CHEM
`
`ousTRIAL&
`
`ISPLA
`fEB ? ~ 1994
`
`A PUBLICATION OF THE AMERICAN CHEMICAL SOCIETY
`
`

`
`INDUSTRIAL &
`ENGINEERING
`CHEMISTRY
`RESEARCH
`
`Editor:
`DONALD R. PAUL,
`University of Texas at Austin
`
`Senior Editor:
`Jerome A. Seiner,
`PPG Industries
`
`Associate Editors:
`John L. Anderson,
`Carnegie-Mellon University
`Milorad P. Dudukovic,
`Washington University
`J. D. Seader,
`University of Utah
`
`Manuscript Office:
`Yvonne D. Curry,
`Manager
`Marie C. Wiggins,
`Senior Associate Editor
`Bryan D. Tweedy,
`Associate Editor
`Barbara V. James,
`Assistant Editor
`
`Advisory Board
`Michael J. Antal, Jr.
`Gary E. Blau
`Nai-Yuen Chen
`James R. Fair
`Gilbert F. Froment
`George R. Gavalas
`Louis Hegedus
`Frederick J. Karol
`Charles L. Liotta
`Richard D. Noble
`Joji Oka
`William R. Parrish
`W. Harmon Ray
`S. I. Sandler
`Peter R. Sperry
`Barry L. Tarmy
`D.T.Wu
`Ralph T. Yang
`
`Industrial & Engineering Chemistry
`Research (ISSN: 0888-5885) is published
`monthly by the American Chemical
`Society at 1155 16th St., N.W., Washing(cid:173)
`ton, DC 20036. Second-class postage
`paid at Washington, DC, and additional
`mailing offices. POSTMASTER: Send
`address changes to Industrial & Engi(cid:173)
`neering Chemistry Research, Member &
`Subscriber Services, P .0. Box 3337,
`Columbus, OH 43210.
`Canadian GST Reg. No. R127571347.
`Printed in USA.
`
`PUBLICATIONS DIVISION
`Robert H. Marks, Director
`Journals Department: Charles R. Bertsch,
`Head
`Editorial Office: Mary E. Scanlan, Man(cid:173)
`ager; Kathleen E. Duffy, Anne C. O'Melia,
`Joseph E. Yurvati, Journals Editing Man(cid:173)
`agers; Jean M. Redmond, Associate Editor
`
`Advertising Office: Centcom, Ltd., 1599
`Post Road East, P.O. Box 231, Westport,
`CT 06881
`
`© Copyright 1993 by the American
`Chemical Society. Copyright permission:
`An individual may make a single repro(cid:173)
`graphic copy of an article in this publica(cid:173)
`tion for personal use. Reprographic copy(cid:173)
`ing beyond that permitted by Section 107
`or 108 of the U.S. Copyright Law is
`allowed, provided that the appropriate per(cid:173)
`copy fee is paid through the Copyright
`Clearance Center, Inc., 27 Congress St.,
`Salem, MA 01970. For reprint permission,
`please write to the Copyright Administra(cid:173)
`tor, Publications Division, at the ACS
`Washington address. Bulk reprints: for
`quotes and information on placing an
`order, please call (202) 872-4539 or write to
`the Distribution Office at the ACS Wash(cid:173)
`ington address.
`The paper used in this publication
`meets the minimum requirements of Amer(cid:173)
`ican National Standard for Information
`Sciences-Permanence of Paper for
`Printed Library Materials, ANSI Z39.48-
`1984.
`Editorial Information
`GUIDE FOR AUTHORS, published
`in the first issue of the year, gives copy
`requirements for manuscripts. A copy(cid:173)
`right status form is also published in the
`first issue.
`Manuscripts for publication should
`be submitted in quadruplicate to Manager,
`Manuscript Office, at the ACS Washington
`address [(202) 872-4542; FAX (202) 872-
`6325]. Include a signed copyright status
`form.
`Correspondence regarding accepted
`papers and proofs should be directed to
`the Journals Department at the address
`indicated.
`The American Chemical Society and its
`Editors assume no responsibility for the
`statements and opinions advanced by con(cid:173)
`tributors. Views expressed in the editori(cid:173)
`als are those of the authors and do not nec(cid:173)
`essarily represent the official position of
`the Society.
`Registered names and trademarks, etc.,
`used in this publication, even without spe(cid:173)
`cific indication thereof, are not to be con(cid:173)
`sidered unprotected by law.
`Subscription and Business Information
`1993 subscription prices, including
`postage. (For membership information,
`contact Office of Member Services at the
`ACS Washington address.)
`
`All
`Canada
`Other
`and
`U.S. Mexico Europe* Countries*
`
`Printed
`Mem bers- 1 yr $ 64
`2 yr $115
`Nonmembers
`$567
`Microfiche
`Members-
`
`I yr $ 64
`2 yr $115
`Nonmembers
`$567
`Supplementary
`$ 15
`material (microfiche)
`* Air service included.
`
`$ 84
`$155
`$587
`
`$ 64
`$115
`$567
`$ 21
`
`$108
`$203
`$611
`
`$ 74
`$135
`$577
`$ 21
`
`$120
`$227
`$623
`
`$ 74
`$135
`$577
`$ 21
`
`Rates quoted do not apply to nonmem(cid:173)
`ber subscribers in Japan, who must enter
`subscription orders with Maruzen Com(cid:173)
`pany Ltd., 3-10 Nihonbashi 2-chome,
`Chuo-ku, Tokyo 103, Japan. Tel:
`(03)
`272-7211.
`New and renewal subscriptions should
`be sent with payment to American Chemi(cid:173)
`cal Society, Department L-0011, Colum(cid:173)
`bus, OH 43268-0011. Subscriptions are
`available only on a calendar year basis.
`Printed edition single issue price:
`current year, $48.00; 1992, $52.00; prior
`year, $46.00. Foreign postage additional.
`Mail orders and requests for back volume
`prices should be sent to Microforms &
`Back Issues Office at the ACS Washington
`address.
`Phone orders can be placed for printed,
`microfiche, and microfilm editions by call(cid:173)
`ing the ACS toll free at (800) 333-9511
`from anywhere in the United States and
`Canada; elsewhere call (614) 447-3776.
`Phone orders can be charged to Visa, Mas(cid:173)
`terCard, or American Express accounts.
`Changes of address must include both
`old and new addresses with ZIP code and a
`recent mailing label. Send all address
`changes to Member & Subscriber Services.
`Please allow 6 weeks for change to become
`effective. Claims for missing number•
`will not be allowed if loss was due to failure
`of notice of change of address to be
`received in the time specified; if claim is
`dated (a) North American: more than 90
`days beyond issue date, (b) all other for(cid:173)
`eign: more than 180 days beyond issue
`date. Claims are handled by Member &
`Subscriber Services.
`Supplementary material is noted in the
`table of contents with a • . It is available as
`photocopy ($10.00 for up to 3 pages and
`$1.50 per page for additional pages, plus
`$2.00 for foreign postage) or as 24X micro(cid:173)
`fiche ($10.00, plus $1.00 for foreign post(cid:173)
`age). Canadian residents should add 73
`GST. See supplementary material notice
`at end of journal article for number of
`pages. Orders must state whether for pho(cid:173)
`tocopy or microfiche and give complete
`title of article, names of authors, journal,
`issue date, and page numbers. Prepayment
`is required and prices are subject to
`change. Order from Microforms & Back
`Issues Office at the ACS Washington
`address. Supplementary material except
`structure factors also appears in the micro(cid:173)
`film edition.
`
`American Chemical Society
`1155 16th St., N.W.
`Washington, DC 20036
`(202) 872-4600
`TDD (202) 872-8733
`FAX (202) 872-4615
`
`Journals Department
`American Chemical Society
`2540 Olentangy River Road
`P.O. Box 3330
`Columbus, OH 43210
`(614) 447-3600, Ext. 3171
`TELEX 6842086
`FAX (614) 447-3745
`
`Memb~r & Subscriber Services
`American Chemical Society
`P.O. Box 3337
`Columbus, OH 43210
`(614) 447-3776; FAX (614) 447-3671
`(800) 333-9511
`
`

`
`Ind. Eng. Chem. Res. 1993, 32, 3117-3122
`
`3117
`
`Supercritical Fuel Deposition Mechanisms
`
`Tim Edwards*
`USAF Wright Laboratory, WL/POSF Bldg 490, 1790 Loop Rd N, Wright-Patterson AFB, Ohio 45433-7103
`
`Steven Zabarnick
`300 College Park, KL463, University of Dayton Research Institute, Dayton, Ohio 45469-0140
`
`Jet fuel is the primary coolant used in high-speed aircraft. A decrease in surface deposition (fouling)
`is often seen as jet fuels and pure hydrocarbons are heated above approximately 370 °C, as measured
`by the wetted wall or film temperature. This temperature is near the critical temperature of most
`jet fuels. Two explanations have been offered for this decrease in deposition under supercritical
`conditions. One explanation is that the decrease reflects the temperature where hydroperoxide
`precursors to solids formation are depleted by thermal decomposition, interrupting the radical
`chain reactions forming solids. Another explanation is that the solvent properties of the fuel become
`enhanced under supercritical conditions, with the surface deposition reduced essentially by keeping
`the solids in solution. In single-tube heat exchanger tests with pure hydrocarbons and jet fuels of
`widely varying critical temperature, it was found that the deposition decrease is insensitive to fuel
`critical temperature but very sensitive to residence time/heating rate, indicating that the deposition
`decrease is a fuel chemistry effect rather than an effect of the supercritical nature of the fuel.
`
`Introduction
`Development of thermally stable jet fuels has been
`recognized as a critical research area because an aircraft's
`fuel is the only practical means of absorbing dramatically
`increasing on-board heat loads. The thermal stability of
`a fuel is a measure of its tendency to cause fuel system
`problems, such as surface deposits or filter plugging, upon
`heating during aircraft operation (Hazlett, 1991). Indeed,
`the limited thermal stability of current fuels must be
`overcome in order to fully realize the system capabilities
`of advanced aircraft. The current practice for Air Force
`engine designers is to limit the fuel to 163 °C (325 °F) bulk
`temperature and 205 °C (400 °F) wetted wall temperature.
`This "rule-of-thumb" is based on engine fouling experience.
`In the long term, the heat management requirements of
`future Air Force aircraft are expected to be sufficiently
`demanding that the aircraft fuel system will need to operate
`at much higher temperatures. The goal for this long term
`fuel is to increase the thermal stability temperature limit
`to480°C (900°F) (Edwards, 1993). This high temperature
`is above the fuel thermodynamic critical temperature,
`which typically ranges from 370 to 415 °C (700-780 °F)
`for current jet fuels (Martel, 1978). Typical fuel system
`pressures are approximately 34-68 atm (500-1000 psia),
`well above the fuel critical pressure. Calculated critical
`data for typical fuels are shown in Table I. The physical
`and chemical properties of supercritical fuels are expected
`to be dramatically different from gases and liquids. There
`is little information available on the properties of fuels
`when operated at supercritical conditions (Edwards, 1993).
`One area of concern for fuels operating at high tem(cid:173)
`peratures is their thermal decomposition and the formation
`of insoluble products that deposit onto metal surfaces
`(fouling). This area has been studied in heated metal tubes
`(simulating aircraft nozzles and heat exchangers) by
`various investigators (Hazlett, 1991; Hazlett et al., 1977;
`Bradley et al., 1974; TeVelde and Glickstein, 1983;
`Marteney and Spadaccini, 1986; Taylor, 1974; Faith et al.,
`1971; Chin and Lefebvre, 1992). These high-temperature
`tests cover a wide range of flow rates and residence times.
`The residence time of a given fuel molecule inside the
`heated tube is typically on the order of seconds, while the
`test time (or the amount of time the metal surface is
`
`Table I. Critical Properties of Current Jet Fuels (Martel,
`1978) and n-Paraffins for Reference•
`Pc,atm
`Pc, psia
`Tc, °F
`Tc, °C
`34.0-27.2
`500-400
`288-365
`55o-690
`JP-4 (Ce.0H16.9)
`20.4-19.0
`382-415
`720-780
`300-280
`JP-5 (Cu.9H22.2)
`JP-7 (Cl2.1H2u)
`18.4
`270
`405
`760
`350-275
`370-405
`700-760
`23.8-18.7
`JP-8 (C10.9H20.9)
`296
`564
`24.8
`365
`n-CeH1e
`726
`17.9
`263
`386
`n-C12H2s
`13.6
`200
`447
`836
`n-C1sH34
`a Critical temperatures for fuels were calculated from the specific
`gravity and boiling curve of the fuels by the method outlined in
`ASTM D2889. Fuel critical properties are listed as a range because
`of varying boiling curves for different samples. Commercial Jet A
`(C11.sH22.o) and Jet A-1 fuels are similar to JP-5 and JP-8. Fuel
`composition data from (Martel, 1988).
`
`exposed to heat and fuel flow) is on the order of hours.
`Most researchers have seen a peak in surface deposition
`near 370 °C (700 °F) when the deposition is plotted as a
`function of wetted wall or film temperature. This is
`illustrated in Figure 1 for two sets of data from different
`test devices (Taylor, 1974; TeVelde and Glickstein, 1983).
`The peak disappears if the dissolved oxygen in the fuel,
`typically about 65-75 ppm (w/w) in air-saturated fuel, is
`removed (Bradley et al., 1974; Taylor, 1974). One goal of
`our work is to investigate the source of this deposition and
`determine ways to reduce, avoid, or eliminate it without
`resorting to removing the dissolved oxygen from the fuel.
`The decrease in deposition above approximately 370
`°C has been given both physical and chemical explanations.
`According to Taylor (197 4), "the sharp drop in autoxidative
`deposit formation rates ... would appear to reflect an effect
`of the transition between the liquid phase and the
`supercritical vapor phase." Hazlett (1977) prefers a
`chemical explanation: "The deposit rating decreases above
`385 °C. Since hydroperoxide is depleted at this temper(cid:173)
`ature, the tie-in between hydroperoxide decomposition
`and deposit formation is reinforced." Similarly, Marteney
`and Spadaccini (1986) state "the sudden decrease in
`deposit formation at temperatures above 645 K [372 °C]
`may be indicative of a depletion of oxygen in the fuel,
`related to the known decomposition of hydroperoxides at
`elevated temperatures."
`
`0888-5885/93/2632-3117$04.00/0
`
`© 1993 American Chemical Society
`
`

`
`3118
`
`Ind. Eng. Chem. Res., Vol. 32, No. 12, 1993
`
`0.05
`
`air sparge vent to hood
`
`570 mVmin
`
`500
`
`0.00
`600
`
`0.00
`100
`
`200
`
`300
`400
`Wall temperature, ·c
`Figure 1. Surface deposition data for JP-5 in UTRC single-tube
`heat exchanger test (TeVelde and Glickstein, 1983) and in Esso
`Advanced Kinetic Unit (AKU) (Taylor, 1974). UTRC conditions: 28
`atm, 570 mL/min (60 lb/h, inlet Re= 3000), 2.4-m-long, 2.16-mm-i.d.
`316 SS tube, 480 °C (900 °F) fuel outlet temperature, 19 h test
`duration. AKU conditions: 69 atm, 10 mL/min (inlet Re= 64), 122
`-cm-long, 2.1-mm-i.d. 304 SS tube, 4 h test duration.
`
`The research described in this paper was motivated by
`a desire to discover if the decrease in surface deposition
`above about 370 °C, which creates the deposition peak, is
`indeed due to supercritical fuel properties. Although other
`investigators have studied a wide variety of conditions,
`most of the fuels and hydrocarbons tested had very similar
`critical temperatures. One of the approaches of this work
`was to investigate several similar hydrocarbons (e.g., the
`normal paraffins) with widely varying critical tempera(cid:173)
`tures, but presumably similar deposition chemistry, to
`determine the effect of fuel critical temperature on the
`deposition behavior. A second approach was to test the
`effect of fuel flow rate on the deposition-temperature
`behavior above 370 °C. The flow rate affects the wall/fuel
`temperature difference, the residence time, and the heating
`rate. Pure hydrocarbon fuels are fairly expensive, so JP-7
`was used in most of the tests as the baseline hydrocarbon
`fuel. JP-7 (MIL-T-38219) is the Air Force's highest
`thermal stability fuel and consists mainly of paraffins and
`cycloparaffins. Finally, a representative Jet A-1 fuel was
`studied for comparison to the JP-7 and pure hydrocarbon
`results. These results will be used to develop computa(cid:173)
`tional fluid dynamic and chemistry (CFDC) models of
`fuel system components (Katta and Roquemore, 1993).
`
`Experimental Section
`The apparatus used in this study is shown in Figure 2.
`The fuel is air-sparged to assure air saturation of the fuel.
`The flow system is initially purged with nitrogen prior to
`fuel introduction to remove any air present. The fuel is
`pumped through the system with a high pressure liquid
`chromatography pump. An initial 0.45-µm filter is used
`to remove any particulates present from fuel handling and
`to maintain a consistent filtration level between fuels. The
`test section consists of an 89-cm-long, 3.2 mm (0.125-in.)
`o.d., 1.4-mm (0.055-in.) i.d. 304 stainless steel tube passing
`through a Lindbergh laboratory furnace. The tube makes
`a 180° bend inside the furnace, thus passing through the
`30-cm (12-in.) actively heated portion of the furnace twice.
`After exiting from the furnace, the products are cooled to
`room temperature and filtered before passing through a
`back pressure valve which regulates the system pressure.
`A 5-µm filter is used to prevent particulates from fouling
`the back pressure valve; the filter also gives a visual
`
`lo AKU, 10 mVmin
`"' ~
`E
`~ 0.30
`E
`a.
`a.
`c
`2
`"iii
`0
`a.
`Cl)
`"O
`c:
`~
`"'
`
`0.40
`
`0.35
`
`0.25
`
`0.20
`
`0.15
`
`0.10
`
`0.05
`
`(,)
`
`"'
`~ 0.04 E
`~
`E
`a.
`a.
`c
`0
`:;::;
`"iii
`0
`a.
`Cl)
`"O
`c:
`0
`-e
`"'
`
`0.03
`
`0.02
`
`(,)
`
`0.01
`
`0
`
`0
`
`I
`l
`
`fuel
`tank___.,
`(atmospheric
`pressure)
`
`purge
`
`N
`2
`
`0.45 µm
`filter
`
`5 µm filter
`
`Figure 2. Schematic of apparatus.
`
`assessment of particulates formed in the heated section.
`A three-way valve downstream of the back pressure valve
`allows the product flow to be diverted to a sampling system
`where the liquid products are analyzed (off-line) at ambient
`conditions. The stainless steel test section is cut into 5-cm
`(2-in.) sections and analyzed by carbon burnoff (Leco
`surface carbon analyzer) to measure carbon deposition.
`The tube wall temperature distribution is measured by
`K-type thermocouples spot-welded to the outside of the
`tube. The fuel temperature at the furnace outlet is
`measured by a thermocouple inserted into the flow.
`Pressure was measured between the filter and the back
`pressure valve and also at the pump. The fuel temperature
`distribution is estimated with a heat transfer program
`(Edwards, 1992; Edwards and Anderson, 1993) from the
`measured wall temperature distribution, fuel flow rate,
`and the initial temperature of the fuel entering the furnace
`(20 °C). The fuel properties (density, thermal conduc(cid:173)
`tivity, viscosity, heat capacity) vary significantly through(cid:173)
`out the tube in a given test where the fuel temperature
`might vary from 20 to 620 °C (70-1150 °F) (Edwards,
`1993), so the calculation uses temperature-dependent fuel
`properties. It is found that test times on the order of
`10-20 hare required for JP-7 and pure hydrocarbons to
`give reproducible deposition results above the measure(cid:173)
`ment background. For more typical (lower thermal
`stability) fuels, deposition amounts are much larger and
`test times can be shorter.
`
`Results and Discussion
`
`A typical result of a JP-7 experiment is shown in Figure
`3. As described in the Introduction, the data is.reduced
`by plotting the deposition (scaled by the tube surface area
`and the total fuel flow) as a function of wall temperature,
`excluding points of decreasing wall temperature after the
`tube exits from the furnace. Thus, the deposition is
`reported as µg of deposit carbon/(g of fuel fed·cm2), or
`ppm/cm2• The deposition-wall temperature results for
`several 12 mL/min flow rate experiments are shown in
`Figure 4. Various tests are shown spanning different
`ranges of wall temperature. The decrease in deposition
`near 440 °C (825 °F) is very similar to the behavior shown
`in Figure 1, although the deposition decrease begins at a
`temperature significantly higher than the fuel critical
`
`

`
`-B-- WallT, C
`-8- Fuel T, C (calc:) ~ C depos~ion, µg
`~ Fuel T, C (meas)
`
`140
`
`120
`
`80
`
`60
`
`100 C)
`::I.
`c
`.g
`'iii
`0 a.
`(Jl
`'O
`c:
`0 -e
`"' (.)
`
`40
`
`20
`
`600
`
`500
`
`400
`
`300
`
`200
`
`100
`
`~
`::>
`~
`~
`E
`~
`
`20
`
`40
`60
`Distance along tube, cm
`Figure 3. Experimental results from 12.0 mL/min test. Air-sparged
`JP-7 fuel, 69 atm, 21 h.
`
`80
`
`0
`100
`
`---e- 6/92-4, 21 hrs
`--B-- 7/92-2, 25 hrs
`~ 5/92-10, 21 hrs
`~ 5/92·1, 20 hrs
`--+-- 5/92-2, 7 hrs
`12 mVmin, JP-7
`
`fuel T - 266 'C
`res time - 2 sec
`Re-700
`
`/
`
`¥
`
`N
`
`0 .005
`
`~
`~ 0 .004
`g
`~ i!l
`a.
`Cl)
`'O
`~ 0.002
`"' (.)
`
`0.003
`
`0.001
`
`0
`
`100
`
`200
`300
`400
`500
`Wall temperature, ·c
`Figure 4. Combined results from 12 mL/min tests covering various
`wall temperature ranges. Air-sparged JP-7 fuel, 69 atm. Residence
`times and Reynolds number of bulk fuel at peak deposit location
`calculated with temperature-dependent fuel properties. Lines connect
`points in order of axial distance.
`
`600
`
`700
`
`temperature of 400 °C (760 °F). The dependence of this
`deposition decrease on residence time/flow rate was
`investigated by running similar tests at 33.5 and 3.1 mL/
`min flow rates. The results of these tests are shown in
`Figures 5 and 6, respectively. It is immediately apparent
`that the autooxidation deposition decrease has shifted
`significantly with the varying flow rates and residence
`times. As fuel flow rates are increased, the difference
`between the fuel and wall temperature increases, a simple
`result of the increasing heat-transfer requirements to heat
`up the increasing amount of fuel. Thus, the fuel tem(cid:173)
`perature at the peak has also changed. If the deposition
`decrease is a simple consequence of the enhanced solid
`solubility of the fuel under supercritical conditions, the
`peaks in Figures 4-6 should occur at the same wall
`temperature (since the deposit would be dissolved by the
`fuel near the wall which should be at a temperature near
`that of the wall). The bulk fuel temperature at the point
`where the deposition decrease begins is well below the
`critical temperature for most tests. Since the deposition
`decrease occurs at differing temperatures at different flow
`
`Ind. Eng. Chem. Res., Vol. 32, No. 12, 1993 3119
`
`---e- 4/92-3, 4.5 hrs
`--B-- 4/92-4, 17.5 hr
`
`33.5 mVmin, JP-7
`
`fuel T - 225 ·c
`res time - 0.8 sec
`Re - 1500
`
`N8 0.008
`l
`g 0.006
`E
`"' ~ "O
`~
`(.)"' 0.002
`
`0.004
`
`o
`
`0.000-+-r~~~~~~~~~~~~~~~~~+-
`100
`200
`300
`400
`500
`600
`700
`800
`Wall temperature, ·c
`Figure 5. Results from 33.5 mL/min tests. Air-sparged JP-7 fuel,
`69 atm.
`
`---e- 8/92-1, 13 hr
`--B-- 8/92-2, 30 hr
`
`3.1 mVmin, JP-7
`
`fuel T - 294 ·c
`res time - 7 sec
`Re-200
`
`0.020
`
`0.015
`
`0.010
`
`0.005
`
`0
`
`100
`
`200
`300
`400
`500
`Wall temperature, ·c
`Figure 6. Results from 3.1 mL/min tests. Air-sparged JP-7 fuel, 69
`atm.
`
`600
`
`700
`
`rates, the deposition-temperature behavior is more likely
`a result of a chemical, rather than a physical (supercritical)
`process.
`It is notable in Figures 4-6 that the background
`deposition levels are highest for the shortest test times.
`For short test times (<10 h), it is found that the deposition
`results at temperatures above and below the autooxidative
`peak were sufficiently close to background detection levels
`that the data are fairly scattered. This can be seen in
`Figure 4 (test 5/92-2) and Figure 5 (test 4/92-3). Apparent
`background levels varied somewhat for the various tests,
`but typical levels were on the order of 7 µg/cm 2• Recent
`tests with lower stability fuels have shown an induction
`time, or delay before the initiation of deposition, of about
`3 h under conditions similar to Figure 4 (Edwards and
`Liberio, 1994).
`The results for the varying flow rates are summarized
`in Table IL In addition to the change in deposition peak
`location, the deposition (expressed as ppm) decreases with
`increasing flow rate. The only published results for varying
`flow rates at high temperatures were those of UTRC
`(Te Velde and Glickstein, 1983; Marteney and Spadaccini,
`1986). UTRC examined a wider range of flow rates in
`their heated tube tests. For inlet Reynolds numbers
`ranging from 400 to 21 000, the deposition rate (in µg/
`(cm2·h) appeared relatively constant in the 225-325 °C
`(435-615 °F) wall temperature range (Marteney and
`
`

`
`3120
`
`Ind. Eng. Chem. Res., Vol. 32, No. 12, 1993
`
`Table II. Autoxidation Deposition Peak (Averaged over
`5-cm Tube Segment) Parameters for JP-7 as a Function of
`Flow Rate•
`
`parameters for
`autooxidation deposition peak
`wall temperature, °C
`fuel temperature, °C
`residence time to peak, s
`Reynolds number at peak
`peak deposition, ppm/cm2
`peak deposition, µg/(cm2·h)
`
`flow rate, mL/min
`12.0
`33.5
`441
`523
`225
`266
`0.8
`2
`1500
`700
`0.0036
`0.0037
`5
`2
`• Deposition values reported are for longest test reported, with
`background (7 µg/ cm2) subtracted.
`
`3.1
`390
`294
`7
`200
`0.0060
`1
`
`Spadaccini, 1986). Expressed as ppm/cm2, the UTRC
`deposition levels decrease with increasing flow rate. The
`tests were not carried out to high enough wall temperatures
`for the lowest and highest flow rates to determine if the
`deposition decrease shifted with varying flow rate. Aside
`from the UTRC data, no published results for varying
`flow rates were found which would confirm this shift in
`the autooxidative peak location. However, recent work at
`Southwest Research Institute has produced similar results
`(Moses, 1992).
`Focusing on the wall temperature alone is an obvious
`oversimplification, since the deposition (as a chemical
`process) is also a function of bulk fuel temperature and
`residence time. Flow velocity may also be important for
`deposit erosion. The importance of other parameters is
`apparent in results obtained by Shell Development (Faith
`et al., 1971), where two deposition peaks (autooxidative
`and pyrolytic) are seen in a heated tube test where the
`wall temperature is relatively constant at 650 °C (1200
`°F), but the fuel temperature increases from 260 to 510
`°C (500-950 °F). It seems likely also that deposition peaks
`could form in tests with constant fuel and wall temper(cid:173)
`atures, where the peaks would arise from consumption
`and/or exhaustion of reactive species. Nozzle fouling
`simulations are usually performed with preheated fuel
`flowing through hot metal tubes at relatively constant
`temperature. Under these conditions, a deposition peak
`may be absent (Chin and Lefebvre, 1992; Hazlett, 1991).
`Another way of testing the supercritical solvency hy(cid:173)
`pothesis is to test fuels of similar structure but varying
`critical temperature. This was done by testing n-octane
`(Tc = 296 °C (564 °F)) and n-hexadecane (Tc = 447 °C
`(836 °F)) under conditions similar to those for the JP-7
`tests. The results are shown in Figures 7 and 8. Again,
`it can be seen that the critical temperature of the fuel does
`not appear to control the deposition behavior in these
`tests. The deposition decrease occurs at essentially the
`same temperature (both fuel and wall) for all of the
`hydrocarbons, regardless of their critical temperature. The
`deposition levels for the pure hydrocarbons are very similar
`to that of JP-7.
`The deposition rates of typical fuels are much higher
`than that of JP-7 or pure hydrocarbons (Bradley et al.,
`1974; Taylor, 1974; Frankenfeld and Taylor, 1980). The
`thermal stability of an aircraft fuel can be assessed by the
`temperature at which it fails the specification jet fuel
`thermal stability test (the Jet Fuel Thermal Oxidation
`Tester (JFTOT), ASTM 3241). This is referred to as the
`breakpoint temperature. Thus, Taylor measured a dep(cid:173)
`osition of 0.4 ppm/cm2 for a JP-5 (274 °C breakpoint)
`(Taylor, 197 4) and UTRC measured a deposition of 0.04
`ppm/cm2 for another JP-5 (breakpoint 270 °C) (TeVelde
`and Glickstein, 1983), as shown in Figure 1. These
`deposition levels are an order of magnitude larger than
`the levels seen in JP-7 and the pure hydrocarbons in
`
`0.005
`
`-6- JP·7, 21 hr
`--8- n-C
`H
`, 13.5 hr
`34
`16
`
`0.004 ~ n-C H
`, 19 hr
`16
`34
`12 ml/min
`
`"'
`E
`-!::!
`E a. a.
`c 0.003
`,g
`'iii
`0 a.
`Cl)
`"O
`c
`0 -e
`"' () 0.001
`
`0.002
`
`O.OOO-+-.~~~~~~~~~~~~~~~T"'r'.-.-1-
`0
`1 00
`200
`300
`400
`500
`600
`700
`Wall temperature, ·c
`Figure 7. 12 mL/min deposition results for n-hexadecane and JP-7.
`Air-sparged fuels, 69 atm.
`
`·--e- JP-7, 13 hr
`--8- JP-7, 30 hr
`, 22 hr
`
`~ n-C H
`8
`18
`
`3.1 ml/min
`
`0.015
`
`0.020
`
`"'e
`~
`8:
`-
`§
`~
`0 a.
`~ 0.010
`c
`~
`"' () 0.005
`
`O.OOO-+-.~~~~~~~~~~T"'r'T"'r'T"'r'T"'r'T"'r'T"'r'.-.-1-
`0
`1 00
`200
`300
`400
`, 500
`600
`700
`Wall temperature, ·c
`Figure 8. 3.1 mL/min deposition results for n-octane and JP-7.
`Air-sparged fuels, 69 atm.
`
`--9- JP-7, 21 hr
`--8- Jet A-1 2747, 22 hr
`
`"'e
`-!::! 0.15
`E
`a. a.
`~ 'iii 0.10
`g_
`
`Cl)
`"O
`c
`~ "' 0.05
`
`()
`
`0.00 --h-.-r.Q;=;ij!@~-Eil~liir#~;$l:~9!!-~~+
`0
`1 00
`200
`300
`400
`500
`600
`700
`Wall temperature, ·c
`Figure 9. Deposition results for various fuels. Air-sparged fuels, 12
`mL/min, 69 atm.
`
`Figures 4-8. In Taylor's test, the total carbon deposit in
`the tube from JP-5 was 10 times that of JP-7 (Taylor,
`1974). The results in the present test for a Jet A-1 fuel
`(333 °C breakpoint) and JP-7 are shown in Figure 9. The
`deposition level is similar to that reported by Taylor for
`aJP-5fuelatasimilarflowrate (Figure 1). More extensive
`
`

`
`Table III. Results of Gas Chromatographic Analysis of
`.n-Octane and n-Hexadecane after Thermal Stressing
`
`species
`propane and propylene
`1-butene and n-butane
`1-pentene
`n-pentane
`1-hexene
`n-hexane
`1-heptene and n-heptane
`1-octene
`n-octane
`n-octane and impurities
`1-nonene
`n-nonane
`1-decene
`n-decane
`1-undecene
`n-undecane
`1-dodecene
`n-dodecane
`1-tridecene and n-tridecane
`1-tetradecene
`n-tetradecane
`1-pentadecene
`n-pentadecane
`Ca dimers
`n-hexadecane and impurities
`other unidentified products
`
`GC area 3
`n-hexadecane test
`n-octane test
`0.125
`0.241/0.274
`0.352
`0.328
`0.448
`0.048
`0.188
`
`0.052
`0.056
`0.089
`0.109
`0.101
`0.197 /0.106
`0.182
`0.119
`
`97.65
`
`0.275
`
`0.071
`
`0.178
`0.125
`0.180
`0.129
`0.207
`0.130
`0.196
`0.126
`0.314
`0.150
`0.148
`0.107
`0.066
`
`95.99
`0.94
`
`tests with various fuels and pure hydrocarbons are
`underway (Edwards and Liberio, 1994).
`
`Chemical Analysis of Products
`Gas chromatographic (GC) analyses were performed on
`samples of stressed n-octane and n-hexadecane for de(cid:173)
`termination of the reaction products. The products were
`identified by a mass spectrometer detector interfaced to
`the GC and quantified using a flame ionization detector
`(Fill). The n-octane sample was subjected to maximum
`fuel and wall temperatures of 480 and 540 °C (900 and
`1000 °F), respectively (Figure 8). The n-hexadecane was
`subjected to maximum fuel and wall temperatures of 470
`and 620 °C (875 and 1150 °F), respectively (Figure 7). The
`product yields were estimated by assuming that the FID
`sensitivity to individual compounds is proportional to the
`number of carbon atoms present in the molecule. There(cid:173)
`fore at small conversions of the parent alkane to products,
`the integrated areas of the GC peaks can be related to the
`fraction of the carbon atoms in the original alkane that
`react. The products observed for these two experiments
`are shown in Table III. The data in the table indicate the
`3-4 3 of the carbon of the reactant alkane has reacted and
`the primary products observed are a series of alkanes and
`alkenes, with shorter carbon backbones than the parent
`alkane. For n-octane, a series of six peaks that appear to
`be dimers of C8H 17 radicals, that is isomers of C16H34, are
`also found. The mass spectra of these peaks do not yield
`the molecular ion; identification as C1s species was obtained
`by comparison of GC retention times with other alkanes,
`comparison with mass spectra of other large alkane isomers,
`the prominent C8H 17 mass spectral peak, and the number
`of isomers found as discussed below. For n-hexadecane
`the corresponding Ca2Hss species was not found, although
`this could be due to these species not eluting from the
`chromatographic column due to their expected very high
`boiling points under the chromatographic conditions used
`for these analyses.
`The C1sH34 isomers result from the recombination of
`sec-C8H 17 radicals. The formation of secondary alkyl
`
`Ind. Eng. Chem. Res., Vol. 32, No. 12, 1993 3121
`radicals is favored as the secondary C-H bond strength
`is 3 kcal/mol less than the primary C-H bond. There are
`three possible sec-CsH17 radicals that can be formed by
`H-atom abstraction from n-octane. These three sec
`radicals can recombine to yield six possible C8 dimers.
`The fact that we found six dimer species as products further
`confirms their identity.
`The major products found, the homologous series of
`alkanes and alkenes, are similar to those observed for GC
`analyses in the low-temperature (160-180 °C) reaction of
`n-dodecane (Zabarnick and Striebich, 1992), and in the
`pyrolytic cracking of alkanes at high temperature. It is
`thought that these alkanes and alkenes are products Of
`the decomposition of alkyl hydroperoxides (Reddy et al.,
`1988) at temperatures below 480 °C (900 °F) and from
`pyrolysis reactions above 480 ° C (Hazlett, 1977). The alkyl
`hydroperoxides are produced from the au