% 9$/^9% #$-#% #/% -% 1.,9#% -KK,/B.J-#./*%#$+%J/9#%.JK/,#-*#%2$-,-2#+,.9#.2%/1%-%K,/2+99%.9%.#9%,-#+%/1%K,/=52#./*F%%P$.9%.9%O+2-59+%#$+%9K+2.1.2%+0+2#,.2-0%+*+,@8%,+e5.,+J+*#%.9%/1#+*%=/J.*-#+=%O8%#$+%95KK/,#%1+-#5,+9%/1%#$+%+e5.KJ+*#%,-#$+,%#$-*%#$+%-2#5-0%K$89.2-0%J+2$-*.9J%

Specific energy requirements of various manufacturing

processes varied by process rate (Gutowski, et al. 2006)

■  Michigan energy mix

0 20 40 60 80 1000

1

2

3

4

0.5

1.5

2.5

3.5

4.5

% of Nominal Ra

Change in P

E D

em

and (

MM

BT

U)

x = 0.15x = 0.60