SAE TECHNICAL PAPER SERIES
2000-01-1780
Overview of Techniques for Measuring Friction Using Bench Tests and Fired Engines Mike T. Noorman ExxonMobil Research and Engineering Company
Dennis N. Assanis and Donald J. Patterson University of Michigan
Simon C. Tung and Spyros I. Tseregounis General Motors Research and Development Development Center
Reprinted From: Advances in Powertrain Powertrain Tribology Tribology 2000 (SP–1548)
and International Spring Fuels & Lubricants Meeting & Exposition Paris, France June 19-22, 2000 400 Comm Commonw onweal ealth th Drive, Drive, Warre Warrenda ndale, le, PA PA 15096-0 15096-0001 001 U.S.A U.S.A..
Tel: Tel: (724) (724) 776-48 776-4841 41 Fax: Fax: (724) (724) 776-5 776-5760 760
The appearance of this ISSN code at the bottom of this page indicates SAE’s consent that copies of the paper may be made for personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay a $7.00 per article copy fee through the Copyright Clearance Center, Inc. Operations Center, 222 Rosewood Drive, Danvers, MA 01923 for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying such as copying for general distribution, for advertising or promotional purp oses, for creating new collective works, or for resale. SAE routinely stocks printed papers for a period of three years following date of publication. Direct your orders to SAE Customer Sales and Satisfaction Department. Quantity reprint rates can be obtained from the Customer Sales and Satisfaction Department. To request permission to reprint a technical paper or permission to use copyrighted SAE publications in other works, contact the SAE Publications Group.
All SAE papers, standards, and selected books are abstracted and indexed in the Global Mobility Database
No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior written permission of the publisher. ISSN 0148-7191 Copyright © 2000 CEC and SAE International.
Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE. The author is solely responsible for the content of the paper. A process is available by which discussions will be printed with the paper if it is published in SAE Transactions. For permission to publish this paper in full or in part, contact the SAE Publications Group. Persons wishing to submit papers to be considered for presentation or publication through SAE should send the manuscript or a 300 word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE.
Printed in USA
2000-01-1780
Overview of Techniques for Measuring Friction Using Bench Tests and Fired Engines Mike T. Noorman ExxonMobil Research and Engineering Company
Dennis N. Assanis and Donald J. Patterson University of Michigan
Simon C. Tung and Spyros I. Tseregounis General Motors Research and Development Center Copyright © 2000 CEC and SAE International.
longer life, higher operating temperatures, and reduced friction.
ABSTRACT This paper presents an overview of techniques for measuring friction using bench tests and fired engines. The test methods discussed have been developed to provide efficient, yet realistic, assessments of new component designs, materials, and lubricants for incylinder and overall engine applications.
Tribological bench testing in the laboratory can provide rapid and cost effective information, and is often used for screening or ranking purposes in the development process of new engine materials and lubricants. The authors will review the current development of selected bench tests employing either rotary or reciprocating motion for evaluating friction and energy-conserving characteristics of lubricants. The main advantage of the developed bench tests is that parts of real components (cylinder liners and piston rings) are tested, preserving the geometry and metallurgy of the engine, thereby permitting a representative evaluation of surface finishes and oils. Results from materials and lubricants studies will be presented and correlated with vehicle testing.
A Cameron-Plint Friction and Wear Tester was modified to permit ring-in-piston-groove movement by the test specimen, and used to evaluate a number of cylinder bore coatings for friction and wear performance. In a second study, it was used to evaluate the energy conserving characteristics of several engine lubricant formulations. Results were consistent with engine and vehicle testing, and were correlated with measured fuel economy performance.
Since the engine environment cannot be completely simulated in bench tests, engine tests are often needed to verify and validate findings. The authors have developed several techniques to measure overall and cycle-resolved, reciprocating component friction in fired engines. These techniques, and current research and development efforts (including investigation of new engine designs and energy-conserving lubricants) will be reviewed.
The Instantaneous IMEP Method for measuring incylinder frictional forces was extended to higher engine speeds and to modern, low-friction engine designs. A comparison of historical cylinder friction measurements shows reductions of 85% for late model piston /cylinder bore designs. A technique for accurately measuring overall engine friction was developed and used to assess the benefits of friction modifiers with an ability to measure changes in friction less than 1%.
TRIBOLOGICAL BENCH TESTS INTRODUCTION This section reviews bench testers employing either rotary or reciprocating motion for evaluating the friction, wear, and energy-conserving characteristics of interface materials and lubricants. Several configurations, as shown in Table 1, are in common use for friction and wear evaluation [1]. General information is presented on experience and practice with one rotary (Falex Type) and one reciprocating tester (Modified Cameron-Plint Machine) developed in the authors' laboratory.
Engine design and tribology engineers are constantly challenged to innovate advanced products to meet more demanding emissions and fuel economy targets. Current research and development on engine cylinder components include new designs, materials, coatings and surface treatments, with the goals of reduced weight,
1
Table 1. Common Bench Test Classification
(versus the conformal contact of the piston ring against the cylinder bore) limits the relevance of the test. In general, the main disadvantage of rotary bench test methods is that real components cannot be tested; hence, the geometry of engine components is not preserved, and representative surface finishes cannot be evaluated.
Figure 2.
Schematic of Cameron-Plint High Frequency Friction Tester
RECIPROCATING TEST METHOD - MODIFIED CAMERON-PLINT FRICTION AND WEAR MACHINE –
Figure 1.
A Cameron-Plint High-Frequency Friction and Wear Tester has been modified (Figure 2), to ser ve as a tool for evaluating advanced lubricants or advanced engine materials for possible use on engine cylinder bores. The modified Cameron-Plint High Frequency Friction and Wear Tester provides a reciprocating motion which makes it suitable for simulating piston/cylinder liner dynamics [6]. The main advantage of this bench test method is that real components can be tested, and since the geometry of the engine is preserved, representative surface finishes can be meaningfully evaluated. A specimen holder is moved back and forth across a cylinder segment as would occur in an engine. In addition, the holder has been modified to allow the piston ring to move in a slot, simulating the movement that occurs in a piston ring groove. Friction and electrical resistance are continuously measured during the test, and the wear of the cylinder liner and piston ring sections are measured with the use of a surface analyzer, using standard techniques [7]. In order to evaluate the stable friction characteristics of engine bores, friction coefficient measurements are performed after a sufficiently long sliding test time to ensure stabilization, typically 5 hours. Friction force is reported as the root mean square (RMS) of the instantaneous friction force over one cycle. The reciprocating bench test procedures are described in detail in [6,7]. Selected results from representative studies of advanced engine materials, coatings and energy-conserving lubricants are presented below to illustrate the potential and use of this method.
Schematic of Falex Type Block-On-Ring Friction and Wear Tester
ROTARY BENCH TEST METHOD – Standard rotary bench devices include pin-on-disk (ASTM G99), blockon-ring (ASTM G77), cross cylinders, and four ball tests [1,2,3]. While these tests can be designed for unidirectional, rotary or reciprocating motion, they all involve a non-conformal contact geometry. In the last 25 years, the rotary block-on-ring tester [4] (designed as LFW-1, also referred to as Falex friction and wear tester), shown in Figure 1, has been used in automotive laboratories for evaluation of piston rings, engine blocks, liners, valve guides, as well as surface coatings. Scuff results have been obtained by incrementally increasing load until failure is detected by friction measurement. Several procedures have been used for wear testing with this device. These have involved variations in load, speed, lubrication, temperature, and duration. The coefficient of friction is determined at the beginning and end of each test from the weight applied and the force transducer output. Block wear is most often evaluated by calculating the volume removed from the block wear scar area. The ring wear is evaluated by weight loss or a subjective evaluation of visual appearance. Test procedures for this rotary block-on-ring tester, used extensively for piston ring and cylinder liner evaluations, have been described in a reference paper published by Patterson, Hill, and Tung [5]. The non-realistic, non-conformal contact of standard specimens used in the block-on-ring test 2
0.125
Results show that the coefficient of friction did not vary significantly between the different test cylinder bores, as shown in Figure 3. After 5 hours of sliding, all materials reached stable friction characteristics. Friction was also measured after 20 hours and 40 hours to determine the surface impact over extended periods of sliding. The average coefficient of friction of the thermal sprayed steel coatings (either PTWA 1040 steel or DJ 1025 steel coatings) was 10% lower than the thermal sprayed bronze coatings, but slightly higher than the cast iron bore, as shown in Figure 3; the latter finding is attributed to the rougher surface finish of the thermal sprayed coatings.
5th hour 20th hour
0.120
40th hour
n e i c 0.115 i f f e o C 0.110 n o i t c 0.105 i r F
0.100 0.095 Cast Iron Bore
Aluminum 390
Nikas il
P TW A DJ 1040 Steel 1025 Steel
PTWA Bronze
DJ Bronze
Figure 3 Friction Coe fficients of Tested Engine Ma terials
Figure 3.
Friction Coefficients of Tested Engine Materials
The reciprocating bench test is especially useful in studying the wear patterns of surface coatings, which can differ from those of iron. Thermal spray microstructures consist of layered structures, called “splats,” from the flattening of the molten metal as it hits the surface. This results in layers of the parent material interlaced with oxides. Oxide layers between the splats are the weak link in the coating, as cracking occurs along the oxide boundaries, eventually releasing the splats [7,8]. This wear mechanism is usually referred to as “splat delamination.” Figure 4 shows bore wear depth as a function of test time for all specimens. Clearly, the diamond jet (DJ) aluminum bronze coating experiences the greatest wear. The PTWA aluminum bronze performs similarly to that of the aluminum alloy 390. The ferrous PTWA 1040 coating performed best, and was as good as or better that the Nikasil coating.
20 18 16
DJ - Bronze
n o r 14 c i 12 m [ h t 10 p e D 8 r a 6 e W 4
Cast Iron Aluminum 390 PTWA - Bronze Nikasil PTWA - 1040 Steel DJ - 1025 Steel
2 0 0
20
40
60
80
Time [hours] Figure 4 We ar Resistance of Tested Engine Materials
Figure 4.
Wear Resistance of Tested Engine Materials
RECIPROCATING BENCH TEST RESULTS ON SEVERAL ADVANCED ENGINE MATERIALS – New thermal spray processes are used to coat aluminum bores with either aluminum bronze or iron oxide coatings to provide a hard surface with good lubrication characteristics. Thermal spray coatings are considered advantageous replacements for cast iron since they have reduced weight, better heat transfer properties and lower costs. In order to assess possible friction and wear merits of such coatings, engine bore and piston ring components have been used in bench tests under conditions typical of the operating engine. In particular, temperatures and loads were chosen to represent the actual operating conditions at top-dead center, immediately after combustion. Advanced bore materials and coatings that were considered as gray cast iron alternatives for a particular 3.8L engine design included two thermal sprayed alloys produced by plasma transfer wire arc processes (PTWA-bronze and PTWA-1040 steel) [7]; two thermal sprayed alloys produced by diamond jet processes (DJ -bronze and DJ-1025 steel) [8]; and 390 Al-Si alloy and Nikasil composites.
These findings correlate very well with previously reported engine friction and wear tests [6,7]. Friction benefits followed the same trends, and consistent wear mechanisms were observed in both bench and engine tests. RECIPROCATING BENCH TEST RESULTS ON ENERGY-CONSERVING OILS – Bench tests can also be used to assess the tribological properties of engine conserving lubricants, including studies of base stock formulation, viscosity index modifier, and friction modifier. For instance, Table 2 lists a set of oils, which is explored in detail using bench tests in a companion paper [9]. These oils have been also described in reference [10], where related fuel economy characteristics in vehicle tests were reported. The selection of test oils spans a range of viscosity grades. The influence of friction modifiers, including organo-molybdenum additives (MoDTC), on friction and wear was also investigated. By including oils REO8 and REO9, the effects of using a formulation with a high viscosity index basestock was compared to one with a conventional basestock. Note that the ASTM FM-8 formulation is a friction-modified version of HR-4, the Sequence VI reference oil.
3
Table 2. Engine Oils Used in Cameron-Plint Bench Tests and in Fuel Economy Tests
on a vehicle tested over the FTP driving cycle, as shown in Figure 6.
0.090 t n e i c i f f e o C n o i t c i r F
0.080
1 0 3 0
H R -4 Oi l B B C
0.070
Figure 6.
0.060 2 0 5 0
Fuel Economy vs. Molybdenum Concentration
F M -8
0.050 0.040
R E O
0.030
R E O
8
9
52 9
Table 3. Vehicle Fuel Economy Data, Viscometric Properties of the Oils, and Friction Coefficient Data
0 . 0 3 4 M -2
0.020 0.010 0.000 0 .0
0 .5
1 .0 2 .0 1 .5 M o C o n c e n t r a t i o n , m i c r o g r a m s / c m 22
2 .5
3 .0
F i g u r e 5 . F r i c ti o n C o e f f i c i e n t v s . M o C o n c e n t r a t i o n . S o l id L i n e R e p r e s e n t s a P o w e r F i t o f a l l D a t a P o i n t s . T h e D a s h e d L i n e i s t h e A v e r a g e V a l u e o f t h e F r i c t io n C o e f f i c i e n t o f t h e F o u r O i ls w i t h t h e L o w e s t F r i c t i o n C o e f fic i e n t V a l u e s ( R E O 8 , R E O 9 , 5 2 9 , M - 2 )
Figure 5.
Friction Coefficient vs. Mo Concentration. Solid line represents a power fit of all data points. Dashed line is the average value of the friction coefficients of the four oils with the lowest friction coefficient values (REO 8, REO 9, 529, M-2)
Table 4. Fuel Economy Regression Results
A detailed surface mechanism study on the impact of energy-conserving lubricants has been published in a companion study [9]. Very significant differences in cylinder bore wear depths were observed after 5 hours of sliding using the different lubricants. Surface analysis techniques were used to reveal the presence of elements found in oil formulations, and to correlate them with friction reduction mechanisms of the energy conserving lubricants. A plot of measured friction coefficient versus the molybdenum concentration on the cylinder surface is shown in Figure 5. These results indicate a relationship between friction reduction and molybdenum concentration. A critical concentration of molybdenum (about 1.0 – 1.5 micrograms/cm 2) appears to be necessary for the friction reducing film formation to become evident. Beyond the critical concentration, friction stabilizes and approaches an asymptotic level. These observed effects of molybdenum surface concentration on bench-measured friction coefficient correlates well with ultimate fuel economy improvements
The bench-measured friction coefficient data and the viscometric properties of the oils (Table 3) were systematically correlated to vehicle dynamometer fuel economy data, as measured in two GM vehicles: a 1993 Buick LeSabre with a 3.8L V6 engine and a 1993 Pontiac Grand Am with a 2.3L Quad4 engine [10]. The engines were chosen to represent the two major types of GM valve train configurations currently produced, a roller follower design with low tension rings and two valves per cylinder, and a bucket tappet, dual overhead cam with four valves per cylinder design. In the correlation, the vehicle fuel efficiency data were assumed to depend on 4
one of the viscometric variables (to describe the hydrodynamic contribution to fuel economy) and on one of the friction variables (taken as a measure of the boundary friction contribution of the oil to the vehicle's fuel economy). Analysis of variance was used to identify which one of the viscometric and friction variables correlated best with the fuel economy data. The resultant correlation equations are shown in Table 4 for the two vehicles, as well as for the average fuel economy data of the two vehicles. A schematic plot of the predicted fuel economy for each vehicle and their average (as calculated from the equations) versus the measured fuel economy is shown in Figure 7.
By examining the relative contribution of the various terms in the equations in Table 4, we conclude that the 2.3L engine has a higher sensitivity to boundary friction than the 3.8L engine. This in agreement with the wellknown notion that in small-displacement engines, a greater portion of the mechanical friction is attributed to boundary losses than in large displacement engines. In addition, the valvetrain of the 2.3L engine (bucket tappets) is more sensitive to boundary lubrication than the valvetrain of the 3.8L engine (roller followers).
FIRED ENGINE FRICTION METHODS CYLINDER FRICTION – Motoring of an engine by a dynamometer is the easiest method of quantifying friction. By progressively disassembling the engine, the friction of various components can be inferred. Literature suggests that the piston ring assembly may make up as much as 60-75% of total engine friction [11]. The primary drawbacks of the motoring technique are the loss of the contribution of firing compression and the in-cylinder heat effects. More advanced techniques seek to study the cylinder friction under firing conditions using moving or floating cylinder liners [12,13,14,15]. These techniques require considerable modifications to the test engine, which raise questions about the representative nature of the resulting modified engine.
1.5 1.0 l i o o t
0.5
e 0.0 v i t a l e r , -0.5 ] g p m [ E F d e t c i d e r P
2.3L, (R^2=0.66) 3.8L, (R^2=0.67)
-1.0
AVG, (R^2=0.75)
-1.5 -2.0 -2.5
-1.5
-0.5
0.5
1.5
2.5
Measured FE [mpg], relative to BC oil
Figure 7.
The Instantaneous IMEP Method was developed in the early 1980’s as a way to measure cylinder friction in a fired engine with minimal modifications [16]. The method is based on an axial force balance around the piston assembly, which includes gas force, crankshaft force, inertial force and friction force (Figure 8). Friction can be determined by simultaneously measuring cylinder pressure and connecting rod strain, and calculating inertia. Instrumentation of the test engine is not trivial, but once in place, is minimally intrusive and therefore has minimal impact on the in-cylinder dynamics of the original system design. Instrumentation includes:
Predicted vs. Measured Fuel Economy in Vehicle Tests
1. A water cooled, piezo-electric pressure transducer in the cylinder head, 2. Four strain gages affixed to the connecting rod, two on each side, positioned parallel and perpendicular to the axis,
W eight
3. A shaft encoder on the crankshaft, capable of indicating 1 degree increments as well as engine TDC, and
friction = (gas pressure, inertia, weight, wrist pin f orce)
Figure 8.
4. Construction of a light-weight, yet durable, “grasshopper” linkage connecting the instrumented connecting rod cap to the engine block. This linkage is used to transport the strain gage wires out of the engine.
Force Balance Around the Piston Assembly
The friction coefficient measured during the (5th) hour in the friction test (FC5th hr) correlated best with the fuel economy data. This is not surprising, since the 5th hour friction coefficient represents friction values measured with broken-in surfaces of the friction pair, a condition that is expected in the engine cylinder-ring area that the bench friction test was designed to simulate.
A high-speed data acquisition unit is used to collect pressure and strain signals at each crank angle. Signals are conditioned and converted to forces. Cylinder pressure is first decreased by barometric pressure (which acts on the underside of the piston) and is multiplied by 5
bore area to determine the gas force acting on the topside of the piston. Strain data are converted directly to force via the conditioning step, but must be decomposed into forces acting parallel and perpendicular to the cylinder axis. This is done using the crank angle information and slider-crank theory. Examples of axial gas and conrod forces over a complete combustion cycle are presented in Figure 9.
80 60 40 e c r o 20 F n 0 o i t -2 0 0 c i r F -4 0
90
180
270
360
450
540
630
720
-6 0 -8 0 Crank Angle [degrees]
8000
F i g u r e 1 0 C y l i n d e r F r ic t i o n A c t i n g o n P i s t o n A s s e m b l y 1500 rpm, 1/6 WOT Load
6000
Gas Force
4000
Axial ConRod Force
[ 2000 e c 0 r o F -2000 0
90
180
270
360
450
540
630
Figure 10. Cylinder Friction Acting on Piston Assembly 1500 rpm, 1/6 WOT Load
720
Complete details of the Instantaneous IMEP method can be found in earlier SAE publications by Uras and Patterson [16,17]. The method has been applied to several engines, and used to evaluate the effects of oil viscosity, friction modifiers, ring tension and piston design on cylinder friction [16,17,18,19]. Generalized findings include:
-4000 -6000 -8000
Crank Angle [degrees] F i g u r e 9 Gas & ConRod Forces Acting on Piston Assembly 1500 rpm, 1/6 WOT Load
Figure 9.
Gas & ConRod Forces Acting on Piston Assembly 1500 rpm, 1/6 WOT Load
1. The ability to measure cylinder friction, on a cycle resolved basis, with enough accuracy to identify boundary, mixed and hydrodynamic lubrication regimes. Boundary lubrication exists at TDC and BDC, where piston speeds diminish to zero, causing the collapse of the oil film. Mixed lubrication, where increasing piston velocity increases oil film thickness causing a decrease in observed friction, is commonly found at the beginning and end of the stroke (first and last 30 degrees). Both mixed lubrication and hydrodynamic lubrication (where increasing speeds cause increased friction) have been observed in the mid stroke region, depending on engine speed and load.
Total inertial force is composed of the inertia of the piston, and the inertia of the length of the connecting rod from the midpoint of the strain gages to the wrist pin opening. Inertia is calculated directly from mass and acceleration (via slider-crank theory). Since the piston, rings and wrist pin travel exclusively in an axially direction, piston inertial force can be determined by assuming a point mass, while the complex motion of the conrod through the cylinder bore requires that its inertia be rigorously determined using a distributed mass approach. Friction force of the piston assembly (Figure 10) is calculated by subtracting the resulting gas force, inertial force and component weights from the axial conrod force.
2. Extremes in oil viscosity (high and low) showed increased cylinder friction. At high viscosity levels (29 cSt @ 100 °C), mixed lubrication regimes gave way to hydrodynamic, increasing friction. The effects of reducing viscosity to very low levels (2 cSt @ 100 °C) were relatively small at mid stroke, but very pronounced approaching TDC and BDC, as mixed lubrication gave way to boundary conditions. The use of friction modifiers has been shown to decrease peak friction force by 3040% at TDC and BDC, depending on oil viscosity. Overall cylinder friction, however, was not significantly reduced.
One drawback to the Instantaneous IMEP Method is that the resulting friction calculation is the difference of two relatively large force measurements determined using two differing techniques; therefore, the accuracy of the method is very dependent on the accuracy of the cylinder pressure and connecting rod force measurements. Calibration is therefore critical. Also, thermal shock of the pressure transducer following the combustion event is a common problem and care must be taken to select a model and an individual unit that is stable. Thermal shock causes a short-term drift in the pressure transducer, and since the strain gages are unaffected by the combustion event, the drift will show up in the calculated friction. Strain gages should be installed on the connecting rod to minimize sensitivity to bending, and at the same time be as close to the center of gravity of the rod as possible. Wiring the gages in a Wheatstone Bridge format increases sensitivity and compensates for temperature variations.
3. Cylinder friction losses under fired conditions were higher than those under motoring conditions, at the same coolant and oil temperatures, engine speed and intake pressure, due to ring loading and heat transfer. At low engine speeds, overall differences of up to 30% were observed, with the biggest changes observed in the power and exhaust strokes. These differences decreased at higher speeds and reduced loads where lubrication was more hydrodynamic.
6
4. While cylinder friction varies with speed and load, results show that it represents only 20-30% of overall engine friction. This is contrary to literature findings [11] which suggest cylinder friction is responsible for 60-75% of overall engine friction.
1600 rpm speed range) than the 60-75% previously suggested by literature. Clampitt compared fired and motored friction measurements, with mixed findings. Virtually no difference between fired or motored conditions were observed for the overall and cylinder friction measurements using stock pistons (Figure 12), while an average of 12 – 15% reductions were found for the light weight piston design (Figure 13). These findings show a considerably smaller difference between fired and motored friction than the 30% reported earlier by Uras and Patterson.
Initial work by Uras and Patterson was conducted at 500 rpm, later increasing to 1600 rpm. The Instantaneous IMEP Method was extended by Clampitt [20] in the late 1980’s to measure cylinder friction at engine speeds up to 4200 rpm, based on advances in high-speed data acquisition systems which made it possible to coll ect data at higher speeds and for a longer duration. Data covering 100 cycles could be captured, up from a maximum of 9 just a few years before. Increased cycle averages helped smooth data considerably, which proved valuable at higher engine speeds.
2500 s s 2000 e r P e v i t c ] 1 5 0 0 e r f f a E b n m a [ 1 0 0 0 e M n o 50 0 i t c i r F
A program [20] was conducted to evaluate the performance benefits of a new, light-weight piston assembly. A 1989 fuel injected 2.5L four cylinder engine was instrumented with pressure transducer, strain gages and grasshopper linkage on cylinder 1. Operated from 2000 to 4200 rpm at full load conditions, increased brake power measurements indicated improved performance for the light weight piston over the stock assembly; while no substantial differences were observed in cylinder friction. Subsequent work by Clampitt [20] showed that benefits were related to improved combustion with the new design rather than weight savings. These results are consistent with work done by Uras and Patterson [16] which showed that reduced weight pistons can actually increase friction slightly (due to deformation), depending on speed. Additional work by Uras and Patterson targeted at investigating the effect of piston weight at constant design confirmed that decreasing weight alone has negligible effects on cylinder friction.
0 2000
Fired Overall FME P Motored Overall FME P Fired Cylinder FMEP Motored Cylinder FMEP
2400
2800
3200
3600
4000
4400
Engine Speed [rpm] Figure 12 Fired vs. M otored Friction - Stock Piston
Figure 12. Fired vs. Motored Friction – Stock Piston 2500
s e r P 2000 e v i t c e ] f 1500 f r a E b n m a [ e 1000 M n o i t 50 0 c i r F
0 2000
Fired Overall FMEP M o t o r e d O v e r a ll F M E P Fired Cylinder FMEP Motored Cylinder FMEP
2400
2800
3200
3600
4000
4400
Engine Speed [rpm] f o % s a , n n o o i i t t c c i i r r F F r e d n i l y C
100% 9 0%
Stock Piston Assembly
8 0%
Lightweight Piston Assembly
7 0%
Figure 13 Fired vs. Motored Friction - Lightweight Piston
Figure 13. Fired vs. Motored Friction – Lightweight Piston
6 0% 5 0% 4 0% 3 0% 2 0%
A new program is underway designed to extend the Instantaneous IMEP Method to late model, low-friction engine designs. Tests have shown limited success in applying the method to a 1995 2.5L V6 engine. The allaluminum, split-block design uses low-friction, rollered, multivalve heads and low-friction, molybdenum coated short skirt pistons. Limited testing has been conducted at 1500 rpm and 1/6 full load, which approximates a steady road cruise operation of 45-50 mph. Cylinder friction was measured at 20% of overall engine friction, consistent with the low end of the 20-30% range reported by Uras and Patterson.
1 0% 0% 2000
2400
2800
3200
3600
E n g i n e S p e e d [ r p m] Figure 11 Normalized Cy linder Friction Stock and Lightweight Piston Assemblies
Figure 11. Normalized Cylinder Friction Stock and Lightweight Piston Assemblies Cylinder friction results from the 2.5L4 engine, normalized to overall engine friction, are presented in Figure 11. The light-weight piston design exhibited overall similar friction characteristics to the stock units, with slightly higher levels observed in the 2600 to 3000 speed range. Results ranged from 30-40% over the 2000 to 3400 rpm range, which are in closer agreement to the Uras and Patterson results of 20-30% (over the 500 to
The most challenging aspect to the success of this program has been the low absolute friction levels of the engine design. In terms of friction mean effective pressure (FMEP), nominal cylinder friction values averaged 110 mbar. This compares to nominal values of 7
480 mbar and 730 mbar for engine work done by Assanis in the late 1980 ’s and Uras and Patterson in the early 1980’s, respectively [16,20]. On a force basis, average cylinder friction forces of 60 N were sought from a difference between gas and conrod forces that peaked over 6000 N.
The general developmental strategy was to operate an engine at a given steady-state operating condition, accurately measure cylinder pressure and power absorption at the dynamometer, introduce the friction modifier additive, and then re-measure cylinder pressure and brake power. The largest obstacles were achieving true steady-state operation in a modern, microprocessorcontrolled engine, and data processing.
Other design issues also raised new challenges. Limited space in the four-valve cylinder head necessitated the use of the smallest available pressure transducers. Nonwater-cooled, Kistler 6125A units equipped with thermal shock resistant covers were installed through the water jacket of the cylinder head. Thermal shock was minimized, but still apparent for 30 degrees following the start of combustion. Powdered metal technology connecting rods made strain gage selection difficult, as compensation for temperature related expansion of the connecting rod is done by selecting strain gages with likemetal backings. This problem was circumvented by calibrating the instrumented connecting rod at in-use temperatures.
A 1995 2.5V6 engine was used in this program, mounted on an electric dynamometer designed to maintain a set engine speed by absorbing flywheel power. Engine load was controlled by throttle position. All accessories were removed from the engine, leaving a single drive belt that operated the stock water pump. Engine coolant and oil temperatures were controlled to 95 and 105 °C, respectively. The engine was equipped with complete factory emissions equipment and was controlled by an electronic control module (ECM) through an OEM wiring harness. Provisions were made to monitor and control such input parameters as spark advance and barometric pressure. In addition, the ECM was removed from adaptive learning mode. While these changes do not provide absolute control over the engine operation, the ECM was more likely to operate using a limited number of operational maps. This had a marked effect on maintaining steadystate operation.
OVERALL ENGINE FRICTION – The Instantaneous IMEP Method has shown that cylinder friction makes up much less of the overall engine friction than was originally expected. While the Method remains a sensitive tool for understanding in-cylinder events, it may be inappropriate to generalize findings to overall engine friction performance. This would be particularly true when assessing lube oil properties such as viscosity and friction modifier content.
Cylinder-to-cylinder variations in IMEP can exceed 10%. Since FMEP is a relatively small value determined as the difference between much larger IMEP and BMEP measurements, extrapolating engine IMEP data from a single cylinder can introduce substantial error. Therefore, six Kistler pressure transducers (Model 6125A) were machined one in each combustion chamber, adjacent to the spark plug. The four-valve configuration of this engine necessitated the use of extremely small, non-water-cooled, pressure transducers. However, the transducers were mounted through the cylinder head ’s internal cooling passageways, essentially providing constant temperature operation to the bulk of the unit.
While numerous bench tests and motored engines are available for measuring frictional changes and benefits, there is often a strong desire to augment this information with data generated from “real world” engine operation. A research program was begun in early 1996 with the primary objective of developing a testing protocol in which the benefits from engine oil friction modifier could be quantified in a fired-engine stand. This involved developing a method for measuring engine friction, and then applying the technique to measure changes in friction. Work lost to engine friction is the difference between the work generated within the cylinders and the sum of the work available at the flywheel and the work used to run the engine accessories (generator, water pump, power steering pump, air conditioning compressor etc.). While internal work can be determined through the use of cylinder pressure indicators (commonly expressed in terms of mean effective pressure as "indicated mean effective pressure" or IMEP) and flywheel work by the power absorbed by a dynamometer (brake mean effective pressure or BMEP), the work given up to drive accessories is harder to measure. By eliminating or minimizing the number of accessories driven by the engine, friction can be approximated as the difference between the internal work and the work at the flywheel:
A Tektronix high-speed data acquisition unit collected data once every crank angle, triggered by a Kistler shaft encoder. The unit converted analog signals to digital and held the data for the duration of the measuring event, which typically lasted 87 cycles. Once completed, the entire data set was uploaded to a Pentium-based personal computer for processing. The upload data set included the following measurements for each crank angle of each cycle: six cylinder pressures, six cylinder volumes, manifold absolute pressure (MAP), barometric pressure, and dynamometer load. At 720 crank angles per cycle for 87 cycles, the entire data set consists of almost one million data pieces, too large to save as is. The results were processed by first correcting for drift in the pressure transducers by pegging the cylinder pressure
Friction Mean Effective Pressure (FMEP) = IMEP – BMEP 8
measurements at bottom dead center of the intake stroke, for each cylinder, to the MAP reading. The data at each crank angle was then used to make individual cylinder IMEP and PMEP calculations for each cycle. Using the average BMEP measured over the cycle, 87 individual estimates of FMEP were made.
any, were accounted for by normalizing BMEP to IMEP. Power benefits generated by friction modifier use were determined by comparing these normalized power figures between the stabilized regions. 2800
2760
At the beginning of each test, the engine was flushed and fresh lube oil installed. Speed and load conditions were set, and the engine was allowed to stabilize. Data acquisition was started at engine start up and automatically repeated measurements at roughly twominute intervals for the duration of the test. Friction dropped rapidly as the engine warmed up, and stabilized in as little as one hour, depending on speed and load. Stabilization was signified by 6 to 10 data sets with FMEP values within 10 mbar (preferably 5 mbar) of each other. Figure 14 is given as an example. Once stabilized, the FMEP of the base engine oil was measured, then friction modifier was weighed into a syringe and injected into the running engine. Friction reduction resulting from the additive was evident within two minutes. Stabilization at the new friction level typically occurs within five acquisition cycles. The test was either terminated at this point, or a second additive dose was injected.
b 2720 m [ P 2 7 0 0 E 2680 M I 2660
1540 1520 1500 1480
a b [ P E M B
1460
2640
1440
2620
1420
0 :0 0
1400 1 :0 0
2 :0 0
Time [hr] Figur e 15 E n g i ne I M E P & B M E P 1 5 0 0 R M P , 1 / 6 W O T L o a d , R e f e r e n ce O i l
Figure 15. Engine IMEP & BMEP 1500 rpm, 1/6 WOT Load, Reference Oil Overall Engine Friction Test Results – Friction modifier effects were investigated at engine speeds of 1500, 2000 and 3000 rpm at selected loads varying from 1/12 to 3/4 WOT load. The test results are presented in Figure 16. Initial inspection shows that additive-related friction benefits increase with decreasing engine speed, while power benefits increase with decreasing load. Maximum friction changes were observed at 1500 rpm and 1/6 WOT load. Interestingly this speed/load condition is consistent with steady-state driving at suburban speeds, and is similar to the operating conditions of the Sequence VI test (1500 rpm, approximately 6 hp), an engine lubricant testing standard used for fuel economy certification.
610
Friction Modifier Added Friction Modifier Added
490 470 450 1:00
Friction Modifier Added Friction Modifier Added
2600
630
0:00
1580 1560
2740
650
] 590 r a b 570 m [ 550 P E 530 M F 510
1600 <- IMEP -> BMEP
2780
Closer inspection of Figure 16 yields an intuitively obvious finding: the magnitude of engine friction (FMEP) changes are equal to the increases in corrected power (BMEPc). This firmly establishes the relationship between friction and power. Less intuitive is the unusual finding that the magnitude of the friction drop (and hence power increase) is nearly constant over the speed/load ranges investigated. This explains why friction benefits (change on a percentage basis) are maximized at low speeds and power benefits are maximized at low loads: absolute friction levels are lowest at low speeds and absolute power levels are lowest at low loads.
2:00
Time [hr]
Figure 14. Engine FMEP 1500 rpm, 1/6 WOT Load, Reference Oil Effects of the friction modifier were calculated by comparing the average values for each stabilized FMEP region following an injection to the values generated for the base region prior to the initial injection (Figure 14). Since additive effects occurred almost instantly, there was no need to account for drift in engine or lube performance. The method has since been extended to account for time-dependent events, such as viscosity shear, and will be the subject of a future publication.
A series of subsequent tests were conducted at this same 1500 rpm & 1/6 WOT load condition to evaluate test repeatability, the effects of various lube formulations, and ring tension. Using a commercially available 5W-30 mineral oil meeting API Energy Conserving II standards, repeat testing over several days yielded an average FMEP of 562 mbar, +/- 23 mbar at the 95% confidence interval. When used to determine instantaneous changes, sensitivities less than 1% were achieved at the 95% confidence interval.
A correction factor for BMEP was calculated to determine the power increases resulting exclusively from the friction modifier. Ideally, IMEP remains constant under “steady state” conditions once the engine warms up. In reality though, slight changes sometime occur from one stabilized region to another (Figure 15). IMEP changes directly affect changes in BMEP, so if more or less power is generated in the engine, a similar increase or decrease is seen at the flywheel. Friction losses are essentially constant over the small variations observed in IMEP at a given steady-state condition. These changes in IMEP, if
Examples of friction measurements with a variety of different commercial lubes are presented in Figure 17. The FMEP of four commercial engine oils varied from 9
488 – 604 mbar, and the effectiveness of friction modifier, at constant dosage, varied between 6-11%. Power increases ranged from 2.5-4%, and followed the observed friction reductions.
three times greater tension. The molybdenum disulfide coatings were also removed from the piston skirts. After break-in, tests were conducted using the reference oil at 1500 rpm & 1/6 WOT load. FMEP increased 29% to 724 mbar. Addition of friction modifier resulted in FMEP reductions of 16% (compared to 11% for stock pistons), suggesting that advances in piston design have reduced rubbing friction.
To get a qualitative estimate for the benefits associated with low-friction piston design, the stock pistons were replaced with a duplicates equipped with flat-faced, gapless compression rings, and oil control rings with
RPM Torque, ft*lbs
Test Conditions 1500 Power, hp 90 Rated Load Test Results
Base Condition, mbar FM Added, mbar Change, mbar Change, %
25.7 3/4
BMEPc 5865 5930 65
FMEP 589 522 -67
1.1%
-11.4%
RPM Torque, ft*lbs
Base Condition, mbar FM Added, mbar Change, mbar Change, % RPM Torque, ft*lbs
Test Conditions 2000 Power, hp 60 Rated Load Test Results
Base Condition, mbar FM Added, mbar Change, mbar Change, % RPM Torque, ft*lbs
Test Conditions 1500 Power, hp 20 Rated Load Test Results
Base Condition, mbar FM Added, mbar Change, mbar Change, %
Test Conditions 3000 Power, hp 90 Rated Load Test Results BMEPc 6031 6100 69
FMEP 860 792 -68
1.1%
-7.9%
22.8 1/2
BMEPc 3875 3939 64
FMEP 713 645 -68
1.7%
-9.5%
5.7 1/6
RPM Torque, ft*lbs
Test Conditions 3000 Power, hp 10 Rated Load Test Results
BMEPc 1460 1525 65
FMEP 523 458 -65
Base Condition, mbar FM Added, mbar Change, mbar
4.5%
-12.4%
Change, %
FMEP 752 671 -81
9.8%
-10.8%
800 Neat Lube Oil
700 ] r 6 5 0 a b m [ P E M F
600
w/ added Friction Modifier
Change from Lube A ->
Effect of Friction Mo difier
8%
6%
500
-14%
-6 % -8 %
550
-11% -9 %
450 400 L ube A (5 W -3 0 )
L ube B (2 0 W -5 0 )
L ube C (5 W -3 0 )
L ube D (5 W -3 0 )
Figure 17. Effect of Lube Oil Formulation on FMEP 1500 RPM & 1/6 WOT Load
10
5.7 1/12
BMEPc 853 937 84
Figure 16. Friction and Power Measurements at Various Speeds & Loads
750
51.4 3/4
SUMMARY AND CONCLUSIONS
6. Hartfield-Wunsch, S., Tung, S. and C. Rivard, “Development of a Bench Test for the Evaluation of Engine Cylinder Components and the Correlation with Engine Test Results, ” SAE 1993 Transactions, Section 3, P. 1131-1138, Paper No. 932693, October 15, 1993.
• A modified Cameron-Plint High-Frequency Friction and Wear Tester was used to evaluate the effect of engine materials and energy-conserving engine oils on friction and wear of the bore/piston ring interface.
7. Hartfield-Wunsch, S. and S. Tung, “ The Effect of Microstructures on the Wear Behavior of Thermal Sprayed Coatings,” Reprint from the 1994 7th Thermal Spray Conference Proceedings, Boston, Massachusetts, June 20-24, 1994.
• Friction coefficient data from bench tests correlated well with vehicle-dynamometer fuel economy data as measured in GM vehicles. • The Instantaneous IMEP Method has been successfully applied to higher engine speeds and modern, low-friction engine designs. Comparison of the measured cylinder and engine friction figures generated since its development shows great advances on the part of engine builders to reduce friction, on the order of 80-85%.
8. Fessenden, K. S., Zurecki, Z., and T. P. Slavin, “ Thermal Sprayed Coatings: Properties, Processes, and Applications”, ASM Materials Park, Ohio, 1991. 9. Tung, S., and S. Tseregounis, “An Investigation of Tribological Characteristics of Energy-Conserving Engine Oils Using A Reciprocating Bench Test, ” To be presented at the SAE Spring Fuels and Lubricants Conference, June 20, 2000.
• Based on numerous tests using the Instantaneous IMEP Method, cylinder friction makes up about 20 – 40% of overall engine friction, depending on speed and load.
10. Tseregounis, S. and M. McMillan, “Engine Oil Effects on Fuel Economy in GM Vehicles - Comparison with the ASTM Sequence VI-A Engine Dynamometer Test,” SAE Paper No. 952347, 1995 SAE Fuels and Lubricants Meeting, October 16-19, 1995.
• A method has been developed to determine overall engine friction capable of measuring differences in engine friction as small as 0.4%. The method has been successfully used to evaluate the effects of friction modifiers and lube oil formulations.
11. McGeehan, J. A., “A Literature Review of the Effects of Piston and Ring Friction and Lubricating Oil Properties,” SAE Paper No. 780673, 1978.
• Consistent with expectations, measurements confirm that decreases in friction result in increases of similar absolute magnitude in power at constant throttle conditions.
12. Furuhama, S. and M. Takiguchi, “Measurement of Piston Frictional Force in Actual Operating Diesel Engine,” SAE Paper No. 790855, 1979.
• For the friction modifier evaluated, absolute changes in friction levels, resulting from friction modifier use, remained constant over a wide range of engine speed/load conditions. On a percentage basis, friction reduction was maximized at low speed conditions, while power was maximized at low load conditions.
14. Sherrington, I and E. H. Smith, “The Measurement of Piston-Ring Friction by the ‘Floating-Liner’ Method,” SAE Paper No. 884707, 1988.
13. Furuhama, S. et al., “Effect of Piston and Piston Ring Designs on the Piston Friction Force in Diesel Engines,” SAE Paper No. 810977, 1981.
15. Kitahara, T. et al., “Studies on the Characteristics of Piston Ring Friction,” SAE Paper No. 928434, 1992. 16. Uras, H. M. and D. J. Patterson, and “Measurement of Piston and Ring Assembly Friction – Instantaneous IMEP Method, ” SAE Paper No. 830416, 1983.
REFERENCES 1. Benzing, R., M. Peterson, et.al, “Friction and Wear Devices, ” ASLE Publication, Park Ridge, Illinois, 1976.
17. Uras, H. M. and D. J. Patterson, “Effect of Some Lubricant and Engine Variables on Instantaneous Piston and Ring Assembly Friction ”,” SAE Paper No. 840178, 1984.
2. ASTM G99 Standard: Test Method for Ultra Testing with a Pin-On-Disk Apparatus, ASTM, PA, 1990.
18. Uras, H. M. and D. J. Patterson, “Oil and Ring Effects on Piston Ring Assembly Friction by the Instantaneous IMEP Method, ” SAE Paper No. 850440, 1985.
3. ASTM G77 Standard: Test Method for Ranking Materials to Sliding Wear Using Block-On-Ring Wear Tester, ASTM, PA,1983. 4. United States Steel (USS) Lubrication Engineers Manual, Edited by Charles Bailey and Joseph Aarons, 1st Edition, 1971.
19. Uras, H. M. and D. J. Patterson, “Effect of Some Piston Variables on Piston and Ring Assembly Friction,” SAE Paper No. 870088, 1987.
5. Patterson, D., Hill, S., and S. Tung, ”Bench Wear Testing of Engine Power Cylinder Components, ” Presented at the ASME Fall Technical Conference, Muskegon, Michigan, 1991.
20. Clampitt, G. D., “The Effects of Lightweight Reciprocating Components on Engine Friction, Heat Transfer, and Performance, ” Master's Thesis, D. N. Assanis (Advisor), University of Illinois, May 1991.
11
