Skip to main content

Full text of "A Materials Compatibility and Thermal Stability Analysis of Common Hydrocarbon Fuels"

See other formats


(^CC3- loQL 



A Materials Compatibility and Thermal Stability Analysis 
of Common Hydrocarbon Fuels 

M. L. Meyer 

NASA Glenn Research Center 

Cleveland, Ohio 

B. R. Stiegemeier 

University of Toledo 

Toledo, Ohio 



ABSTRACT 

A materials compatibility and thermal stability investigation was conducted using five 
common liquid hydrocarbon fuels and two structural materials. The tests were performed at the 
NASA Glenn Research Center Heated Tube Facility under environmental conditions similar to 
those encountered in regeneratively cooled rocket engines. Scanning-electron microscopic 
analysis in conjunction with energy dispersive spectroscopy (EDS) was utilized to characterize 
the condition of the tube inner wall surface and any carbon deposition or corrosion that was 
formed during selected runs. Results show that the carbon deposition process in stainless steel 
tubes was relatively insensitive to fuel type or test condition. The deposition rates were 
comparable for all fuels and none of the stainless steel test pieces showed any signs of corrosion. 
For tests conducted with copper tubing, the sulfur content of the fuel had a significant impact on 
both the condition of the tube wall and carbon deposition rates. Carbon deposition rates for the 
lowest sulfur fuels (2 ppm) were slightly higher than those recorded in the stainless steel tubes 
with no corrosion observed on the inner wall surface. For slightly higher sulfur content (25 ppm) 
fuels, nodules that intruded into the flow area were observed to form on the inner wall surface. 
These nodules induced moderate tube pressure drop increases. The highest sulfur content fuels 
(400 ppm) produced extensive wall pitting and dendritic copper sulfide growth that was 
continuous along the entire tube wall surface. The result of this tube degradation was the inability 
to maintain flow rate due to rapidly increasing test section pressure drops. Accompanying this 
corrosion were carbon deposition rates an order of magnitude greater than those observed in 
comparable stainless steel tests. The results of this investigation indicate that trace impurities in 
fuels (i.e. sulfur) can significantly impact the carbon deposition process and produce 
unacceptable corrosion levels in copper based structural materials. 

INTRODUCTION 

NASA and the United States Air Force have had extensive interest in advanced 
propulsion systems (rocket, air-breathing, and combined cycle systems) operated with 
hydrocarbon fuels. In many cases these propulsion systems are intended to be reusable, low 
cost, and safe to operate. As these propulsion devices evolve and designers look for ways to 
improve engine performance, primarily through increasing operating pressures, the strain placed 
on fuels used to cool engine components simultaneously increases. In some cases, the heat 
load, which is almost directly proportional to increases in operating pressure, may more than 
double. With the flow rate of fuel available for cooling purposes limited, the only way to absorb 
this increase in heat load is to increase the temperature at which the fuel must operate. 
Additionally, to transfer the energy from the combustor to the coolant, it is desirable to have 
combustor walls operating at temperatures as high as possible. 



Approved for public release. Distribution is unlimited. 



Unfortunately, experience has shown that hydrocarbon fuels, at sufficiently high 
temperatures, begin to decompose resulting in the formation of gums and solids that can deposit 
on wetted fuel surfaces. The deposit then acts as an insulating layer, causing a further increase 
in wall temperature, which can eventually lead to loss of structural integrity and propulsion system 
failure. Excessive deposits may also reduce the coolant flow area and increase surface 
roughness, resulting in increased coolant pressure drop or reduced coolant flow rate. The rate at 
which the deposition process occurs is driven by many factors such as wail temperature, fuel 
composition (e.g. sulfur and oxygen content), velocity (residence time), and coolant passage 
material. At the present time, the interaction of these effects is not completely understood and 
the prediction of deposition remains difficult. However, it is generally accepted that at lower 
temperatures, less than about 900 °F, the deposition process occurs as the result of auto- 
oxidation reactions, whereas the deposition process at higher temperatures is driven by the 
pyrolysis of the hydrocarbon molecules.^ An additional complication has been reported when fuel- 
bound sulfur is present in fuels used to cool copper and copper alloys structures. The sulfur 
reacts with the copper to form copper sulfides. This sulfide corrosion can both damage the copper 
surface and disturb the flow. 

This study is part of a larger effort coordinated between NASA and the US Air Force to 
understand the feasibility of applying a variety of fuels with different characteristics to the various 
propulsion systems being investigated by the respective organizations. Thus, a series of tasks 
have been initiated to study both combustion and cooling performance of several readily available 
aerospace fuels. The purpose of the effort discussed here was to experimentally investigate, 
using resistively heated tube sections, the heat transfer, thermal stability, and compatibility (with 
stainless steel and copper) characteristics of five common hydrocarbon fuels: JP-7, JP-8, JP- 
8+100, JP-10, and RP-1 . The experiments were conducted under conditions similar to those 
encountered in current expendable, regeneratively cooled, hydrocarbon rocket engines.^' ^ The 
fuels were chosen as representing most of the fuels used today for aerospace vehicle propulsion: 
JP-7 is the Air Force's highest thermal stability fuel, used in the SR-71 ; JP-8 is the standard 
kerosene fuel for the United States Air Force and Army; JP-8+100 is JP-8 with an additive 
package that has been shown to markedly improve thermal stability; JP-10 is a higher density 
hydrocarbon fuel used in cruise missiles and a candidate for several trans-atmospheric vehicles 
(also the only pure component fuel tested, exotetrahydrodicyclopentadiene); and RP-1 is the long 
accepted standard propellant for U.S. hydrocarbon rocket engines. 

An earlier report presented the results of carbon deposition measurements from these 
experiments."* Recently, a number of remaining pieces of the test articles from those experiments 
were sectioned and examined with Scanning Electron Microscopy (SEM), and the interior 
surfaces were analyzed for elemental composition. The results of that effort provide additional 
insight into the mechanisms involved in the degradation of cooling capability and deposit 
formation and are presented in this report. 

BACKGROUND 

Scanning electron microscopy (SEM) and associated compositional analytical techniques 
have been used in previous efforts to characterize the deposits formed in both laboratory and 
fielded components. In the gas turbine field, Schirmer published an extensive set of micrographs 
of the deposits found in an assortment of rigs and engine fuel components.^ The deposits were 
primarily carbon compounds, which had condensed from the fuel in the form of spherical particles 
and collected on the fuel system surfaces. Schirmer characterized these particles as soft and 
subject to fusion on the heated surface. And noted that under certain conditions, they could be 
transformed into different forms of deposits: smoother, denser varnishes, for example. In 
references 6 - 8, which were more similar to the present experiments in operating conditions, 
dendritic structures were also observed protruding from copper or copper alloy surfaces. Analysis 
indicated that these dendrites were composed largely of copper and were also referred to as 
"copper wool." In experiments reported in reference 8, the possibility that the dendritic structures 
were caused by reaction of the fuel-bound sulfur with the copper in the test section surface was 



investigated by doping fuels (RP-1, methane, propane) with additional sulfur compounds 
(thianaphthene, benzyl disulfide), and indeed an increase in the dendritic formations was 
observed after increasing the fuel sulfur content. In addition, the dendrite formations were 
identified as copper sulfides. 

The carbon deposits formed in sections of each test piece discussed in this paper were 
previously quantified with a burn-off technique and reported in reference 4. One of the key results 
reported was a dramatic difference between the tests in SS 304 and Copper 101 test pieces. 
While the results in SS 304 tubes showed little fuel-to-fuel difference in the rate of deposit 
formation, in Copper 101 the JP-8 and JP-8+100 had much higher deposit rates than the other 
fuels (figure 1). Furthermore, the Copper 101 and JP-8 and JP-8+100 tests were terminated after 
only a few minutes due to significant increases in pressure drop required to maintain coolant flow 
rate. In reference 4 it was suggested that the increase in deposits was due to the significantly 
higher sulfur content in the JP-8 and JP-8+100 fuels. 



"E 2000 

I. 



O 1000 



& 

c 
o 
.a 

w 

a 
O 























1 




674 °F < T,.|, < 801 °F 


2185 


- 


4,377 < Re < 43,91 5 


] 






\ 








^y^ 


1170 1210 
■ ■ 


1352 






— 


1 1 


»* 
M 




98 


100 134 


114 


184 


nil 







r 



r 



f 



r 






K^ 



f 



r 



Figure 1 : Carbon deposition rates for copper tubes (Tw = 1000 °F; V = 25 ft/s). 

EXPERIMENT 

A series of heated tube tests was conducted at the NASA Glenn Research Center 
Heated Tube Facility (HTF) in Cleveland, Ohio using the combustible liquid flow system. A 
detailed discussion of the experimental design, the facility and the test hardware was presented 
in reference 4. A brief summary of the experiment is provided here for clarity. 

DESIGN OF EXPERIMENT 



The primary variables investigated in these experiments were fuel type (five types), 
wetted wall material (two types), flow velocity (two levels), and inside wall temperature (two 
levels). Coolant bulk outlet temperature and pressure were held constant. In order to further 
reduce the number of tests and increase the statistical significance of the results, a half-factorial 
test matrix was selected with several conditions repeated. The tests conducted for each fuel are 
defined in table 1 . One disadvantage of the half-factorial test matrix selected is that a direct test- 
to-test comparison of the variation of a single parameter, other than fuel type, is not possible. The 
two conditions shaded in table 1 were the focus of the present work, as they were the most 
severe conditions tested for each tube wall material. 



Table 1 : Test Matrix Summary for Each Fuel 



Material 


Inner Wall Temp. (°F) 


Average Velocity (FT/SEC) 


SS304 
Zy SS 304t ^: 
Copper 101 
Copper 101 1 


750 


25 


750 
1000 


75 mm 

75 

25 •'* 


Bulk Outlet Temperature 500°F Average Pressure 1 000 psi tRepeat Condition 



The wall temperatures and coolant flow velocities in table 1 were selected as 
representative of current expendable RP-1 fueled engines.^' ^ Coolant outlet temperatures for all 
tests were set to 500 °F, while mean test section pressures were fixed at 1000 psi. The outlet 
temperature was chosen to be sufficiently below (~ 250 °F) the critical temperature of the fuels to 
avoid the dramatic thermophysical property changes of the bulk flow that would result from 
operating near reduced temperatures (T/Tc) of one. Run times for all tests were nominally set to 
twenty minutes. This was predicted to be long enough to produce sufficient coke for analysis, but 
would still keep fuel quantity requirements low. In some cases test were stopped short due to 
increasing coolant pressure drop. 



FACILITY AND HARDWARE 

A simplified schematic of the facility architecture is shown in figure 2.^ Ambient 
temperature fuel is loaded into the run tank and pressurized up to 1500 psig. Flow rate and test 
section pressure are regulated by two valves. Two independent closed-loop controllers operate 
these valves based on flow rate measurement and test section backpressure measurement to 
maintain the desired test section flow and pressure conditions. 



Vent 



-ih 



-m- 



— C^ 
HSI-t3ch 




nil 

valve 




Heat 
exchanger 

■* — Water supply 



f 



c-tJ5j. 



jg, Regulated 

PriSunStion GN2 supply 





=^ 


|PT1. 


TC| 




r^ 


] 


^^^ 


te 


i 


i 


3 


IptI- 


^Tc| 



-Thermal 
expansion 
stress relief 



Electrical 
' heating power 
connection 



Flow meter 



valve 



Filter 



I Inlet 
plenum 



Row control valve 



Figure 2. Simplified schematic of the NASA Glenn Heated Tube Facility. 



The test sections for these tests were fabricated from drawn tubing of oxygen free 
electrical copper (Copper 101) and 304 stainless steel (SS 304). The test section dinnensions are 
shown in tables 3 and 4, and they were typically instrumented with four or five outer wall 
thermocouples, which were spot welded to the SS 304 test sections and brazed to the copper test 
sections. The properties of JP-10 were sufficiently different from the other fuels to require a 
change in the test section dimensions to meet the test matrix conditions. 

Table 3: Dimensions and Relevant Operating Conditions for JP-7, JP-8, JP-8+100, and RP-1 

Test Pieces 



Material 


Velocity 

(ft/sec) 


Wall Temp. 

cn 


Inside 
Dia. (in) 


Heated 
Length (in) 


Heat Flux 

(BTU/in'- 
sec) 


Flow Rate 

(Ibjsec) 


SS304 

SS304 

Copper 101 

Copper 101 


25 

, 75 

25 

75 


750 
1000 
1000 
750 


0.06175 

0.06175 

0.061 

0.061 


14.40 


2.20 


0.024 
0^072 1 
0.024 ?l 
0.070 


11.70 
8.10 

16.75 


8.10 
3.90 
5.60 



Table 4: Dimensions and Relevant Operating Conditions for JP-10 Test Pieces 



Material 


Velocity 

(ft/sec) 


Wall Temp. 

(°F) 


Inside 
Dia. (in) 


Heated 
Length (in) 


Heat Flux 

(BTU/ in^-sec) 


Flow Rate 

(Ibjsec) 


SS304 


25 


750 


0.06175 


13.50 
10.00 
7.60 
15.50 


2.00 0.027 

8.10 0.081 ; 

-3.60 :^i:i, 0.027 
5.20 0.080 


SS 304 
Copper 101 
Copper 101 


75 
25 

75 


1000 
"1000^^ 
750 


0.06175 
X- 0.061 
0.061 


FUEL ANALYS 


IS 













Each of the fuels tested in this study was delivered with a vendor test report verifying that 
they met the requirements of the respective military specification. Since sulfur has been identified 
as an important contaminant in previous work with copper, it was necessary to have accurate 
characterization of the sulfur contamination. In general, the standard sulfur analyses required by 
the specifications were not sufficiently accurate for the low sulfur fuels. Thus, additional analyses 
for total fuel-bound sulfur present (e.g. disulfides, mecaptans, thiophenes) were conducted 
according to standard ASTM test methods. The lower sulfur content fuels were analyzed using 
the ASTIVI D 5453, while the ASTM D 4294 was performed on JP-8 and JP-8+100. The 



Table 5: Fuel Total Sulfur Content 



Fuel 


Total Sulfur Content (ppm) 


JP-7 


2^ 


JP-8andJP-8-(-100 


400 


RP-1 


23^ 


JP-10 


Not Analyzed 


' ASTM D 5453 ASTM D 4294 



reproducibility of the test methods is dependant upon the amount of sulfur in the sample, but for 
the ranges seen in these tests the reproducibility is approximately 2 ppm for the ASTM D 5453 
and 1 00 ppm for the ASTM D 4294. The results of the total sulfur content tests are shown below 



in table 5. It has been reported previously that certain fuel-bound sulfur compounds are more 
reactive than others, however, results of speciation of the sulfur compounds in the test fuels is not 
available at this time. 

MICROSCOPIC ANALYTICAL TECHNIQUES 

Segments of each test section were milled to expose the interior surface for examination 
under a scanning electron microscope (SEM) as shown in the sketch in figure 3. In conjunction 
with the SEM imaging, energy dispersive spectroscopy (EDS) was utilized to perform elemental 
composition analysis on the inner wall of the tube samples. The EDS technique works by 
measuring the number and energy of X-rays emitted by the sample after being irradiated by an 
electron beam. By matching the energy of the emitted X-rays to the known characteristics of 
each element, the composition of the sample can be determined. It is important to recognize that 
the highly irregular surfaces analyzed in this investigation are not optimal for EDS analysis and 
that corrections for the atomic number, absorption, and possible fluorescence of the elements are 
all possible sources of errors in the quantification process. Even with these limitations, the 
qualitative information provided by the EDS analysis provides significant insight into the 
deposition and corrosion processes. All of the results given below were obtained using a 
standardless calibration at an accelerating voltage of 10kV. 




10 um 



Figure 3. SEM micrograph of a clean section of Copper 101 tubing showing the as 
delivered internal surface due to the drawing process. 

RESULTS AND DISCUSSION 



A series of micrographs are presented in figures 4-8 and illustrate the most significant 
results of the microscopic imaging effort. In figure 4, four images magnified lOOOx are presented 
from approximately the same location in four different test sections, which were operated with 
each of the fuels at approximately the same test conditions. The one exception was that the JP-8 
test was only 2 minutes in duration compared with 20 minutes for the other fuels (note that JP-8 
and J,P-8+100 were similar and so separate results are not presented). The four fuels show a 
significant range in corrosion/deposition behavior. It is also apparent that the sulfur content of the 
fuel is a significant contributor to the corrosion/deposition process. 

JP-8, with the highest sulfur content (400 ppm), in just two minutes formed dendritic 
structures that dominate the corrosion/deposition process and nearly fill the test section. The flow 
passage was not completely plugged in this test, but coolant pressure drop more than doubled, 
forcing early termination of the test. At approximately 23 ppm sulfur content, RP-1 formed some 
nodules (lighter colored structures) and darker deposits on the surfaces between the nodules. 
JP-7, a very low sulfur fuel at 2 ppm total sulfur, did not form any nodules in the 20 minute test. 



3 




w 


0) 






(11 


3 


3 


c 


00 


E 


a. 


o 


i 




"S" 


E 




Q. 




Q. 





O 


-J 


T ■ 


•^- 


*-• 


<ii 


c 


x: 




o 


o 
o 


c 


x_ 


o 


3 


r 


*♦— 


o 

Q. 


3 
(0 


■D 
CD 


03 

3 


m 


*♦" 


03 


t^ 


^ 


n 


(1) 


-3 




"o" 


o 








c 


c 
o 






0) 


03 




k_ 


(1 


3 c 


c 


■n o 




(D E n' 
Q. Q.~3 

^■^ if 

03 c .2 
S: 03 -^j 

=> ^ 2 







03 

3 



00 

Q. 



^ o 
■;= o 

03 

w 



03 



O B 

r- CO 

o o 
"^E 



to i: 
03 0) 



=! 

CO 
o 

!■§ 

.2 03 



(0 



o 



<0 CO 



o .- 
w -c 

Q..E 

2^ 

o ^ 
.9 "3 

8.2 

CO CO 

■5^ 

"to « 

C 

11 

O (M 

- 
03 D- 

r o 

02 



Sb 

LL W 



The deposits formed are thin enough that the as delivered roughness of the tube wall is still 
visible. JP-10, a synthetic fuel, should have extremely low levels of sulfur, and formed only a thin 
deposit on the tube wall as well. 

The deposition process is also typically dependent on the residence time that the fuel is 
exposed to the heated environment. Thus, it is of interest to examine any changes in the structure 
of the corrosion/deposition products from the inlet to the exit of the test section. Figures 5 and 6 
are micrographs at 30x and 1000x from various locations along the test section from the test with 
JP-8. Although, it is clear that the amount of deposits increased progressing downstream in the 
test section, even near the entrance of the heated section the amount of material formed is 
significant. From the closer examination presented in figure 6, the structure of the material is 
dendritic all along the test section, though it appears to be more densely packed toward the exit. 

A final notable observation can be made from the images taken from just downstream of 
the heated portion of the test section. Only a thin deposit is present and none of the dendritic 
structure is observed. Since even in this area the bulk temperature of the fuel is 500 °F, the 
formation of the dendrites appears to be driven by the higher wall temperatures in the heated test 
section. A similar series of micrographs (lOOOx only) for an RP-1 test section is presented in 
figure 7. In this case the dominant material formed on the walls is nodular in shape, with a thin 
layer of darker deposit between the nodules. Although there is a moderate change in the amount 
and appearance of the nodules progressing downstream, it would be difficult to draw conclusions 
based on these differences. As with the JP-8 test section, the micrograph from the portion of the 
tube just downstream of heating shows only a thin layer of darker deposit. 

A representative micrograph from a SS 304 test section tested with JP-8 is included in 
figure 8 to illustrate the test section material effects. Only a thin darker deposit was present on the 
SS 304 test section without dendritic or nodular structures. The loose particles present on the 
surface are believed to be contamination due to an earlier sectioning technique, which used an 
abrasive cut-off wheel. 

Qualitative observations based on the elemental compositional analyses from the EDS 
technique are presented in figures 9-1 1 . One of the limitations of the EDS technique is that the 
focal point of the probe is larger than many of the structures of interest in the micrographs. 
Analyses that focused on the dendrites and nodules of figures 4(a) and 4(b) indicated that the 
nodules and dendrites had higher concentrations of sulfur and copper, where as the darker 
deposits between had higher carbon concentrations. This supports the conclusion of reference 4 
that the dendritic structures contain large amounts of copper sulfides. The results of broader focal 
point EDS analyses were more consistent and are presented graphically. Figure 9 shows the 
results of the EDS analysis for each of the JP-8 test section images in figure 6. The composition 
appears to be relatively constant progressing down the test section. Copper and sulfur are the 
main components with carbon and oxygen present in lesser amounts. Figure 10 presents similar 




285 yum 

Figure 8. Interior surface of SS 304 
tube (magnification 35x) tested with 
JP-8 for 20 minutes; 75 ft/s, 1000 °F. 




1.5 4 6 Outlet 

Axial Distance from Start of Heating (in) 

Figure 9. Tube inner wall surface 
elemental composition with JP-8 as a 
function of location in the test section. 





RP-1 Fuel 
Copper 101 
Tw = 1000 °F 
V = 25 ft/sec 



6 Outlet 

Axial Distance From Start of Heating (in) 




JP-8 



RP-1 



JP-7 



JP-10 



Figure 10. Tube inner wall surface 
elemental composition after test with RP-1 
as a function of location in the test section; 
20 minute duration. 



Figure 11. Tube inner wall surface 
elemental composition after test after 20 
minutes test duration (2 minute duration 
for JP-8). 



results for the RP-1 images in figure 7. In this case, it appears that the relative amount of carbon 
present is greater, and the amount of sulfur on the surface is less than the JP-8 test. This result is 
additional support that the dendrites that dominate the JP-8 corrosion/deposition material are due 
to the higher sulfur concentration. This point is made more clearly in figure 1 1 , which presents 
results for each of the four fuels (images in figure 4) side-by-side for comparison. Even 
qualitatively, the results suggest that the sulfur content of the fuel is key to the 
corrosion/deposition process with copper tube materials at these conditions. 

Quantitative data on the amount of carbon present on the tube surface was presented in 
figure 1 , and indicated that significantly more carbon was deposited in the JP-8 tests than in tests 
with the other fuels. Based on the micrographs in figure 6, it is likely that the dendritic structure 
due to sulfur corrosion of the copper surface contributes to increased carbon deposition in two 
ways. First the dendrites serve to greatly increase the surface area where deposits can collect. 
Second, the dendrites form cavities where fuel can be trapped near the hot wall under stagnant or 
low velocity conditions and form additional solids. 

SUMMARY AND CONCLUDING REMARKS 



A series of electrically heated tube experiments were conducted with five readily 
available hydrocarbon military fuels to compare their relative thermal stability and compatibility 
with Stainless Steel 304 and Copper 101 materials. A previous report presented that significant 
carbon deposits were observed with JP-8 and JP-8+100 in Copper 101 test sections and were 
believed to be the result of higher sulfur content in those fuels. In this report, qualitative 
observations based on SEM images and EDS surface elemental composition analysis of portions 
of those same test sections were presented to lend insight into the mechanism of corrosion and 
deposition in the experiments. The images and composition analysis, although qualitative, 
support the proposal that the fuel-bound sulfur reacts with the copper wail material to form copper 
sulfides. Depending on the amount of fuel-bound sulfur present, these copper sulfide structures 
can be dendritic and extensive, forming a significant flow blockage. 

It has been proposed that for the higher sulfur fuels, JP-8 and JP-8+100, the dendrites 
formed result in additional carbon deposition both through the additional surface area for 
collection of deposits and by creating cavities where fuel is held near the hot walls at near 
stagnant conditions where additional deposits form. 



Finally, it is recommended that more quantitative and spatially precise (than EDS) 
analytical techniques be utilized to conclusively support this proposed mechanism of corrosion 
and deposit formation. 

ACKNOWLEDGMENTS 

The authors wish to acknowledge the support of Dr. Tim Edwards, Air Force Research 
Laboratory at Wright Patterson AFB, in providing fuels for the tests, carbon deposition 
measurements, and fuel sulfur analysis. 

The authors are also indebted to Mary Ann Dembowski and Drago Androjna of the NASA 
Glenn Research Center for their efforts to obtain the high-quality SEM images and EDS analysis 
results presented in this report. 

REFERENCES 



1. Katta, V.R., Jones, E.G., and Roquemore, W.M., " Modeling of Deposition Process in Liquid 
Fuels," Combust. Sci. and Tech, Vol. 139, pp.75-111, 1998. 

2. Volkmann, J.C, " Development of Diazidoalkane Fuel Additives for LOX/RP-1 Booster 
Engines," Al AA 92-31 30, 1 992. 

3. Linne, D.L., and Munsch, W.M., "Comparison of Coking and Heat Transfer Characteristics of 
Three Hydrocarbon Fuels in Heated Tubes," 32"*^ JANNAF Combustion Meeting, CPIA 
Publication 631, Vol. II, pp. 95-101, October 1995. 

4. Stiegemeier, B., Meyer, M. L., and Taghavi, R., 'Thermal Stability and Heat Transfer 
Characteristics of Five Hydrocarbon Fuels: JP-7, JP-8, JP-8+100, JP-10, and RP-1," AIAA-2002- 
3873, 2002. 

5. Schirmer, R.M., "Morphology of Deposits in Aircraft and Engine Fuel Systems," SAE Paper 
700258, Society of Automotive Engineers, 1970. 

6. Giovanetti, A.J., Spadaccini, L.J., and Szetela, E.J., "Deposit Formation and Heat Transfer in 
Hydrocarbon Rocket Fuels," NASA-CR-1 68277, October 1983. 

7. Roback, R., Spadaccini, L.J., and Szetela, E.J., "Deposit Formation in Hydrocarbon Rocket 
Fuels," NASA-CR-1 65405, August 1981. 

8. Homer, D.G., and Rosenberg, S.D., "Hydrocarbon Fuel/Combustion-Chamber-Liner Materials 
Compatibility," NASA CR-187104, April 1991.