the effect of coatings and liners on heat transfer in a ... · and the patran solid modeler...
TRANSCRIPT
![Page 1: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/1.jpg)
NASA AVSCOMTechnical Memorandum 102513 Technical Report 90-C-006
The Effect of Coatings and Linerson Heat Transfer in a DryShaft-Bush Tribosystem
AD-A229 669Mihir K. GhoshLewis Research CenterCleveland, Ohio
and
David E. Brewe* r
Propulsion Directorate * ,.*
U.S. Army Aviation Systems CommandLewis Research Center 11Cleveland, Ohio
August 1990 4 3"a
TUTION 5TAhtm. A j--" oe tor P',b' c rel~c"
ADiWbutin orU plai bt|ed US ARMY_...-AVIATIO
SYSTEMS COMMANDAV1ATION RAT ACTMIY
9' 1
![Page 2: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/2.jpg)
THE EFFECT OF COATINGS AND LINERS ON HEAT TRANSFERIN A DRY SHAFT-BUSH TRIBOSYSTEM
Mihir K. Ghosh*National Aeronautics and Space Administration
Lewis Research CenterCleveland, Ohio 44135
and
David E. BrewePropulsion Directorate
U.S. Army Aviation Systems CommandNASA Lewis Research Center
Cleveland. Ohio 44135
Summary tribologists in recent years. Friction and wear of dry contactsare dependent on the overall temperature increase at the
The temperatures due to frictional heating within a solid interface. Thermomechanical effects, such as the initiation oflubricated or coated journal bearing were analyzed by using thermal cracks and the low-cycle thermal-fatigue failure in dry
a finite element method. A solid model of the shift-bush bearings and seals, are dependent on the heat dissipationtribocontact was generated with an eight-node. three- mechanism and the overall temperature rise. Floquet. Play.dimensional, first-order isoparametric heat-transfer element and Godet (ref. 1) used a two-dimensional Fourier transformand the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contactMarc) translator was used to help develop the Marc-based temperatures in dry bearings operating with a plastic liner.finite element program for the system: this software was used Later. Floquet and Play extended this technique to three-on the Cray X-MP super computer to perform a finite element dimensional prblems (ref. 3). They showed that small changesanalysis of the contact. The analysis was performed for various in the design may bring large differences in contactliner materials, for thin. hard, wear-resistant coated bearings, temperatures, and that to reduce the maximum temperature.and for different geometries and thermal cooling boundary the rubbing surfaces must be as close as possible to a controlledconditions. The analyses indicated that thermal conductivity temperature heat sink and must not be shielded by any thermalof the liner or coating material is the most vital thermal barrier. Gecim and Winer (refs. 4 and 5) extended the Fourierparameter that controls the interf'-:e temperature. In addition transform method to calculate steady temperatures in a rotatingto design variations, the proximity of the cooling source to cylinder that had multiple surface heat sources and a varietythe heat-flux-generating interface is critically important to the of geometries and thermal boundary conditions.
temperature control in the system. In recent years. high-temperature applications brought about
the need for hard. wear-resistant coatings (viz., AO 3. andIntroduction ZrO,) to be used as thermal barriers in mechanical
components. When these components are subjected to heavily
The temperature rise due to frictional heating at the interface loaded sliding conditions, "heat checking" often occurs: that
of tribocontacts has attracted the attention of engineers and is. cracking in the neighborhood of the contact zone due tothe combination of thermal heating and mechanical load. This
*N~iiinaI Research (Cm I - NASA Research A~soctate at LCwIAIS RCwartch often occurs in such mechanical devices as seals and brakes.
Ccnicr. on Ica'.c Iri 1 Mcchanical nuinccring )cpt. IT.. B.ttU.. Kennedy. et al. (refs. 6 to 8) have done thernomechanicalVarnaii. 221005. Intia. analyses of dry sliding contacts, using a finite element
![Page 3: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/3.jpg)
technique. In particular. a recent investigation (ref. 8) R heat load matrixaddressed the problems of hard. wear-resistant coatings on a R1,, inner radius of the hearing. inductile metallic substrate, as applied to mecnanical face seals.A thermo-elasto-plastic analysis of stresses and deformationswas used. S surface area, m2
Ju and Chen (ref. 9) used a Fourier transform method to T temperature, °Cinvestigate the mechanical stresses, temperature field, and time. sec
thermal stresses due to a moving frictional load. Ju and LiuLV rubbing velocity. nm/see(refs. 10 and 11) extended this technique to study thephenomenon ofthermonechanical cracking in layered media. xy.z Cartesian coordinatesLeroy. Floquet. and Villechaise (ref. 12) also used a fast F surface of integration. m
Fourier transform algorithm to study thermomechanical K thermal diffusivity, m/secbehavior of layered media.
Fourier transform techniques generally provide useful P density. kg/nvanalytical expressions, but they are limited in applications 12 volume of integration. m3
dealing with irregular geometries and boundaries. Insofar as W, angular speed of the shaft, rad/secwe know. dry bearing temperatures have not been investigatedwith the finite element technique that accommodates irregular Subscripts:boundaries. Herein. results of a finite element analysis of drycontact temperatures due to frictional heating at the interface c conductanceof a shaft-bush-housing system are presented. A solid model e element, in a senseof the system was generated with the Patran solid modeler lobalsoftware. and the finite clement analysis wa, done with the h surface convectionMarc finite element software. The temperature distribution in /I maximumthe system was determined with various liner materials, coating q surface heatingmaterials, coating thicknesses, and bearing geometries. s surface
This work also shows that the method can serve as acomputer-aided design technique to model mechanical T nodal temperatures
components (such as the cylinder-piston of an advanced lowheat rejection (LHR) diesel engine or Stirling engine, and Superscript:aircraft control and undercarriage bearings) for temperaturerise and thermal distortions under specific operating e element, in a local sense
conditions.Theory
Symbols The theory discussed herein is taken from reference 13.
[B] temperature gradient interpolation matrix Finite Element Formulation
[CI capacitance matrix The steady-state or transient heat transfer in a three-
c specific heat, J/kg 'C dimensional anisotropic solid moving with velocity V isgoverned by the Fourier heat conduction or energy equation..f friction coefficient at the contactwhcisrtena which is written as
I film heat transfer coefficient. W/m 2 °C
IKI heat transfer or coefficient matrix [paT) 7"0
k thermal conductivity. W/m TC v•k vT-pc - V. vT1 + Q 0 (I)
M number of finite elements
[NI temperature interpolation matrix where Q(xvz,t) is the internal heat generation rate per unit
n unit vector volume. The solution for the temperature distribution mustp contact pressure. Pa satisfy the boundary conditions and initial conditions on all
cnatportions of the surface. The inertia conditions, which specifyQ internal heat generation the temperature at time zero, are required for the transient
q heat flux distribution, W/m or N/cr-sec analysis. The heat conduction boundary conditions are as follows:
![Page 4: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/4.jpg)
T= T, where
q.n-=KvTn (2) [C= pcN [NJdi
h(T,- T,,) .n =k vT. n c = BI[KJJBdQ
where
T, prescribed temperature in the specified region [K1 = h N [Nidi
S1 of the surface [ = NN'3
-K v T-n prescribed heat flux normal to the surface S,k v T.n convection to exchange temperature T, from the RQ = Q N d2 (6)
surface with unknown temperature T U0
The solution domain in the solid is discretized into M finiteelements possessing r nodes. Temperature and temperature Rq = q, N dIgradients within the element are expressed in terms of nodal AItemperatures as
Rh, = hT, N dFTe(x.y,z,t) = [N(x.y,z,t)] T(t) '3
and RT =- (q.n) N dF
OT(xy.,t)
at The coefficient matrix [C] of the time derivative of the nodal
aT temperatures is the element capacitance matrix. The coefficient- (xyz') = [B(x,y,z)] T(t) (3) matrices [K,.] and [K,] are element conductance matrices andat relate to conductance and convection.
aT The vectors Rr. RQ. Rq, Rh are heat-load vectorst (x.,y,z,t) arising from specified nodal temperatures, internal heat
at generation, specified surface heating. and surface convection.
respectively. For steady-state analysis, equation (j) reduces towhere [NJ is the temperature interpolation matrix, and [B] is
the temperature gradient interpolation matrix such that,
[N(x.y,z)] = [NI, N2 ... . [,] liKce + [Kh1J T = RQ + pq + Rh (7)
and Problem Statement and Solution Methodology
aN, ON, aN, In the present work a shaft-bush/liner-housing type of
Ox 7A ax tribosystem was analyzed. The system, shown in figure 1.
aN, ON, ONr consists of a hollow shaft (1) rotating in a bush-liner (2) fitted[B(x.y.z)J = N a, (4) within a stationary housing (3). Alternatively. the shaft rotates
Ov yin a housing with a thin, hard, wear-resistant ceramic coating
at its internal surface. There is frictional rubbing at theN1 ON2 .. ONr interface of the liner or coating and the shaft. Frictional heating
_ L a z_ at the interface is conducted to the shaft and the bearing andhousing, thereby causing a temperature rise in both. The main 4
Finite element equations are derived from the energy equation objective of the present study was to develop a finite element(eq. I) by using the method of weighted residuals (Galerkin's model of the previously described system by using the Patranmethod): solid modeler software, and thereafter to obtain a Marc finite 0
element program by using a Patran-Marc (Patmar) translator.(O7) Following this, we analyzed the problem for temperature
ICI + [IK,IJ + IKh1i 11 distribution within the solid. using the Marc finite elementsoftware, subject to the proper boundary conditions and the
= RQ + Rq + R4 + RT (5) appropriate material thermal properties. The results were -des
k3
![Page 5: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/5.jpg)
postprocessed with the Patran postprocessor. The boundary Case I-Bearings with Soft Liner Materialsconditions for the systenm analyzed wereconitions A he tansteroeffiintanatoz8dwThermal properties of various liner materials are shov, n in
equal t8W/n table Ill. The shaft and the housing were made of steel. Fora reference temperature of 20 *C. caused by natural
convection a 6.0-m thick linier, the maximum interface temperaturesobtained for various liner materials are shown in table IV. Thelowest of the maximum temperatures "as 108.5 °C for the
the shaft, because of direct cooling(3) An imposed temperature of 35 "C at the periphery of graphitized metal liner, which had the highest thermal
conductivity and diffusivitv values. As the material'sthe hosd becpatuodreof C cling conductivity and diffusivity decreased, the temperature at the(4)interace increased. In the absence of convection the thermal
by direct cooling at the inner radius of the shaft inderfacevi ofrte In the matil' tietermsconductivity of' the liner was the material's single most
Shaft speeds of 2.6. 6.5. and 13.0 rad/sec were used tocalculate the heat flux generated at the interface. A nodal heat important thermal property' influencing both the temperature
rise at the interface and the temperature distribution as sellflux q equal to the product of the contact pressure p. the friction (fi. 2). Table V shows the variation of maimum interfacecoefficient f and the rubbing velocity V was input to the C.
interface between the shaft and bush (or coated) housing. An emperature with five different heat distribution functions forh o ca resin liner material (ref. I ). while keeping all other conditionsharmonic contact pressure distribution extending from + 90' L
was assumed (i.e.. the journal contacts of half the bearing), constant. We observed that the maximum interface temperature"a sue ietejunlcnat fhl h ern) and the temperature distribution can differ significantly %\ithEffects of various liner materials, coating materials, coatingZ_ t the nature of the heat distribution in the contact recion (fie!. 3).thicknesses, and geometries on the temperature distribution The nature of the heat distribution dn the contactThe nature of the heat distribution depends on the contact
and the maximum temperature at the interface wereinvestigated. pressure in the contact region, which in turn depends on thegeometry and the type of loading. The nature of the loading
may vary with the operating conditions.Results and Discussion The effects of the magnitude of the load and the thickness
of the liner material are shown in table VI. Not surprisinglv.
The maximum interface temperatures and the temperature a thicker liner material can result in a lower interface
distribution in the shaft-bearing system are presented herein. temperature, provided the conductivity of the liner material
This system was composed of either a bearing with liner is higher than the conductivity of the bush material. Thus. for
material in a housing. or a bearing with a thin. hard. wear- higher loads or pV factors, a thicker graphitized metal liner
resistant coating on its inner surface. The temperature rise was would help reduce the temperatures.
due to frictional heating at the interface. The hollow shaft haddifferent dimensions (see table 1) for the three cases studied: Case l-Bearings with a Thin. Hard Coating on th(I ) bearings with soft liner materials: (2) bearings with thin. Internal Surface
hard coatings on the internal surface: and (3) variations in Thermal properties of thin. hard. coating materials are givengeometry (table II) and the cooling condition. Nou ,i heat flux in table VII. For a 1.0-mm-thick coating. the variations ingenerated in the contact was estimated to be equal to the maximum interface temperatures are shown in table VIII. Asproduct of the contact pressure. rubbing velocity, and friction is evident from table VII for the thermal properties and tablecoefficient. The coefficients of friction for the liner and coating VIII for the maximum temperatures. the thermal conductivit\materials ranged from 0. 15 to 0.3. However, an intermediate of the coating material determines the magnitude of thevalue of 0.2 was assumed. Our objective was to evaluate the maximum interface temperature. When a coating material withtemperature rise in the contact for the same amount of heat low thermal conductivity was used as a thermal barrier toflux input. A harmonic pressure distribution extending from shield the bearing, high temperatures occurred at the shaft-+90 ° was assumed. bearing interface. thereby dissipating more heat to the shaft
Thus. heat flux distribution and causing higher shaft temperatures.
q = q,,, cos In order to study the effect of coating thickness. calculationswere made for titanium nitride coatings ranging from
where 0.125 to 1.0 mm thick. The resulting maximum interfaceq =fVp,,, temperatures arc shown in table IX. Note that maximum
interface temperatures decreased with decreasing coatingand thickness. This means that thicker coatings with lower
f friction coefficient at the contact. assuned to be 0.2 conductivity than the substrate (such as titanium nitride onsteel) are better thermal barriers than very thin coatings.V rubbing scoC. ity: R,,e, Temperature distributions in the system. which result from
p,,, maximum contact pressure usin" three different thicknesses of ,itaniv'n nitrid- coatiz.
![Page 6: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/6.jpg)
are shown in figures 4 to 6. Similar calculations for various Summary of Resultsthicknesses of tungsten carbide coatings showed almost nochange in maximum interface temperature, since tungsten A Patran solid modeling package was used to generate acarbide has a higher thermal conductivity than the steel finite element model for the thermal analysis of a shaft-bushsubstrate. tribosystem. The Marc finite element software was used to
determine the temperature distribution in the tribosystem.Case III-Variations in Geometry and the Cooling These software permit finite element modeling of bearings andCondition seals so that temperature distributions can be calculated in
specific applications. Mechanical components (viz., theTables X and XI show the variation of maximum interface cylinder-piston of an advanced LHR diesel engine, the Stirling
temperatures for a number of bearing outer radii and bearing cycle engine for space power applications, aircraft control andlengths. Other dimensions remained th same. Higher interface undercarriage bearings, and so on) can be modeled fortemperatures occurred with larger bearing diameters, since temperature-rise and thermal distortions under specificthe cooling source was located at the outer bearing diameter. operating conditions. The results of our analysis can beAn increase in bearing length did not alter the maximum summarized as follows:temperature. For a hollow shaft that was cooled internally, I. If a constant friction coefficient is assumed, thermalthe interface temperature was drastically reduced-from 203 conductivity of the liner or coating material is the mostto 109 °C. The thermal properties of the materials are important thermal property for controlling temperatures inconsidered to be temperature-independent here. Figures 7 to tribocontacts.10 show the changes in the temperature distribution in the 2. Design changes in bearing diameter and length can resultsystem because of geometry variation and inner surface cooling in significant differences in the interface temperature.of the hollow shaft. 3. Interface temperatures can be substantially reduced by
An increase in rubbing velocity or pV factor can cause a cooling the inner surface of the hollow shaft.tremendous rise in the maximum interface temperatures. Table 4. Designs that can provide a cooling medium in closeXII shows that temperatures as high as 928.0 °C can be proximity to the heat-flux-generating interface would be mostobtained at pV = 5362.5 N/cm-sec with a l.O-mm thick effective in reducing the temperature rise in the contact.titanium nitride coating. Figure I I illustrates the thermaldistribution for this case.
References
I. Floquet, A.: Play. D.: and Godet, M.: Surface Temperature in Distributed 8. Kennedy. F.E.: and Hussaini. S.Z.: Thermo-Mechanical Analysis of Dr%Contacts: Application to Bearing Design. J. Lubr. Technol., vol. 99, Sliding Systems. Comput. Struct., vol. 26, no. 1/2, 1987. pp. 345-355.no. 2. Apr. 1977. pp. 277-283. 9. Ju. F.D.: and Chen. T.Y.: Thermrnomechanical Cracking in Layered Media
2. Ling. F.F.: Surface Mechanics. Wiley. 1973. from Moving Friction Load. J. Tribology. vol. 106. no. 4. Oct. 1984.3. Floquet, A.: and Play. D.: Contact Temperature in Dry Bearings: Three pp. 513-518.
Dimensional Theory and Verification. J. Lubr. Technol.. vol. 103. no. 2. 10. Ju. F.D.: and Liu. J.C.: Parameters Affecting Thermo-Mechanical
Apr. 1981. pp. 243-252. Cracking in Coated Media Due to High-Speed Friction Load.4. Gecim. B.: and Winer. W.O.: Steady Temperature in a Rotating Cylinder J. Tribology. vol. 110, no. 2, Apr. 1988, pp. 222-227
Subject to Surface Heating and Convective Cooling. I. Tribol.. vol. 106. II. Ju. F.D.: and Liu. J.C.: Effect of Peclet Number in Thermo-Mechanicalno. I. Jan. 1984, pp. 120-127. Cracking Due to High-Speed Friction Load. J. Tribology. vol. 110.
5. Gecim. B.: and Winer. W.O.: Steady Temperatures in a Rotating no. 2. Apr. 1988. pp. 217-221.Cylinder-Some Variations in the Geometry and the Thermal Boundary 12. Leroy. J.M.: Floquet. A.: and Villechaise. B.: Thermochemical BehaviorConditions. J. Tribology. vol. 108. no. 3. July 1986. pp. 46-454. of Multilayered Media: Theory. J. Tribology. vol. I 11. no. 3. July 1989.
6. Kennedy. F.E.. Jr.: Surface Temperatures in Sliding Systems-A Finite pp. 538-544Element Analysis. 1. Lubr. Technol.. vol. 103. no. I. Jan. 1981. 13. Huebner. K.H.: and Thornton, E.A.: The Finite Element Method forpp. 90-96. Engineers. 2nd ed., Wiley, 1982.
7. Kennedy. F. E., et al.: Improved Techniques for Finite-Element Analysis 14. Malik. R.A.; and Freeman, P.: Development ofa Modified PV-Criterionof Sliding Surface Temperatures. Developments in Numerical and for Dry Bearings. ASLE International Conference on Solid Lubrication.Experimental Methods Applied to Tribology. D. Dowson. et al., eds.. ASLE. 1971. pp. 143-156.Butterworths. 1984. pp. 138-150.
5
![Page 7: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/7.jpg)
TABLE I.-SHAFT SPECIFICATION AND BEARING TABLE IV.-MAXIMUMDIMENSIONS FOR THREE CASES INTERFACE TEMPERATURES
FOR VARIOUS LINERCase Shaft Bearing MATERIALS
Outer Inner Length, Outer Inner Length. [Heat flux distribution q = qcoso:radius, radius, cm radius, radius. cm pV = 1072.5 N/cm-sec: liner
cm cm cm cm thickness = 6.0 mm.]
'I and all 2.5 1.5 8.4 6.0 2.5 1.2 Liner material Maximumhill 2.5 1.5 16.0 5.0 2.5 1 1 .6 temperature.
aBEcring i,.et I? ,halt to righ h% I 2 cn 0C
hBcaring i, .entra, hated on hat , Graphitized metal 108.5
PTFE-bronze 118.6Carbon graphite 147.0PTFE-thermoset 152.8PTFE-ceramic 152.7
TABLE II.-GEOMETRY MoZS-nyloii 152.8IResin 152.8VARIATIONS FOR Resin_152.8
CASE III
Bearing Bearing outerlength. radius.
cm cm TABLE V.-MAXIMUM INTERFACE
1.6 3.2 TEMPERATURES FOR VARIOUS HEAT
1.6 5.0 FLUX DISTRIBUTION FUNCTIONS
1.6 8.03.2 5.0 IResin liner thickness = 6.0 mm: pV = 1072.5
N/cm-sec.]
Heat flux distribution. Maximum Maximumq. temperature. temperature
N/cm-sec *C location.deg
q,,cos6 152.8 0q,,,(4/r)cos2O 179.7 0q,,(2/7r)(1 +sino) 169.7 +90q,*(2/7r) 114.0 0
q,• (2/70)( I - sinO) 99.6 0
TABLE Ill.-THERMAL PROPERTIES OF LINER MATERIALS
IFrom ref. 14.1
Liner Material Specific Thermal Density, Diffusivity,heat. conductivity, P. K/PC.
c, k. kg/m 3 m2/secJ/kg °C W/m *C TABLE VI.-MAXIMUM INTERFACE TEMPERATURES
FOR pV FACTORS AND LINER THICKNESSESGraphitized metal 377.0 127.4 7039.0 4.8 x 10-5
IHeat flux distribution q = q. cos (b]PTFE-bronze 394.0 42.5 6501.6 1.66 x 10-5
(b) Graphitized metal liner:Carbon graphite 754.0 3.35 2101.7 2.1 x 10-6 (a) Resin liner. 6.0 mm thick pV = 1072.5 N/cm-sec
PTFE-thermoset 942.8 .263 2293.2 1.1 X 10-6 p,,,V factor. Temperature, Liner Temperature.N/era-see *C thickness. *
PTFE-ceramic 838.0 .337 2376.0 1.69 x 10- 7 MTI1072.50 152.8
MoS-nylon 188.6 .23 1198.0 1.22 x 10-7 536.25 93.9 6.0 108.5107.25 46.8 12.0 103.5
Resin 1214.0 .233 1200.0 1.6 < 10- 7 18.0 99.8
6
![Page 8: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/8.jpg)
TABLE VII. -THERMAL PROPERTIES OF COATING TABLE X.-MAXIMUM INTERFACEMATERIALS AND SUBSTRATE TEMPERATURES FOR VARIOUS
BEARING DIMENSIONSMaterial Specific Thermal Density. Diffusivity
heat, conductivity, p. K/pC. ICoating. I-mm-thick TiN: heatc. k. kg/m 3 m2/sec flux distribution q = q,,cosO:
J/kg *C W/m *C pV = 1072.5 N/cm-sec.1
Aluminum 869.8 267.7 2 699.0 1.136x I0 5 Bearing and housing Maximum
outer radius, temperature.Zirconium 17 769.0 131.9 6513.0 1.142x10 6 cm C
Silicon carbide 711.0 104.4 2997.0 4.9x 10-5 3.8 182.05.0 202.1
Tungsten carbide 100.0 100.0 15 800.0 5.Ox 10-6 8.0 228.1
Aluminum oxide 1 039.0 20.73 3 997.0 4.99x 10-6
Stellite 421.9 9.7 8 300.0 2.77x 10 -6
TABLE XI.-MAXIMUMTitanium nitride 590.0 4.0 5 200.0 1.3 x 10-6 INTERFACE TEMPERATURES
FOR VARIOUS BEARINGZirconiaa 400.0 1.8 5 750.0 7.8x 10- 5 LENGTHS AND COOLING
CONDITIONSSteel 460.0 46.0 7 800 .0 1.28xI0 - 5
rJ l [Coating. I-mm-thick TiN:pV = 1072.5 N/cm-see: bearingouter radius. 5 cm.]
Bearing length. Maximumcm temperature..C
1.6 202.13.2 202.53.2 109.0
aWith shahi inner surface coo~ling.
TABLE VII.-MAXIMUM TABLE IX.-MAXIMUM
INTERFACE TEMPERATURES INTERFACEFOR COATING MATERIALS TEMPERATURES FOR
VARIOUS TITANIUM[Coating thickness = 1.0 mm: heat NITRIDE COATING TABLE XII.-MAXIMUM
flux distribution q = qcos4: THICKNESSES INTERFACE
pV = 1072.5 N/cm-sec.] TEMPERATURES FOR[Heat flux distribution VARIOUS pV FACTORS
Coating material Maximum q = q,,,cosO, pVtemperature. = 1072.5 N/cm-sec.[ [Coating. I-mm-thick TiN:
.C heat flux distribution
Coating Maximum q = q,,cos6.1Aluminum 176.8 thickness, temperature.Zirconium 181.1 mm C pV factor. MaximumSilicon carbide 182.2 N/cm-sec temnprature.Tungsten carbide 182.4 1.0 213.0 C
Aluminum oxide 191.3 .75 208.0Stellite 199.5 .5 201.0 1072.5 213.0Titanium nitride 213.0 .25 194.0 2681.25 48I.7Zirconiaa 234.7 .125 190.0 5362.5 928.0
a Prtally. , lh/ed
![Page 9: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/9.jpg)
Bush (3)
7r 2
I
167T1
fK~4jiji
xz
(b)
(m hII uh IC In %le"l~fh
FiLIure I h hII~h bc' rinv, trbho'\ ,Icm
![Page 10: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/10.jpg)
153
125
1978
90 0
82 1
74 -3
66 4
586
50 7
Y 42 9
35 0
z
figurte I riho~ ... i te ctiipcrialurc conr'l~or ca'e I IBearing had' 6 0) il;i hlitk o'It rt ttlef
![Page 11: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/11.jpg)
emperature,
nc
165
156U
147 C
/ >. \~ ~ -L ~138 D>( 129
120
iISG
102
93 4
844
75 4
664
57.4
y
V 485
x 395z
(ai Heat distribution function q t= q,(2 /r (nI miot.
F gure .1 -TriNho% stern temperature conrour% for case 1. Bearing hits f-mm-tick soft re-sin liner and \ariou, heat distribution tunctu,..
It0
![Page 12: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/12.jpg)
Temperature,
-C106
~.. 101
)< 95.6D
90.3
85.0
79.8G
74.5
69.2
64.0
58.7
53.4
y 48.2
42.9
x 37.6
(b) Heal distribution function q ,2r.
Figure 3.-Continued.
![Page 13: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/13.jpg)
Temperature,.r
97.4
93.1I
88.8 c
80.2
75 9
71 6G
67,3
63.0
58.7
"'< !X'54.4
k ~50..
V 45.8
41.5x
z 37.2
mC Heat~ distribution function q =q,,(2/7r) (I -in,
Figure 3.-Concluded.
12
![Page 14: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/14.jpg)
Temperatute.
213
202
190
178
166
154
142
130
118
106
945
82 6
70 7
588
46 9
z350
1Itnitc -4 ITI1~ \l1 Ictlpct.Iltirc o'ntom, l'r Lci~c 11 licotrim! i~i 1t I mmin thick I A\~.imm
![Page 15: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/15.jpg)
Tempeiature-C
201
190
168
157
146
135
124
113
102
905
79 4
683
572
46 1
z350
I- tiurc S Irib-N ' c\ IcTtc'Pe tlurc omfouir, fo~r :,iI Iic,11itiwi Iml' 0) 111111 011A~ I I mnw
![Page 16: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/16.jpg)
Temperature
190
179
169
159
148
138
128
117
107
968
865
76 2
55 6
45 3
z35 0
1 't 1~ Irrh '' 1Njjj Icrllperr1I ' 101t ! I'llur 'C 11 IICJF)' j I I 0 1 2i MIII III),k Ir IN 0111L
![Page 17: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/17.jpg)
'.5
25
2027
47 9
50
![Page 18: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/18.jpg)
Tcmpcraur&2
c
182
172
162
152
143
133
123
113
Ad 104
93 7
83 9
74 2
64 4
546
44 8
350
Ii .I L ~ i''\ lCm wiITTil~tTIIIT ''1 liII t''r Ill 1311I 11' Ild' 1 1111 1111k I I~ IN 1IIITITA. II'.Ii inc l cr .Ii' OH Sl l Lill
![Page 19: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/19.jpg)
?C3
'
,69
*58
79 7
68 5
5.'
![Page 20: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/20.jpg)
93
693
69 5
y 398
34 8
/x
1 0 11 'l I. ll T 11 11 t. 1 i , In 11'1KI
![Page 21: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/21.jpg)
Temperature.
929
869 f810
750
690
631
571
512
452
393
333
273
214
Y 154
94 6
zz 35 0
1Ilic I I I rill, ' I ,, I C 1kfI~ I* MJ)WI'J I, IJ ow 1 I1 im-rt!1aZ I ()nii' t I I \ TUTU tic ntN, aIrlT1 "6 N Ini ~ ' C '
![Page 22: The Effect of Coatings and Liners on Heat Transfer in a ... · and the Patran solid modeler software. The Patmar (Patran- method developed by Ling (ref. 2) to calculate the contact](https://reader034.vdocument.in/reader034/viewer/2022042123/5e9e0fd222bbaa25313a15e8/html5/thumbnails/22.jpg)
AReport Documentation PageSpace Adrmlnistration
1. Report No. NASA TM- 102513 2. Government Accession No. 3. Recipient's Catalog No.
AVSCOM TR 90-C-0064. Title and Subtitle 5. Report Date
The Effect of Coatings and Liners on Heat Transfer in August 1990a Dry Shaft-Bush Tribosystem 6. Performing Organization Code
7. Author(s) 8. Performing Organization Report No.
Mihir K. Ghosh and David E. Brewe E-5316
10. Work Unit No.
9. Performing Organization Name and Address 505-63-IA
NASA Lewis Research Center IL161102AH45Cleveland, Ohio 44135-3191and 11. Contract or Grant No.
Propulsion DirectorateU.S. Army Aviation Systems CommandCleveland, Ohio 44135-3191 13. Type of Report and Period Covered
12. Sponsoring Agency Name and Address Technical Memorandum
National Aeronautics and Space AdministrationWashington. D.C. 20546-0001 14. Sponsoring Agency Code
and
U.S. Army Aviation Systems CommandSt. Louis, Mo. 63120-1798
15. Supplementary Notes
Mihir K. Ghosh. National Research Council-NASA Research Associate at Lewis Research Center. on leavefrom Mechanical Engineering Department, 1.T.,B.H.U.. Varnasi, 221005, India: David E. Brewe. PropulsionDirectorate, U.S. Army Aviation Systems Command.
16. Abstract
The temperatures due to frictional heating within a solid lubricated or coated journal bearing were analyzed byusing a finite element method. A solid model of the shaft-bush tribocontact was generated with an eight-node,three-dimensional, first-order isoparametric heat-transfer element and the Patran solid modeler software. ThePatmar (Patran-Marc) translator was used to help develop the Marc-based finite element program for the system:this software was used on the Cray X-MP super computer to perform a finite element analysis of the contact.
The analysis was performed for various liner materials, for thin, hard, wear-resistant coated bearings, and fordifferent geometries and thermal cooling boundary conditions. The analyses indicated that thermal conductivity ofthe liner or coating material is the most vital thermal parameter that controls the interface temperature. Inaddition to design variations, the proximity of the cooling source to the heat-flux-generating interface is criticallyimportant to the temperature control in the system.
17. Key Words (Suggested by Author(s)) 18. Distribution Statement
Heat transfer: Journal bearing coatings: High Unclassified-Unlimitedtemperature: Dry lubrication: Finite elements: Subject Category 34Thermal distribution: Frictional heat
19. Security Classif. (of this report) 20. Security Classif. (of this page) 21. No. of pages 22. Price*
Unclassified Unclassified 23 A03
NASA FORM 1626 OCT 86 "Fr sale by the Nationni Technical Information Service, Springfield, Virginia 22161