Full-Scale Engine Endurance Test of Swift Enterprises UL102 Fuel
Full-Scale Engine Endurance Test of Swift Enterprises UL102 Fuel
Full-Scale Engine Endurance Test of Swift Enterprises UL102 Fuel
David Atwood
July 2010
Final Report
12. Sponsoring Agency Name and Address 13. Type of Report and Period Covered
U.S. Department of Transportation Final Report
Federal Aviation Administration
Air Traffic Organization NextGen & Operations Planning
Office of Research and Technology Development
Washington, DC 20591 14. Sponsoring Agency Code
ANE-110
15. Supplementary Notes
16. Abstract
Researchers at the Federal Aviation Administration William J. Hughes Technical Center Aviation Fuel and Engine Test Facility
(AFETF) entered a Cooperative Research and Development Agreement (CRDA) with Swift Enterprises. Swift has plans to produce a
high-octane, nonoxygenated, binary fuel from biomass for spark ignition, piston aircraft engines. Due to the binary nature of the fuel
and the use of heavy aromatics, the Swift fuel will require engine and airframe testing to determine the performance effects of
noncompliance with the current aviation gasoline specification properties regarding distillation slope. This research was performed on a
Swift binary blend, manufactured in a refinery and not in bio-process, to determine whether there were any initial major engine
performance-related findings that would prevent further research into the use of a Swift binary blend of these components. It is noted
that trace materials found in this Swift binary blend may differ from those that arise from manufacturing of the components from a
biomass fermentation process and will probably differ depending on sugar source and bio-enzymes selected. Trace materials, even at the
100-ppm level, can impact engine performance.
A Lycoming IO540-K, six-cylinder, 300-horsepower, piston aircraft engine was tested using the Swift fuel for a total of 150 hours. The
remanufactured engine was torn down and the critical, high-stress components were measured by Textron Lycoming and shipped to the
AFETF. The engine was broken in and operated solely on the Swift binary blend at the AFETF. After of the endurance test was
completed, the engine was sent back to Textron Lycoming for teardown, inspection, and remeasurement. The tests followed the Title 14
Code of Federal Regulation 33.49 endurance test block power setting procedures. The Swift binary blend weighed approximately 1
pound per gallon more than 100 low-lead (100LL) and contained 96.7% of the specific energy content, thus on a volumetric basis, the
Swift fuel contained 13% more energy per gallon than 100LL. The Swift binary blend did not meet the current aviation gasoline
specification ASTM D 910 at the 50%, 90%, or end distillation points. The endurance test results indicated that the engine experienced
normal levels of engine wear during the 150-hour test, and combustion, oil, and fuel deposits were light. Marks in the secondary fuel
pump diaphragm were discovered, and the fuel pump pressure remained within specification limits throughout the test. It was difficult to
start the engine the mornings after the engine was left to sit overnight. When the engine was warm, it started immediately. The effects
on aircraft structure or pilot operating handbooks from the additional fuel weight and fuel aging effects were beyond the scope of this
research and were not addressed. Previous AFETF research addressed the detonation and power performance of the Swift binary blend.
It is recommended that Swift perform additional engine, airframe, and fuel pump rig tests using a fuel made in their biomass process and
perform the engine tests using a straight weight mineral oil.
17. Key Words 18. Distribution Statement
Aviation, Avgas, High octane, Unleaded fuels, Piston engine, This document is available to the U.S. public through the National
Aircraft, Alternative fuels, Endurance, IO540-K engine, EPA, Technical Information Service (NTIS), Springfield, Virginia
Clean Air Act, 100LL, Biofuel, Swift, Binary blend fuel, UL102 22161. This document is also available from the Federal Aviation
Administration William J. Hughes Technical Center at
actlibrary.tc.faa.gov.
19. Security Classif. (of this report) 20. Security Classif. (of this page) 21. No. of Pages 22. Price
Unclassified Unclassified 70
Form DOT F 1700.7 (8-72) Reproduction of completed page authorized
TABLE OF CONTENTS
Page
EXECUTIVE SUMMARY ix
1. INTRODUCTION 1
1.1 Purpose 1
1.2 Background 1
1.3 Related Documents 2
2. DISCUSSION 3
3. ANALYSES 10
5. REFERENCES 59
iii
LIST OF FIGURES
Figure Page
iv
22 Piston Rings and Side-Clearance Measurement Locations 40
33 Intake and Exhaust Valve Pictures Taken at the end of Endurance Test
(cylinders 1 through 3) 48
34 Intake and Exhaust Valve Pictures Taken at the end of Endurance Test
(cylinders 4 through 6) 49
38 Oil Pump Housing, Crankcase Housing Bottom, Oil Sump and Induction Housing,
and Accessory Housing Showing Light Deposit Formation 53
v
LIST OF TABLES
Table Page
8 Kinematic Viscosity of Base Oil With Varying Dilution Levels of 100LL and
Swift Binary Blend Fuel 18
11 Valve Measurements 21
vi
23 Piston Land, Skirt, and Boss Measurements for Cylinders 3 and 4 34
vii
LIST OF ACRONYMS
viii
EXECUTIVE SUMMARY
Researchers at the Federal Aviation Administration (FAA) William J. Hughes Technical Center
Aviation Fuel and Engine Test Facility (AFETF) have considerable experience performing
engine tests on unleaded aviation fuels. In 2008, the AFETF entered a Cooperative Research
and Development Agreement with Swift Enterprises, Ltd. Swift has plans to produce a high-
octane, nonoxygenated, binary fuel from biomass fermentation process for spark ignition, piston
aircraft engines. Due to the binary nature of the fuel and the use of heavy aromatics, the Swift
fuel requires further engine and airframe testing to determine the performance effects of
noncompliance with the current aviation gasoline specification properties regarding distillation
slope. This research was performed on a Swift binary blend, manufactured in a refinery, not in a
bioprocess, to determine whether there were any initial major engine performance issues that
would prevent further research into the use of a Swift binary blend of these components. It was
noted that trace materials found in this Swift binary blend may differ from those that arise from
manufacturing the components from a biomass fermentation process and may vary depending on
sugar source and bio-enzymes selected. Trace materials, even at the 100-ppm level, can have an
effect on engine performance and safety. This potential variation was not addressed by this
research.
This work is a follow-on to the detonation and power performance tests that were performed at
the AFETF on Swift’s binary blend. The primary goal of these tests was to determine if using
the suggested high-aromatic hydrocarbon component would result in excessive combustion
chamber deposits, spark plug fouling, excessive valve seat wear, or materials compatibility
issues.
Researchers at the AFETF completed a 150-hour endurance test on a Swift binary blend supplied
by Swift Enterprises that was formulated in accordance with a proposed Swift binary blend
ASTM specification and consisted of approximately 85% aromatic hydrocarbon (mesitylene or
1,3,5 trimethyl benzene) and 15% pentane isomerate added to achieve vapor pressure. Swift
Enterprises supplied 4500 gallons of the Swift binary blend, which was stored in a tanker truck
on-site during the tests.
The tests were performed in coordination with engineers at the Textron Lycoming Company, and
a remanufactured six-cylinder, 300-horsepower, Lycoming IO540-K model engine was used.
Lycoming researchers performed initial engine measurements and reassembled the engine for
testing at the AFETF. The Lycoming engine was broken in on the supplied Swift binary blend,
and all operations were performed with this fuel. At the completion of the tests, the engine was
sent back to Lycoming to be dissembled, inspected, and remeasured.
There was no indication of excessive wear on any of the high-contact, high-stress parts of the
engine, and the engine oil analyses showed minimal fuel dilution. There was no evidence of
excess fuel nozzle deposits or fuel maldistribution. Cylinder combustion deposits, including
spark plugs, valves, and piston face deposits, were light. Varnish and sludge buildup were light.
The fuel pump secondary diaphragm showed stretch marks. The fuel pump outlet pressure
slowly degraded but remained within specification at the end of the test. Starting difficulty was
experienced in the mornings after the engine was left to sit overnight. It is recommended that
ix
fuel pump diaphragm rig and soak tests be performed with a high-content mesitylene blend.
Starting with a warm engine was immediate.
This research did not address fit-for-purpose testing that Swift will need to do on their binary
blend made from biomass, nor did it address research requirements requested by voluntary
consensus standards bodies for an industry-recognized specification. The FAA does not approve
fuels, and this testing should not, in any way, be misinterpreted as FAA endorsement of a fuel.
Further, this testing may not conform to Standard Practice for Evaluating the Compatibility of
Proposed Fuel or Additives With Aviation Otto Cycle Fuels and Aircraft Fuel System Materials
currently being developed by ASTM International.
It is recommended that further testing be done by Swift Enterprises on their blend made from
their biomass process and conforming to an approved commercial ASTM fuel specification. It is
also recommended that further testing be done in conformance to the aviation gasoline approval
protocol currently being developed by the ASTM International voluntary consensus standards
body.
x
1. INTRODUCTION.
1.1 PURPOSE.
The endurance performance of a high-octane, unleaded UL102 fuel was tested over a 150-hour
period in a naturally aspirated, fuel-injected Lycoming IO540-K engine. The endurance test
used the six-cylinder Lycoming IO540-K model engine, which had 300 horsepower, 8.7:1
compression ratio. (The IO in the engine model description refers to fuel injection and opposed
cylinder and the numerical value refers to the cylinder displacement.) The purpose of this full-
scale engine endurance test was to determine if there were any major engine performance issues
that arose with the use of the particular aromatic hydrocarbon-content fuel that would prevent
Swift from further pursuing their biomass fuel. Issues of oil dilution, nozzle and fuel system
deposits, engine wear, oil system sludge buildup, combustion and valve deposits, materials
compatibility, and start-ability were reviewed.
1.2 BACKGROUND.
Researchers at the Federal Aviation Administration (FAA) Aviation Fuel and Engine Test
Facility (AFETF) at the FAA William J. Hughes Technical Center have been providing
independent, full-scale engine test data on unleaded fuels to facilitate the development of a safe,
high-octane, unleaded aviation gasoline for spark ignition, piston aircraft engines. The AFETF
works closely with FAA certification officials, the Coordinating Research Council Unleaded
Aviation Gasoline Committee, the ASTM International J0.2 subcommittee on aviation gasoline,
and the General Aviation Stakeholders Coalition Future Aviation Gasoline Strategy and
Transition Plan.
Federal Clean Air Act Amendments prohibit the sale of leaded fuels for on-road vehicles. As a
result, the general aviation (GA) community is the largest domestic consumer of leaded fuels
and, according to the Environmental Protection Agency (EPA), GA is responsible for 45% of the
ambient air lead inventory. The Bluewater Network filed a petition with the EPA that called for
either a complete ban on leaded aviation fuels or a commission to study of the effects of leaded
aviation fuel on public health. This petition has significantly increased domestic environmental
pressures against the use of leaded aviation fuels. An Advanced Notice of Proposed Rule
Making was released on April 28, 2010 by the EPA in response to the Bluewater Network
petition. In a related action, the EPA recently reduced the National Ambient Air Quality
Standard (NAAQS) for lead by 90%, from 1500 nanograms per cubic meter of air (ng/m3) to
150 ng/m3. This significant reduction in the lead NAAQS resulted in the requirement for source-
oriented monitoring by individual states to be operational by January 1, 2010.
Researchers at the AFETF entered into a Cooperative Research and Development Agreement
(CRDA) with Swift Enterprises, Ltd. of Indiana. Under the CRDA, Swift Enterprises supplied
4500 gallons of a high-octane, Swift binary blend. The Swift binary blend was petroleum-based
and not made in their biomass process. Swift Enterprises requested that the AFETF perform
initial engine endurance tests on their Swift binary blend to determine whether there were any
major engine performance-related findings that would prevent their continued pursuit of this
endeavor.
1
The researchers at the AFETF have previously performed detonation and power performance
tests on a Swift binary blend in both a turbocharged Lycoming TIO540-J2BD and a naturally
aspirated Lycoming IO540-K engine. The results were published in an FAA report [1].
A 150-hour engine endurance test, following the Title 14 Code of Federal Regulations (CFR)
33.47 150-hour block test requirements, was performed. Testing was performed in coordination
with engineers at Textron Lycoming. Lycoming supplied a remanufactured IO540-K engine and
performed the initial teardown measurements. The engine was broken in and operated solely on
the Swift-supplied fuel at the AFETF. The tests were performed to provide a cursory review of
issues of wear, performance, materials compatibility, deposit formation, start-ability, and other
issues. The majority of the tests were performed at full-rated power and engine speed under
maximum cylinder head and oil temperatures.
• ASTM D 909, “Standard Test Method for Knock Characteristics of Motor and Aviation
Fuels by the Supercharge Method.”
• ASTM D 2700, “Standard Test Method for Knock Characteristics of Motor and Aviation
Fuels by the Motor Method.”
• ASTM D 6424, “Standard Practice for Octane Rating Naturally Aspirated Spark Ignition
Aircraft Engines.”
• ASTM D 445, “Standard Test Method for Kinematic Viscosity of Transparent and
Opaque Liquids (and the Calculation of Dynamic Viscosity).”
• ASTM D 664, “Standard Test Method for Acid Number of Petroleum Products by
Potentiometric Titration.”
• ASTM D 3524, Standard Test Method for Diesel Fuel Diluent in Used Diesel Engine
Oils by Gas Chromatography.
• FAA Advisory Circular 20-24B, “Qualification of Fuels, Lubricants, and Additives for
Aircraft Engines.”
2
2. DISCUSSION.
Fuel samples were sent to an independent laboratory for comparison against the current ASTM
D 910 leaded aviation gasoline specification, as shown in table 1. The first column lists the
ASTM test method, and the last column on the right lists the current specification limits for the
leaded aviation gasoline D 910 for 100 low-lead (100LL). The items in red indicate differences
with the current ASTM D 910 specification. The Swift binary blend had a specific energy
content that was 3.7% lower than the current 100LL specification. The potential consequences
of reduced energy content are reduced range and increased fuel consumption. However, due to
the much higher mass density of the fuel, the volumetric fuel consumption is lower. Therefore,
in the Swift binary blend case, the 3.7% reduction in specific energy content may not be as
significant. Effects of increased fuel weight, if any, on the structural integrity of the airframe
and on potential revisions to pilot operating handbook performance curves were beyond the
scope of this research.
3
Table 1. Swift Binary Blend Component Specification
Note: Values in red deviate from current ASTM D 910 specification for 100LL.
MJ = Mega Joules kPa = kilopascal BRE = Bracketing equilibrium method
kg/m3 = Kilogram per cubic meter mL = milliliter TEL = Tetraethyl lead
The fuel also did not match the distillation curve points at the T50, T90, and end point. A high-
aromatic content unleaded fuel is not likely to meet the points at the high end of the distillation
curve developed for a conventional 100LL. The aromatic, high-octane components are typically
of high mass density and high boiling point. Previous FAA tests, in conjunction with the
4
Coordinating Research Council showed equivalent unleaded fuel detonation performance to the
leaded 100LL was only attainable with an addition of 6% to 10% volume-to-volume aromatic
amine in aviation alkylate or with a high-aromatic hydrocarbon-content fuel [2]. It is unlikely
that such a fuel will meet the distillation curve parameters that have evolved for the current
performance of an aviation alkylate with lead.
The purpose of this full-scale engine endurance test was to determine if there were any major
engine performance issues that arose with the use of the particular aromatic hydrocarbon-content
fuel that would prevent Swift from further pursuing their biomass fuel. Issues of oil dilution,
nozzle and fuel system deposits, engine wear, oil system sludge buildup, combustion and valve
deposits, materials compatibility, and start-ability were reviewed.
The potential effects of having a high 90% and end distillation point fuel includes excessive
combustion chamber, valve, and spark plug deposits from incomplete combustion; excessive oil
dilution; fuel maldistribution; fuel system deposits; heavy intake deposits; and cold starting
difficulty. Swift has addressed some of these by using an aromatic hydrocarbon, mesitylene,
which has strong solvent properties that may act to counter many of the above effects.
The Swift binary blend had an ASTM D 2700 motor octane number (MON) of 102.2 and an
ASTM D 909 supercharge rich rating of 139.6, which were above the current specification
minimums for 100LL. Further, the 102.2 MON was more than 2.5 MON higher than the 100LL
minimum MON requirement to account for the differences in engine performance between
leaded and unleaded fuels [3].
Table 2 shows the compositional analyses of the Swift binary blend by gas chromatography
(GC) flame ionization detection (FID).
Component Value
Mesitylene, (1,3,5-Trimethylbenzene) 0082.50 mass %
Iso-pentane 0016.59 mass %
Isobutane 0112.3 ppm
Butane 0131.6 ppm
Neopentane 0376.0 ppm
Pentane 0442.5 ppm
Isooctane 0927.5 ppm
Toluene 0238.2 ppm
p-Xylene 0032.8 ppm
o-Xylene 0451.6 ppm
Cumene 0042.6 ppm
4-Ethyltoluene 1229.8 ppm
5
Table 2. Swift Binary Blend Component Composition by GC-FID (Continued)
Component Value
2-Ethyltoluene 0192.3 ppm
1,2,4-Trimethylbenzene 4352.6 ppm
1,2,3-Trimethylbenzene 0014.0 ppm
Other 0511.6 ppm
Note that the Swift binary blend consists of two main components, 82.5% by mass aromatic
hydrocarbon and 16.6% by mass iso-pentane (volatility component), with approximately 0.9%
being typical aviation gasoline component hydrocarbons.
Traditionally, engines and airframes were certified on available leaded fuels that met ASTM
D 910. The actual unleaded fuel motor octane requirement for the portion of the fleet that
consumes the majority of the aviation gasoline for unleaded fuels is more than 100 MON. The
performance of unleaded fuels in spark ignition, piston aircraft engines can vary significantly
from the performance of traditional leaded aviation gasoline [3].
Engine parameter data were recorded at a rate of one full channel scan every second. The engine
was equipped with at least the following sensors:
All the sensors were installed at the manufacturer’s recommended locations whenever possible.
All sensors and thermocouple channels were calibrated through the data acquisition system prior
to engine testing, and the calibrations were verified at the end of the test. The engine was also
equipped with accelerometers to ensure engine detonation was not occurring.
6
All engine break-in and acceptance tests were performed with the Swift-supplied fuel. Prior to
and after the completion of the 150-hour test, power baseline tests were performed. At the
completion of all tests, the engine was sent back to Textron Lycoming, measured again, and
compared to the original engine measurements taken after remanufacture. FAA researchers,
Swift Enterprises representatives, and Lycoming engineers were present to witness and
document the engine tear down.
A series of maintenance runs was performed to verify the engine system’s integrity, proper fuel
and oil pressures, and instrumentation accuracy. Following these test runs, a series of
acceptance and oil consumption tests was completed. All the test runs were performed using
Phillips Type M 20W-50 nondispersant mineral oil. Prior to any engine operation, the mixture
cutoff, full-rich settings, and the throttle idle stop and throw positions were checked. All engine
operations throughout this test were performed with the Swift binary blend; the engine was not
operated on leaded fuel at any time during the test.
All engine operations following the break-in period were performed with Aeroshell 15W-50
multi-viscosity oil and any servicing of the engine with oil was recorded. It is likely that future
endurance tests would be performed using a straight mineral oil without additives. However,
details of a test protocol for alternative aviation gasoline approval are still being formulated at
ASTM International. For the purpose of this test, data were available from other unleaded fuel
endurance tests (performed at the AFETF) with this oil, which can be used for comparison.
The health of the engine was then verified. A dynamic engine power baseline test was
performed by varying the manifold pressure from 20 inches of mercury (inHg) to full-throttle
stop and by varying the revolutions per minute (rpm) from 2000 to 2700 rpm. The manifold
absolute pressure (MAP) was varied by 2 inHg increments, whereas the rpm was varied by 100-
rpm increments.
These power baselines encompass a combination of MAP settings and engine rpm settings over a
practical operating envelope in set increments. The results from the baselines verify the health
of the engine. A MAP setting is chosen, and the engine power data are collected for each
subsequent engine rpm. The MAP is then changed and the process repeated until the engine
power production data have been collected for all combinations of MAP and rpm.
The endurance block test was complied with 14 CFR 33.47 and immediately followed the power
baseline test. The separate phases of the 150-hour endurance test are shown in table 3. It should
be noted that total engine time (which includes start-up, warm-up, transition, and shutdown) is
not the same as the engine point time (only the amount of time the engine is at the specified
power setting).
7
Table 3. Phase Breakdown of Endurance Test
The Lycoming IO540-K engine was equipped with a spin-on oil filter, allowing 50-hour
maintenance intervals. After every 50 hours of engine operation, a scheduled inspection was
performed. (This does not correspond to 50 hours of engine on-point test time.) These
inspections included at least the following:
• A compression test was performed as soon as possible after the engine shutdown to
ensure that the piston rings, cylinder walls, and other parts were well lubricated and at
running tolerance. The test was performed using procedures and a differential pressure
tester incorporating a master orifice device as described in Lycoming Service Instruction
1191a. Eighty psi of air pressure was applied to the cylinder while the pressure across
the orifice device was measured with the piston at top-dead-center of the compression
stroke.
• Oil sampling procedures followed those outlined in the Society of Automotive Engineers
lubricating oil qualification test J1899. The oil system was drained and the oil pump
filter screen was removed and inspected for metal particles and contamination. The
screen was thoroughly cleaned, reinstalled with new gaskets, and safety wired. The
system was serviced to the proper level with Aeroshell 15W-50 multi-viscosity oil.
• The oil samples were analyzed by an independent laboratory, as per ASTM D 445 for
viscosity, ASTM D 664 for total acid number, ASTM D 3524 for fuel dilution, and
ASTM D 5185 for metals content.
8
• After the engine cooled, a special inspection to monitor cylinder valve stem height was
performed. The valve covers were removed and any valve cover gasket material were
removed from the cylinder head. The rocker pin retaining bolts were removed, the pins
pulled out, and the rocker arms removed. To ensure the valves were properly seated, a
rubber mallet was used to carefully tap the valve stem heads. A special measurement
plate was mounted to the cylinder head. Figure 2 shows the installation of a similar
measurement plate on a Continental engine. With the measurement plate installed, the
total valve train recession was obtained through the guide holes using a depth gauge, and
the data were recorded as valve stem height.
• The fuel metering unit throttle and mixture control levers were inspected for freedom of
movement. A beam-type torque wrench was used to measure the resistance to movement
of the lever arms.
• The spark plugs were inspected for electrode wear, tested for spark strength, rotated, and
reinstalled. They were not cleaned so the deposit buildup could be continuously
monitored. A cylinder borescope was used to ensure that rings, valves, valve surfaces,
piston crown, and cylinder walls were in visibly healthy condition.
• The fuel injection nozzles were removed from the engine, and rubber hoses were
connected to the cylinder side of the injectors. These hoses were all of equal length and
were run to graduated cylinders. The connections to the fuel manifold valve, fuel control
unit, fuel pump, and fuel supply were left intact. The fuel tank boost pump was run, and
the amount of fuel in each graduated cylinder was measured and recorded. The test was
repeated three times, and the average of the tests was compared to previous values and to
each other. The nozzles were also inspected using a mechanics light for visible deposit
formation. The nozzles were reinstalled without cleaning.
• All fuel and oil lines were inspected for stiffness or excessive pliability. Hose fittings
were also inspected for leaks or deposit buildup.
• The fuel inlet screen (finger screen) was removed, cleaned, reinstalled, and safety wired.
The system was pressure checked for evidence of leaks at the sealing gasket.
• The engine cylinder assembly was inspected for evidence of overheating, leakage
between exhaust ports and pipes, and warped exhaust port flanges. The baffling was
inspected for condition and security. The fabricated sheet metal cowling was inspected
for cracks, bowing, or other evidence of pressure-cycling fatigue.
9
Figure 2. Valve Stem Height Measurement Tool (Typical)
Upon completion of the inspection, the engine was serviced with oil and inspected for evidence
of oil and fuel leaks.
3. ANALYSES.
Figure 4 shows the Swift binary blend response of MON and Reid Vapor Pressure (RVP) versus
mesitylene content. The mesitylene is a high-MON, low-RVP component, whereas the iso-
pentane has a high RVP and lower MON. When the mesitylene content is higher, the MON is
greater and the RVP of the Swift binary blend is lower. Conversely, when the iso-pentane
component is higher, the MON is lower but the RVP of the Swift binary blend is higher.
The significance of this particular Swift binary blend is that the aging effects of the fuel are such
that, as the iso-pentane of the Swift binary blend evaporates, the octane will increase and the
vapor pressure will decrease. The boundaries of possible Swift binary blend composition will
have to be fully evaluated, both at the high-octane, low-RVP and high-RVP, low-octane limits of
the final fuel specification. Fuel that has been left in vented storage tanks over time may degrade
start-ability and could make altitude relights or engine throttle response more difficult. The
current specification for 100LL is ASTM D 910, which lists the allowable vapor pressure range
as 5.5 to 7.0 psig. According to figure 3, this vapor pressure range would limit the mesitylene
content between 80% and 85% of the Swift binary blend.
10
105.0 10.00
104.5 9.00
104.0 8.00
103.0 6.00
102.5 5.00
102.0 4.00
101.5 3.00
MON (ASTM D 2700)
101.0 2.00
RVP (psig)
100.5 1.00
100.0 0.00
75 77.5 80 82.5 85
Mesitylene (% v/v)
Figure 4 shows the distillation slope of the Swift binary blend fuel compared to a locally
purchased 100LL fuel. Also included on the graph (in red) are the current ASTM D 910
distillation specification limits for a leaded, hydrocarbon-based, crude-derived aviation gasoline.
These limits evolved by industry consensus from extensive experience with traditional leaded
hydrocarbon fuels and their effect on aircraft and aircraft engine performance. Deviations from
this known slope will require testing to address effects on start-ability, vapor lock, hot fuel,
engine combustion deposits, gum formation, varnish buildup, oil dilution, valve sticking, fuel
atomization, fuel mal-distribution, throttle and mixture response, cold-weather operation, and
altitude relight. Another effect of the high distillation temperatures is that the oil temperatures
are not high enough to boil off the fuel that migrates into the oil; thus, the majority of the fuel
that ends up in the oil will remain, which diminishes the lubricating effects of the oil, until the
next oil change. In this research, the oil dilution remained less than 1%.
11
180.0
170.0 max
160.0
150.0
140.0
max
130.0
120.0
Temperature (deg. C)
110.0
max
100.0
90.0
80.0 max min
70.0
60.0
50.0
100LL
40.0
30.0 Swift
D 910 Temperature Limits 97% min
20.0
recovery
10.0
0.0
0 10 20 30 40 50 60 70 80 90 100
Percent Evaporated (v/v)
Table 4 contains the lubricating engine oil analyses at 50 engine-hour intervals. The trace metal
content of iron, silicon, lead, copper, nickel, and tin showed levels higher than typically observed
in past experience for this type of test. It was noted during teardown that the valve tappet faces
in four of the cylinders showed excessive wear (see figure 5). This was not attributed to the fuel
as accelerated wear had been found on this type of tappet material and design, which is no longer
used by Lycoming. This could account for the elevated iron and silicon levels. The lead, nickel,
and tin are typically found in the crankshaft bearings, all of which showed normal wear. The
elevated copper levels could have been from the wearing of the crankshaft bearing coatings and
exposure of the copper and also from the edges of the components with copper plating, including
the crankshaft, camshaft, and some gears. All the metal trace levels peaked at the 150 engine-
hour mark and proceeded to decrease to the 180 engine-hour mark, which is a typical wear trend
pattern. The acid number of the oil also followed this trend.
12
Table 4. Engine Lubricating Oil Analyses
Table 5 shows the main bearing measurements, and figure 6 shows the measurement locations.
The largest bearing thickness change was 1.8 thousandths of an inch in the #1 bottom B location.
All the other measurements were less than a thousandth of an inch change, which are normal
wear levels.
13
Table 5. Main Bearing Thickness Measurements for Numbers 1 and 2
Top Bottom
Bearing After After
No. Location Before Test Change Before Test Change
1 A 0.0947 0.0946 -0.0001 0.0949 0.0945 -0.0004
B 0.0949 0.0945 -0.0004 0.0945 0.0947 0.0002
C 0.0947 0.0917 0.0920 0.0003
D 0.0919 0.0920 0.0001 0.0953
E 0.0950 0.0948 -0.0002 0.0949 0.0950 0.0001
F 0.0951 0.0948 -0.0003 0.0949 0.0947 -0.0002
2 A 0.0952 0.0947 -0.0005 0.0949 0.0946 -0.0003
B 0.0946 0.0946 0.0000 0.0947 0.0943 -0.0004
C 0.0921 0.0920 -0.0001 0.0950
D 0.0948 0.0927 0.0921 -0.0006
E 0.0951 0.0942 -0.0009 0.0947 0.0946 -0.0001
F 0.0950 0.0948 -0.0002 0.0947 0.0947 0.0000
Table 6 shows the main crankshaft bearing measurements for numbers 3, 4, and 5, and figure 7
shows the main crankshaft measurement locations for numbers 3, 4 and 5. This table shows
14
normal bearing wear with less than a thousandth of an inch change in thickness and normal wear
modes.
15
Figure 7. Main Crankshaft Bearings for Numbers 3, 4, and 5 With Associated Drawing Showing
Measurement Locations
Table 7 shows the main crankshaft journal wear, and figure 8 shows the measurement locations.
This table shows that there was no measurable wear in the journals.
Cylinder Dimensions
No. Location Before After Wear
1 A 2.6250 2.6250 0.0000
A' 2.6250 2.6250 0.0000
B 2.6250 2.6250 0.0000
B' 2.6246 2.6246 0.0000
2 A 2.6252 2.6252 0.0000
A' 2.6250 2.6250 0.0000
B 2.6249 2.6249 0.0000
B' 2.6250 2.6250 0.0000
16
Table 7. Main Crankshaft Journal Measurements (Continued)
Cylinder Dimensions
No. Location Before After Wear
3 A 2.6249 2.6249 0.0000
A' 2.6249 2.6249 0.0000
B 2.6252 2.6252 0.0000
B' 2.6250 2.6250 0.0000
4 A 2.6249 2.6249 0.0000
A' 2.6249 2.6249 0.0000
B 2.6250 2.6250 0.0000
B' 2.6250 2.6250 0.0000
5 A 2.6245 2.6245 0.0000
A' 2.6245 2.6245 0.0000
B 2.6246 2.6246 0.0000
B' 2.6246 2.6246 0.0000
As shown in table 4, oil dilution levels from fuel were typically less than 1%. Table 8 shows
viscosity data for varying fuel dilution levels in oil for both 100LL and Swift binary blend. The
data show that the effect on the oil viscosity was relatively the same regardless of whether the
dilution occurred from the Swift binary blend or the 100LL, and at 1% dilution levels, the
viscosity change was minimal. However, as previously shown by the distillation data in figure 4,
the majority of the Swift binary blend (approximately 80%) boils at 162°C (324°F), which is
17
higher than the oil temperatures reach during normal operation. Therefore, the majority of the
Swift binary blend dilution that occurs does not evaporate out of the crankcase breather but
remains in the oil until the oil is replaced. Conversely, more than 50% of the 100LL fuel boils at
a temperature less than 100°C (212°F). If a condition were to develop that resulted in elevated
Swift binary blend oil dilution, the oil lubricity properties could be compromised to a greater
extent than would occur with 100LL.
Table 8. Kinematic Viscosity of Base Oil With Varying Dilution Levels of 100LL and
Swift Binary Blend Fuel
cSt = Centistokes
Valve stem height measurements taken every 50 engine hours during testing show that valve seat
recession was minimal (see tables 9 and 10 and figures 9 and 10). Total valve recession for both
the intake and exhaust valves for any cylinder was less than 3 thousandths of an inch at the end
of the test. One of the reasons this value has been so low is due to the lack of wear grooves on
the exhaust valve faces.
18
3
Cylinder 1 Cylinder 2
Change in Intake Valve Stem Height From Start of Test
Cylinder 3 Cylinder 4
2.5 Cylinder 5 Cylinder 6
1.5
(thousandths of an in.)
0.5
0
0 20 40 60 80 100 120 140 160 180 200
-0.5
-1
-1.5
Engine Hours
19
3
Cylinder 1 Cylinder 2 Cylinder 3
Change in Exhaust Valve Stem Height From Start of Test
2
(thousandths of in.)
1.5
0.5
0
0 20 40 60 80 100 120 140 160 180 200
Engine Hours
Table 11 shows the posttest valve measurements. The measurement locations are shown
graphically in the far right column of the table. As shown in the table, the largest change in
valve length was the intake valve of cylinder 2. This value was still within the manufacturer’s
new limit specifications of 4.580 ±0.007 inches. All other measurements were within the
manufacturer’s new limits, except for the exhaust valve in cylinder 5, which started at the
minimum-allowable diameter and wore 0.0001 inch.
Table 12 shows the valve seat and face run-out measurements, and figure 11 shows the
measurement locations. All the seat run-out measurement changes were less than 0.0013 inch,
except for the exhaust valve in cylinder 1, which showed a 4 thousandths of an inch change.
All the valve face run-out measurements were less than the maximum-allowable level of
0.0015 inch.
20
Table 11. Valve Measurements
21
Table 12. Valve Seat and Face Run-Out Measurements
Seat Face
Cylinder Before After Cylinder Before After
No. Location Test Test Change No. Location Test Test Change
1 Intake 0.0011 0.0011 0.0000 1 Intake 0.0000 0.0000 0.0000
Exhaust 0.0017 0.0057 0.0040 Exhaust 0.0000 0.0005 0.0005
2 Intake 0.0011 0.0022 0.0011 2 Intake 0.0000 0.0000 0.0000
Exhaust 0.0027 0.0038 0.0011 Exhaust 0.0000 0.0005 0.0005
3 Intake 0.0017 0.0023 0.0006 3 Intake 0.0000 0.0000 0.0000
Exhaust 0.0026 0.0027 0.0001 Exhaust 0.0005 0.0005 0.0000
4 Intake 0.0005 0.0015 0.0010 4 Intake 0.0000 0.0000 0.0000
Exhaust 0.0010 0.0013 0.0003 Exhaust 0.0000 0.0010 0.0010
5 Intake 0.0015 0.0022 0.0007 5 Intake 0.0000 0.0000 0.0000
Exhaust 0.0021 0.0018 -0.0003 Exhaust 0.0005 0.0005 0.0000
6 Intake 0.0007 0.0013 0.0006 6 Intake 0.0000 0.0000 0.0000
Exhaust 0.0021 0.0034 0.0013 Exhaust 0.0000 0.0005 0.0005
Table 13 shows the valve guide measurements, and figure 12 shows the measurement locations.
All the intake guides fell out of the manufacturer’s specification range at the A measurement
location. However, these parts are automatically replaced at overhaul, as they are part of the
cylinder assembly. The largest intake valve guide wear was in cylinder 1 with 1.8 thousandths
of an inch, and the largest exhaust valve guide wear was in cylinder 5 with 2.3 thousandths of an
inch. All the other intake valve guides showed wear less than 0.0006 inch and the exhaust valve
guides showed wear less than 1.4 thousandths of an inch. All the clearance measurements for
valve-to-valve guides fell out of specification for at least one measurement location for the
exhaust valves and all the intake valve-to-valve guide clearances were still within new
specification, except for the A location in cylinder 1. All intake clearances were within
serviceable limits of 0.006 inch.
22
Table 13. Valve Guide Measurements
Intake Exhaust
Cylinder Dimensions Cylinder Dimensions
No. Location Before After Wear No. Location Before After Wear
1 A 0.4048 0.4066 0.0018 1 A 0.5003 0.5002 -0.0001
A' 0.4049 0.4063 0.0014 A' 0.5000 0.5004 0.0004
B 0.4048 0.4049 0.0001 B 0.5003 0.5002 -0.0001
B' 0.4048 0.4050 0.0002 B' 0.5003 0.5002 -0.0001
C 0.4047 0.4048 0.0001 C 0.5002 0.5011 0.0009
C' 0.4047 0.4048 0.0001 C' 0.5003 0.5006 0.0003
2 A 0.4046 0.4052 0.0006 2 A 0.5001 0.5000 -0.0001
A' 0.4051 0.4055 0.0004 A' 0.5001 0.5003 0.0002
B 0.4048 0.4049 0.0001 B 0.5003 0.5005 0.0002
B' 0.4048 0.4050 0.0002 B' 0.5003 0.5003 0.0000
C 0.4047 0.4048 0.0001 C 0.5001 0.5014 0.0013
C' 0.4047 0.4048 0.0001 C' 0.5001 0.5001 -0.0000
3 A 0.4047 0.4051 0.0004 3 A 0.5002 0.5001 -0.0001
A' 0.4049 0.4051 0.0002 A' 0.5002 0.4997 -0.0005
B 0.4048 0.4049 0.0001 B 0.5003 0.4999 -0.0004
B' 0.4048 0.4049 0.0001 B' 0.5003 0.5000 -0.0003
C 0.4047 0.4047 0.0000 C 0.5002 0.5021 0.0019
C' 0.4047 0.4048 0.0001 C' 0.5002 0.5007 0.0005
4 A 0.4047 0.4052 0.0005 4 A 0.5006 0.5012 0.0006
A' 0.4050 0.4055 0.0005 A' 0.5004 0.5009 0.0005
B 0.4048 0.4049 0.0001 B 0.5006 0.5004 -0.0002
B' 0.4048 0.4049 0.0001 B' 0.5005 0.5005 0.0000
C 0.4046 0.4047 0.0001 C 0.5004 0.5015 0.0011
C' 0.4046 0.4047 0.0001 C' 0.5004 0.5009 0.0005
5 A 0.4048 0.4052 0.0004 5 A 0.5004 0.5003 -0.0001
A' 0.4049 0.4052 0.0003 A' 0.5002 0.4996 -0.0006
B 0.4048 0.4049 0.0001 B 0.5003 0.5004 0.0001
B' 0.4049 0.4050 0.0001 B' 0.5003 0.5005 0.0002
C 0.4047 0.4048 0.0001 C 0.5002 0.5025 0.0023
C' 0.4047 0.4047 0.0000 C' 0.5002 0.5007 0.0005
6 A 0.4048 0.4052 0.0004 6 A 0.5004 0.5003 -0.0001
A' 0.4049 0.4051 0.0002 A' 0.5000 0.5003 0.0003
B 0.4049 0.4049 0.0000 B 0.5002 0.5006 0.0004
B' 0.4049 0.4049 0.0000 B' 0.5002 0.5003 0.0001
C 0.4046 0.4047 0.0001 C 0.5001 0.5015 0.0014
C' 0.4046 0.4047 0.0001 C' 0.5001 0.5006 0.0005
23
Figure 12. Valve Guide Measurement Locations
The valve tappet bore measurements are shown in table 14, and measurement locations are
shown in figure 13. New and service limits specifications were not shown in the overhaul
manual for this particular engine. The table shows that there was minimal wear in the exhaust
and intake tappet bores.
Intake Exhaust
Cylinder Dimensions Cylinder Dimensions
No. Location Before After Wear No. Location Before After Wear
1 A 0.7228 0.7228 0.0000 1 A 0.7215 0.7215 0.0000
A' 0.7227 0.7227 0.0000 A' 0.7217 0.7217 0.0000
B 0.7215 0.7215 0.0000 B 0.7212 0.7211 -0.0001
B' 0.7214 0.7214 0.0000 B' 0.7213 0.7213 0.0000
2 A 0.7207 0.7209 0.0002 2 A 0.7207 0.7209 0.0002
A' 0.7207 0.7209 0.0002 A' 0.7206 0.7206 0.0000
B 0.7209 0.7207 -0.0002 B 0.7204 0.7204 0.0000
B' 0.7209 0.7207 -0.0002 B' 0.7207 0.7205 -0.0002
3 A 0.7212 0.7213 0.0001 3 A 0.7217 0.7217 0.0000
A' 0.7215 0.7217 0.0002 A' 0.7221 0.7222 0.0001
B 0.7202 0.7204 0.0002 B 0.7213 0.7210 -0.0003
B' 0.7199 0.7202 0.0003 B' 0.7212 0.7214 0.0002
4 A 0.7203 0.7204 0.0001 4 A 0.7197 0.7196 -0.0001
A' 0.7202 0.7206 0.0004 A' 0.7196 0.7196 0.0000
B 0.7212 0.7216 0.0004 B 0.7208 0.7210 0.0002
B' 0.7211 0.7214 0.0003 B' 0.7206 0.7208 0.0002
5 A 0.7195 0.7192 -0.0003 5 A 0.7221 0.7221 0.0000
A' 0.7190 0.7189 -0.0001 A' 0.7224 0.7224 0.0000
B 0.7192 0.7193 0.0001 B 0.7215 0.7214 -0.0001
B' 0.7189 0.7186 -0.0003 B' 0.7216 0.7213 -0.0003
6 A 0.7215 0.7214 -0.0001 6 A 0.7211 0.7212 0.0001
A' 0.7207 0.7206 -0.0001 A' 0.7207 0.7210 0.0003
B 0.7215 0.7215 0.0000 B 0.7206 0.7209 0.0003
B' 0.7215 0.7215 0.0000 B' 0.7204 0.7206 0.0002
24
Figure 13. Valve Tappet Bore Measurement Locations
Table 15 shows the valve tappet measurements, and figure 14 shows the measurement locations.
The C length measurement showed extensive wear in some of the tappets. This was previously
discussed section 3.2.
Intake Exhaust
Cylinder Dimensions Cylinder Dimensions
No. Location Before After Wear No. Location Before After Wear
1 A 0.7174 0.7175 0.0001 1 A 0.7174 0.7175 0.0001
A' 0.7174 0.7175 0.0001 A' 0.7174 0.7175 0.0001
B 0.7174 0.7176 0.0002 B 0.7175 0.7175 0.0000
B' 0.7174 0.7176 0.0002 B' 0.7175 0.7175 0.0000
C 1.3720 1.3700 -0.002 C 1.3720 1.3720 0.0000
2 A 0.7172 0.7173 0.0001 2 A 0.7172 0.7173 0.0001
A' 0.7172 0.7175 0.0003 A' 0.7172 0.7173 0.0001
B 0.7172 0.7174 0.0002 B 0.7173 0.7174 0.0001
B' 0.7172 0.7175 0.0003 B' 0.7173 0.7175 0.0002
C 1.3720 1.3685 -0.0035 C 1.3750 1.3685 -0.0065
3 A 0.7175 0.7176 0.0001 3 A 0.7173 0.7174 0.0001
A' 0.7175 0.7175 0.0000 A' 0.7173 0.7174 0.0001
B 0.7174 0.7177 0.0003 B 0.7172 0.7174 0.0002
B' 0.7174 0.7175 0.0001 B' 0.7172 0.7174 0.0002
C 1.3720 1.3670 -0.0050 C 1.3750 1.3670 -0.0020
4 A 0.7173 0.7174 0.0001 4 A 0.7174 0.7175 0.0001
A' 0.7174 0.7174 0.0000 A' 0.7173 0.7174 0.0001
B 0.7174 0.7174 0.0000 B 0.7175 0.7175 0.0000
B' 0.7174 0.7174 0.0000 B' 0.7175 0.7175 0.0000
C 1.3720 1.3630 -0.0090 C 1.3730 1.3620 -0.0110
25
Table 15. Valve Tappet Measurements (Continued)
Intake Exhaust
Cylinder Dimensions Cylinder Dimensions
No. Location Before After Wear No. Location Before After Wear
5 A 0.7173 0.7174 0.0001 5 A 0.7173 0.7174 0.0001
A' 0.7174 0.7174 0.0000 A' 0.7174 0.7174 0.0001
B 0.7174 0.7175 0.0001 B 0.7174 0.7174 0.0000
B' 0.7174 0.7175 0.0001 B' 0.7174 0.7174 0.0000
C 1.3740 1.3720 -0.0020 C 1.3740 1.3705 -0.0035
6 A 0.7173 0.7174 0.0001 6 A 0.7174 0.7174 0.0000
A' 0.7173 0.7174 0.0001 A' 0.7173 0.7174 0.0001
B 0.7174 0.7173 -0.0001 B 0.7173 0.7174 0.0001
B' 0.7174 0.7173 -0.0001 B' 0.7173 0.7174 0.0001
C 1.3720 1.3690 -0.0030 C 1.3730 1.3680 -0.0050
The maximum serviceable clearance between tappet body and tappet bore is 0.004 inch. The
intake clearances in cylinders 1, 3, 4, and 6 and the exhaust clearances in cylinders 1, 3, and 5
were out of serviceable maximum limits.
Table 16 shows the valve spring tension measurements. The open valve position specifications
for the inner and outer springs are 73-83 lb at 1.330 inches and 114-124 lb at 1.430 inches,
respectively. The closed valve position specifications for the inner and outer springs are 40-44
lb at 1.800 inches and 53-59 lb at 1.930 inches, respectively. All the spring tensions were within
specification limits in the valve open position. The cylinder 4 intake valve outer spring for the
closed valve position was out of specification limits, all the exhaust valves for the valve closed
position were out of specification limits, and the intake valve inner spring in cylinders 2 and 4
were out of specification limits. These out of specification limits were not fuel related. The
valve springs are routinely replaced at overhaul.
26
Table 16. Valve Spring Tension Measurements
The camshaft journal measurements are shown in table 17. The measurement locations are
shown in figure 15. The camshaft journals showed negligible wear.
Table 18 shows the crankshaft pin measurements, and figure 16 shows the crankshaft pins and of
the measurement locations. As shown in the table, no measurable wear had occurred in the
crankshaft pins.
Cylinder Dimensions
No. Location Before After Wear
1 A 1.0275 1.0275 0.0000
A' 1.0275 1.0275 0.0000
B 1.0274 1.0274 0.0000
B' 1.0274 1.0274 0.0000
2 A 1.0272 1.0272 0.0000
A' 1.0273 1.0273 0.0000
B 1.0274 1.0274 0.0000
B' 1.0274 1.0274 0.0000
3 A 1.0273 1.0273 0.0000
A' 1.0273 1.0273 0.0000
B 1.0274 1.0274 0.0000
B' 1.0274 1.0274 0.0000
4 A 1.0278 1.0276 -0.0002
A' 1.0276 1.0276 0.0000
B 1.0273 1.0273 0.0000
B' 1.0274 1.0273 -0.0001
27
Figure 15. Main Camshaft Journal (typical) With Associated Measurement Locations
Cylinder Dimensions
No. Location Before After Wear
1 A 2.2497 2.2497 0.0000
A' 2.2500 2.2500 0.0000
B 2.2497 2.2497 0.0000
B' 2.2497 2.2497 0.0000
2 A 2.2494 2.2494 0.0000
A' 2.2494 2.2494 0.0000
B 2.2494 2.2494 0.0000
B' 2.2494 2.2494 0.0000
3 A 2.2494 2.2494 0.0000
A' 2.2497 2.2497 0.0000
B 2.2494 2.2494 0.0000
B' 2.2497 2.2497 0.0000
4 A 2.2498 2.2497 -0.0001
A' 2.2497 2.2497 0.0000
B 2.2497 2.2497 0.0000
B' 2.2497 2.2497 0.0000
5 A 2.2497 2.2497 0.0000
A' 2.2497 2.2497 0.0000
B 2.2497 2.2497 0.0000
B' 2.2495 2.2495 0.0000
6 A 2.2495 2.2495 0.0000
A' 2.2499 2.2499 0.0000
B 2.2495 2.2495 0.0000
B' 2.2499 2.2499 0.0000
28
Figure 16. Crankshaft Pin and Measurement Locations
Cylinder barrel measurements for cylinders 1 and 2, 3 and 4, and 5 and 6 are shown in tables 19
through 21, respectively. Figure 17 shows the typical posttest cylinder and the measurement
locations. All the cylinder measurements showed minimal wear and were less than the
maximum-allowable cylinder diameter of 5.1305 in. Cylinder crosshatching was still visible.
The maximum wear measurement in cylinders 1 and 2 was 1.5 thousandths of an inch. For
cylinders 3 and 4, all the measurements were less than 1 thousandth of an inch. Cylinder 5 had a
maximum wear measurement of 1.5 thousandths of an inch. Cylinder 6 had the most wear with
one measurement reading 3.3 thousandths of an inch.
29
Table 19. Cylinder Barrels 1 and 2 Measurements (Continued)
30
Table 20. Cylinder Barrels 3 and 4 Measurements
31
Table 21. Cylinder Barrels 5 and 6 Measurements
Piston measurements for cylinders 1 and 2, 3 and 4, and 5 and 6 are shown in tables 22 through
24, respectively. Figures 18 through 20 show the posttest pistons and the measurement
locations. All the clearance measurements for the piston skirt to the cylinder barrel were within
maximum allowable new clearance limit specifications of 18 thousandths of an inch, except for
one direction in cylinder 1 and two directions in cylinder 2. Those clearances were
approximately 1 thousandth of an inch greater than the new clearance limit. Again, these tables
indicate normal wear, and the photographs show typical light scuffing on the piston crowns. The
piston boss measurements showed normal wear.
32
Table 22. Piston Land, Skirt, and Boss Measurements for Cylinders 1 and 2
33
Figure 18. Pistons 1 and 2 and Measurement Locations
Table 23. Piston Land, Skirt, and Boss Measurements for Cylinders 3 and 4
34
Figure 19. Pistons 3 and 4 and Measurement Locations
Table 24. Piston Land, Skirt, and Boss Measurements for Cylinders 5 and 6
35
Figure 20. Pistons 5 and 6 and Measurement Locations
Table 25 shows the connecting rod bushing measurements and the measurement locations. As
shown in the table, the wear was much less than the permissible wear of 0.0015 in. All the after
measurements remained within new limit specifications, except for the A direction of rod 1,
which was 1 ten thousandth of an inch outside the limit. There was no visible scoring of the
bearings.
36
Table 25. Connecting Rod Bushing Measurements (Continued)
Cylinder Dimensions
No. Location Before After Wear
6 A 1.1253 1.1254 0.0001
A' 1.1253 1.1250 -0.0003
B 1.1254 1.1254 0.0000
B' 1.1253 1.1250 -0.0003
Table 26 shows the piston pins and the measurement locations. As shown in the table, there was
negligible wear in the piston pins. All the piston pin-to-piston hole clearances were within
maximum service limits, except for one of the eight directions in cylinder 3 and one of the eight
directions in cylinder 4. All the piston pin-to-connecting rod bushing clearance measurements
remained within the new limit specifications.
37
Table 26. Piston Pin Measurements (Continued)
The piston ring gap and tension measurements are shown in table 27, and the measurement
locations are shown in figure 21. The gage gaps remained within the new limit specifications for
both the oil control and compression rings. All the ring tensions remained within the new limit
specifications.
38
Table 27. Piston Ring Measurements (Continued)
Ring No.
Piston
Piston Rings
Gap Tension
Compression Rings
Oil Control Rings
Table 28 shows the piston ring side clearances, and figure 22 shows the measurement locations.
All the top ring clearances remained the same throughout the test. All the second compression
ring clearances remained within the new specification limits, except for cylinder 2, which was
within the allowable service limit of 0.006 inch. All the oil control ring side clearances remained
within the new limit specifications.
39
Table 28. Piston Ring Side Clearance Measurements
Table 29 shows the connecting rod bearing thickness measurements and the measurement
locations. Figure 23 shows the connecting rod bearings. The measurements show that none of
the bearing thickness changes were greater than 1 thousandth of an inch and the photographs
show normal wear modes.
40
Table 29. Connecting Rod Bearing Thickness
Measurement
Rod Side Crankshaft Side Locations
Cylinder
No. Location Before After Change Location Before After Change
1 A 0.0852 0.0854 0.0002 A 0.0850 0.0852 0.0002
B 0.0850 0.0853 0.0003 B 0.0853 0.0852 -0.0001
C 0.0850 0.0849 -0.0001 C 0.0851 0.0854 0.0003
D 0.0852 0.0851 -0.0001 D 0.0854 0.0852 -0.0002
E 0.0847 0.0855 0.0008 E 0.0848 0.0850 -0.0002
F 0.0850 0.0855 0.0005 F 0.0848 0.0858 0.0010
2 A 0.0853 0.0849 -0.0004 A 0.0854 0.0857 0.0003
B 0.0849 0.0852 0.0003 B 0.0851 0.0851 0.0000
C 0.0855 0.0849 -0.0006 C 0.0849 0.0858 0.0009
D 0.0853 0.0853 0.0000 D 0.0852 0.0854 0.0002
E 0.0854 0.0855 0.0001 E 0.0847 0.0855 0.0008
F 0.0851 0.0852 0.0001 F 0.0854 0.0852 -0.0002
3 A 0.0851 0.0854 0.0003 A 0.0854 0.0850 -0.0004
B 0.0846 0.0854 0.0008 B 0.0854 0.0853 -0.0001
C 0.0851 0.0855 0.0004 C 0.0853 0.0858 0.0005
D 0.0851 0.0853 0.0002 D 0.0852 0.0854 0.0002
E 0.0853 0.0852 -0.0001 E 0.0853 0.0851 -0.0002
F 0.0855 0.0853 -0.0002 F 0.0851 0.0854 0.0003
4 A 0.0848 0.0847 -0.0001 A 0.0852 0.0851 -0.0001
B 0.0848 0.0851 0.0003 B 0.0844 0.0850 0.0006
C 0.0857 0.0848 -0.0009 C 0.0852 0.0852 0.0000
D 0.0849 0.0848 -0.0001 D 0.0853 0.0854 0.0001
E 0.0848 0.0851 0.0003 E 0.0849 0.0854 0.0005
F 0.0857 0.0851 -0.0006 F 0.0850 0.0855 0.0005
5 A 0.0854 0.0853 -0.0001 A 0.0859 0.0859 0.0000
B 0.0853 0.0854 0.0001 B 0.0849 0.0856 0.0007
C 0.0857 0.0852 -0.0005 C 0.0856 0.0852 -0.0004
D 0.0853 0.0852 -0.0001 D 0.0852 0.0852 0.0000
E 0.0850 0.0855 0.0005 E 0.0850 0.0850 0.0000
F 0.0855 0.0853 -0.0002 F 0.0850 0.0855 0.0005
6 A 0.0854 0.0851 -0.0003 A 0.0850 0.0854 0.0004
B 0.0854 0.0852 0.0002 B 0.0852 0.0859 0.0007
C 0.0853 0.0852 -0.0001 C 0.0854 0.0858 0.0004
D 0.0855 0.0854 -0.0001 D 0.0857 0.0858 0.0001
E 0.0850 0.0851 0.0001 E 0.0849 0.0852 0.0003
F 0.0848 0.0851 0.0003 F 0.0860 0.0850 -0.0010
41
Figure 23. Connecting Rod Bearings
The oil pump gear measurements are shown in table 30, and figure 24 shows the measurement
locations. There was visible scoring of wear grooves in the pump housing, and the
measurements show there was negligible wear in the engine drive and driven oil pump gears.
Dimensions
Location Before After Wear
LW 18109 A 1.8604 1.8602 0.0002
A' 1.8603 1.8602 -0.0001
B 1.8604 1.8602 0.0002
B' 1.8609 1.8603 -0.0006
C 0.7488 0.7488 0.0000
D 0.7488 0.7485 -0.0003
LW 18110 A 1.8614 1.8614 0.0000
A' 1.8616 1.8616 0.0000
B 1.8614 1.8615 0.0001
B' 1.8610 1.8615 0.0005
C 0.7487 0.7487 0.0000
D 0.7486 0.7486 0.0000
42
Figure 24. Oil Pump Impellers, Housing, and Measurement Locations
Figure 25 shows the valve rocker arm and pin for the exhaust valve in cylinder 2. All shafts,
valve contact surfaces, and bushings showed normal wear modes for all cylinders. There was no
noticeable wear grooves or scoring.
Figure 25. Rocker Arm and Pin for the Exhaust Valve in Cylinder 2
Table 31 shows the results of the static cylinder compression tests. All compression
measurements were performed with a warm engine. Note, all compression values showed level
trends and remained above 70 psig throughout the test. These results also suggest minimal
cylinder barrel, piston ring, piston, and valve wear was occurring.
43
Table 31. Cylinder Compression Results
The evidence for minimal combustion chamber wear was also found in the power baseline
results, shown in figures 26 through 32 for the Lycoming IO540-K engine before and after the
endurance test. Figure 26 contains corrected brake horsepower (HP) data, as the test was run at
60ºF inlet air temperature and 400ºF maximum cylinder head temperature. The data in figure 26
is for a best-power mixture setting. Full mixture lean-out curves for each respective engine
speed tested are shown in figures 27 through 32. In these figures, each curve is for a constant
manifold pressure.
The engine power loss was minimal and approximated 3-4 HP, or approximately 1% of takeoff
power over the duration of the test. Moreover, the posttest maximum power of 295 HP was
within 1.7% of the 300-rated HP.
Figure 26. Power Baseline Data Before and After the Endurance Test
44
Figure 27. Mixture Lean Curves for 2700 rpm
45
Figure 29. Mixture Lean Curves for 2500 rpm
46
Figure 31. Mixture Lean Curves for 2300 rpm
47
3.3 ENGINE DEPOSITS.
As discussed in section 3.1 and shown in figure 4, the main component of the Swift binary blend
is a heavy aromatic with a high boiling point. One of the major concerns with use of heavy
components is the potential for heavy engine deposits. The teardown showed that excessive
combustion chamber deposits were not found. This is largely attributable to the use of an
aromatic that has strong solvent properties. Figures 33 and 34 show the intake and exhaust
valves for cylinders 1 through 3 and 4 through 6, respectively. The valves show light deposit
buildup and minimal valve face wear grooves. Figure 35 shows the combustion chambers for all
six cylinders. Combustion chamber deposits were light, and the cylinder cross hatching was still
visible. Figure 36 shows the piston face. Again, there was light deposit buildup. Figures 34
through 36 show a dark discoloration, darker than typically found with the use of traditional
leaded aviation gasoline.
Figure 33. Intake (left) and Exhaust (right) Valve Pictures Taken at the end of Endurance Test
(cylinders 1 through 3)
48
Figure 34. Intake (left) and Exhaust (right) Valve Pictures Taken at the end of Endurance Test
(cylinders 4 through 6)
49
Figure 35. Light Combustion Chamber Deposits for all six Cylinders
(shown from top left to bottom right)
50
Figure 36. Light Deposit Formation on Piston Faces for all six Cylinders
(shown from top left to bottom right)
51
The rocker arm covers in figure 37 show very light deposits and some dark discoloration on the
valve area.
Figure 37. Light Deposit Formation on Rocker Arms and Rocker Arm Covers
(shown from top left to bottom right)
Figure 38 shows the oil pump housing, crankcase housing bottom, oil sump and induction
housing, and accessory housing. These show an almost total lack of varnish, sludge, or deposit
buildup in the oil-wetted parts of the engine. Note the absence of wear grooves or scars in the oil
pump housing. The dark discoloration observed in the higher-temperature areas previously
discussed is not shown in figure 38.
As shown in previous figures, the piston sides and ring lands also averaged light deposits. The
oil pickup tube, pushrod tubes, and oil rings were unclogged and all piston rings moved freely.
Table 32 and figure 39 show the nozzle flow test results taken at 50 engine-hour intervals. The
results showed that the use of the Swift binary blend heavy component did not result in nozzle
deposits or nozzle flow imbalances.
52
Figure 38. Oil Pump Housing, Crankcase Housing Bottom, Oil Sump and Induction
Housing, and Accessory Housing Showing Light Deposit Formation
(referenced from top left to bottom right)
53
85.0
84.5
84.0
Percentage of Total Flow
83.5
83.0
82.5
82.0 Cylinder 1 Cylinder 2
Cylinder 3 Cylinder 4
81.5 Cylinder 5 Cylinder 6
81.0 Linear (Cylinder 1) Linear (Cylinder 2)
Linear (Cylinder 3) Linear (Cylinder 4)
80.5 Linear (Cylinder 5) Linear (Cylinder 6)
80.0
0 50 100 150 200
Engine Hours
Figure 40 shows the bottom and top spark plug photographs taken at the end of the test. The
plugs were rotated but not cleaned during the test. The photographs show a dark discoloration
and light deposit formation.
54
Table 33. Sludge, Varnish, and Carbon Deposits Ratings
The use of a high-aromatic fuel could also result in material compatibility issues, particularly
regarding elastomers, o-rings, and seals. The throttle and mixture tension were measured at the
start of each test, and the results are shown in table 34 and figure 41. The lever arm tension
increased at the start of the test and then leveled off at the 50 to 100 engine-hour mark.
55
Table 34. Throttle and Mixture Stiffness
56
5
4.5
Lever Arm Tension (in-lb) 4
3.5
3
2.5 Throttle
Mixture
2
1.5
1
0.5
0
0 50 100 150 200
Engine Hours
Figure 42 shows the diaphragms in the fuel metering unit, fuel pump, and fuel distribution valve.
The fuel metering unit and fuel distribution valve diaphragms did not show excess pliability,
stiffness, or cracking. The fuel pump diaphragm showed creases. The fuel pump outlet pressure
remained within Lycoming’s acceptable range during the tests.
Figure 42. Fuel Metering Unit Diaphragm Showing No Marks or Creases (left), Fuel Pump
Secondary Diaphragm Showing Marks and Creases (center), and Fuel Distribution Valve
Diaphragm Showing No Marks or Creases (right)
Swift Enterprises supplied 4500 gallons of a binary blend to the FAA AFETF for a 150-hour
engine endurance test. This was a joint effort by the FAA AFETF, Textron Lycoming, and Swift
Enterprises. The Swift binary blend’s two main components were made in a refinery, not in a
bio-process. The 150-hour test used a Lycoming IO540-K engine and followed the 14 CFR
33.47 endurance block test. The test was severe, as most of the test time was spent at maximum-
rated power under maximum engine and oil temperatures. Prior to overhaul, the remanufactured
57
engine was torn down and measured by Textron Lycoming, reassembled, and sent to the AFETF
for break-in and testing. At the end of the test the engine was sent back to Lycoming for
remeasurement. The engine was broken in, and operated solely on Swift binary blend, which
consisted of 82.5% mass mesitylene and 16.6% mass isopentane, with approximately 1% of
other hydrocarbon components.
The Swift binary blend met all current aviation gasoline ASTM D 910 specifications for 100LL,
except it was unleaded and undyed and had a 3.7% reduced specific energy content, and it did
not meet the requirements for distillation slope at the 50%, 90%, and end point temperatures.
The fuel weighed an average of 1 lb/gal more than traditional aviation gasoline and, therefore,
had roughly 13% more energy per gallon of fuel. The consequence of additional fuel weight on
an airframe and the potential to affect pilot operating handbook performance curves were beyond
the scope of this research and were not addressed. Fuel storage and aging issues were also
beyond the scope of this research and were not addressed. Detonation and power performance of
another Swift binary blend were evaluated in a previous FAA report [1].
Using the Swift binary blend resulted in 1% oil dilution, normal engine wear, light engine
combustion deposits, light fuel system deposits, and very light oil system deposits. The fuel
injector nozzles did not experience visible deposit formation or noticeable fuel distribution
issues.
Starting difficulty was noted on mornings after the engine was left to sit overnight. Starting was
immediate with a warm engine. It is recommended that start-ability issues be addressed in actual
airframes rather than in a test cell environment.
The only material compatibility issue uncovered by this limited test was stretch marks or creases
on the fuel secondary pump diaphragm. No noticeable creases, softness, or brittleness were
noticed in the fuel-metering unit and distribution valve diaphragms or the throttle and mixture
o-rings. The fuel pump outlet pressure remained within Lycoming’s acceptable range during the
tests. Compatibility of fuel pump elastomers with Swift binary fuel is an area that requires
further extensive rig and soak testing.
This type of endurance test cannot adequately address material compatibility issues. These
issues should be addressed with a material compatibility study involving all the wetted parts of
the engine using typical materials found throughout engine and aircraft fuel systems. Further, it
is recommended that this endurance test be repeated using a straight-weight mineral oil without
deposit control additives.
This test does not constitute FAA certification, endorsement, or approval of any kind. Further, it
is recommended that Swift Enterprises perform further tests to address fit-for-purpose
specification properties using a Swift binary blend made from their fermentation process.
58
5. REFERENCES.
1. Atwood, D., “Full-Scale Engine Detonation and Power Performance Evaluation of Swift
Enterprises 702 Fuel,” FAA report DOT/FAA/AR-08/53, January 2009.
59/60