reformer tube life evaluation

8
Archive of SID Archive of SID RESEARCH NOTE CREEP LIFE ASSESSMENT OF PRIMARY REFORMER HP40-NB MODI-FlED STEEL TUBE OF AN AMMONIA PLANT S. A. Jenabali. Jal1romi and M. NagltiKltalli Department oflvlaterials Science and Engineering, School of Engineering Shiraz University. Shiraz, Iran, [email protected] (Received: August 9, 2003 - Accepted in Revised Form: June 10,2004) Abstract Assessment of creep damage and residual creep life of a cast Ill' 40 'Nb Mod. reformer tube was perfon11cd. wherein the experimental Larson-Miller diagram and area fraction of creep voids were adopted. 'Ille state of damage of the tube in service was metallographically analyzed by using light and electron microscopy. Samples from the serviced reformer furnace tube were cut and prepared for void examination and creep test at 940°C-lOOO"C under 20-30 MPa stress. Microstructural examination was carried out with an Scanning electron microscope with secondary and backscattered electron detectors. Inter-granu lOll' voids. in the microstructure of the worked tube as a result of a creep phenomenon are ranked relating to the remaining life. Key Words Creep Life Assessmcnt. Ammonia Plant, 111'40Nb Modified Steel Tube "-:.J:, y."..4r.J:'.,y;, ~ t...:.,",l:.~ 1II' 40 ~ j1 .sI J) ,_Gl...}~ ~ .J ...,...:~ .)I.T--" J\i. ..:.-::1 .J~ .~ ~~y"';'f.slA .)>- ~ y--S .J },r :..~)'i ".;:..1)4jI ,;le..."! ~ SL,.;yi ...l>-IJ Y'A) .s.l.~I,,-:: '.J;> j! d.i.:.,.1 L J ..}I})\.:.. J J) L ::;.>- ...,;\... ',.i.>- cs:~j) .-: ,1d} ;I} ~)f; ~)Y )l:.>-L... j-:) )~ J Jf ~ -.slA ; ~ ,~"S )5 J) j! ),,&... :..-::1 ...,;';. .-: ,1..L:: r~1 ..;J".:..<]IJ is)~ '7:' J-C J,,-'=" L(.r.-,.,.::.;~ .~I~...:.;L_.,.(.>-.~'''' lIo.'f. .~\...,~.~ "';"'>-~\..."";""';._lo~S.,~ I~J\.:.. ~ -'.T-." ~c' ~"'-'./ L> " -,,-;-'T'~f-. .-'-::~"S ...5--:. ~ ...GL...}~ ,,~ "J..,..- y) '1 .,.100')) Jf :.:\...,.)e;::.\;,; j! d.e 1L ..G...t...:.::\...,.) JLC4 ..J-::~} d..:: 1?,,":. J ,,-:y~ is\... .)J".:..<]Ij1) ..f:-.;)) '-:;"'J-C).T'=" jI ~ ..I) )l:.>-L... j-:) ~ iSlf":. ..L:: ,-~;;, JJ) :..",1~...,...:,">'; .)1;.,..J ~\i. ...s'..I~ ":':.I~ ~ )l:.>-L..;-::)..,;.f.slA ,»- .)I.T--" 1. INTRODUCTION A major aspect of plant life management is estimating the remaining life of high-temperature components, which have a finite life due to creep. Methods based on post-service evaluations of the actual component material and direct estimation of the remaining creep life has gained popularity in the last decades. This is mainly because the results of such methods are expected to be more accurate in view of the fact that no assumption needs to be assumed regarding the materials properties or the past history of the component [1]. In this research the post - service evaluation of an HP-40 Nb modified refonner tube was done based on the accelerated, uniaxiai stress rupture testing of samples excised from the tube. In tubes, which are filled with supported nickel IJE Transactions B: Applications catalyst, methane reacts with steam, carbon dioxide and oxygen into the synthesis gas. Reform~rs are the heart of the fertilizer industry and any failure in this section of the plant results in premature shutdown leading to huge losses in terms of damage to the equipment, production losses and safety hazard [2]. 2. ALLOY DEVELOPMENT Since the catalyst tube assembly can amount to 25% of total cost of the furnace, there is a great incentive to optimize its design from chemical, thermal, and mechanical points of views. In the 80s, HP (25Cr/35Ni) modified alloys were developed by using certain metals, such as, Vo\. 17, No. 2, July 2004 - 183 Archive of SID www.SID.ir

Upload: subramanian-cholaperumal

Post on 31-Mar-2015

312 views

Category:

Documents


17 download

TRANSCRIPT

Page 1: reformer tube life evaluation

Archi

ve o

f SID

Archi

ve o

f SID

RESEARCH NOTE

CREEP LIFE ASSESSMENT OF PRIMARY REFORMER HP40-NBMODI-FlED STEEL TUBE OF AN AMMONIA PLANT

S. A. Jenabali. Jal1romi and M. NagltiKltalli

Department oflvlaterials Science and Engineering, School of EngineeringShiraz University. Shiraz, Iran, [email protected]

(Received: August 9, 2003 - Accepted in Revised Form: June 10,2004)

Abstract Assessment of creep damage and residual creep life of a cast Ill' 40 'Nb Mod. reformertube was perfon11cd. wherein the experimental Larson-Miller diagram and area fraction of creepvoids were adopted. 'Ille state of damage of the tube in service was metallographically analyzed byusing light and electron microscopy. Samples from the serviced reformer furnace tube were cut andprepared for void examination and creep test at 940°C-lOOO"C under 20-30 MPa stress.Microstructural examination was carried out with an Scanning electron microscope with secondaryand backscattered electron detectors. Inter-granu lOll'voids. in the microstructure of the worked tube asa result of a creep phenomenon are ranked relating to the remaining life.

Key Words Creep Life Assessmcnt. Ammonia Plant, 111'40Nb Modified Steel Tube

"-:.J:,y."..4r.J:'.,y;,~ t...:.,",l:.~ 1II' 40 ~ j1 .sI J) ,_Gl...}~~ .J ...,...:~ .)I.T--"J\i. ..:.-::1.J~ .~~~y"';'f.slA .)>- ~ y--S .J },r :..~)'i ".;:..1)4jI ,;le..."! ~SL,.;yi ...l>-IJY'A) .s.l.~I,,-::'.J;>

j! d.i.:.,.1 L J ..}I})\.:.. J J) L ::;.>- ...,;\...',.i.>-cs:~j) .-: ,1d} ;I} ~)f; ~)Y )l:.>-L... j-:) )~

J J f ~ -.slA ;~ ,~"S )5 J) j! ),,&... :..-::1...,;';. .-: ,1..L:: r~1 ..; J".:..<]IJ is)~ '7:'J-C J,,-'="L(.r.-,.,.::.;~ .~I~...:.;L_.,.(.>-.~'''' lIo.'f. .~\...,~.~ "';"'>-~\..."";""';._lo~S.,~ I~J\.:..~ -'.T-." ~c' ~"'-'./ L> " -,,-;-'T'~f-.

.-'-::~"S ...5--:. ~ ...GL...}~ ,,~ "J..,..- y) '1 .,.100')) J f :.:\...,.)e;::.\;,;j! d.e 1L ..G...t...:.::\...,.) JLC4

..J-::~} d..:: 1?,,":. J ,,-:y~ is\... .)J".:..<]Ij1) ..f:-.;)) '-:;"'J-C).T'="jI ~ ..I) )l:.>-L... j-:) ~ iSlf":.

..L:: ,-~;;, JJ) :..",1~ ...,...:,">';.)1;.,..J ~\i. ...s'..I~ ":':.I~ ~ )l:.>-L..;-::)..,;.f.slA ,»- .)I.T--"

1. INTRODUCTION

A major aspect of plant life management isestimating the remaining life of high-temperaturecomponents, which have a finite life due to creep.Methods based on post-service evaluations of theactual component material and direct estimation ofthe remaining creep life has gained popularity inthe last decades. This is mainly because the resultsof such methods are expected to be more accuratein view of the fact that no assumption needs to beassumed regarding the materials properties or thepast history of the component [1]. In this researchthe post - service evaluation of an HP-40 Nbmodified refonner tube was done based on theaccelerated, uniaxiai stress rupture testing ofsamples excised from the tube.

In tubes, which are filled with supported nickel

IJE Transactions B: Applications

catalyst, methane reacts with steam, carbon dioxideand oxygen into the synthesis gas. Reform~rs arethe heart of the fertilizer industry and any failure inthis section of the plant results in prematureshutdown leading to huge losses in terms ofdamage to the equipment, production losses andsafety hazard [2].

2. ALLOY DEVELOPMENT

Since the catalyst tube assembly can amount to25% of total cost of the furnace, there is a greatincentive to optimize its design from chemical,thermal, and mechanical points of views. In the80s, HP (25Cr/35Ni) modified alloys weredeveloped by using certain metals, such as,

Vo\. 17, No. 2, July 2004 - 183

Archive of SID

www.SID.ir

Page 2: reformer tube life evaluation

Archi

ve o

f SID

Archi

ve o

f SID

TABLE 1. Designation and Composition of HP Alloy 131.

TABLE 2. Charactedstic of As Cast HP Alloy 141.

* For centrifugally cast pipes 8 and for static castings 6

1.::1" ,HP-4D Mod Nb

. .. . '" . . HK-40 AlloyOne halftimesnormal, 2

Normal~:.:sIv

oD

~O~iv

.:::<;; IHiv0::

(j .4

Two timesnormal

-------- Normal+28°C

02 Normal

{j, O.2!:. O.!)O (j 7::

Wall Thichness, in

1.0

Figure 1. Shows the relation between w thickness and amount cycli that is an important consideration tube life [5J.

molybdenum, niobium, tungsten and titanium.Designation and composition of the HP alloy areshown in Tables 1 and 2 [1,2,3].

The requirements of a cost effective reformerdesign are maximum reliability, operating stability

and high thermal efficiency. The materials usedmust have a high creep strength confirmed withgood strain relation, good weld ability, andexcellent oxidation resistance and, after aging,good ductility and good weld ability [5].

184 - Vol. 17, No. 2, July 2004 IJ£ Transactions B: Applications

1

UNS ASTM Composition wt%ACI designation number specification %C %Cr %Ni %Si %Mn

HP N08705 A297 0.35-0.75 24-28 33-37 2-2.5 2

Melting DensityCoefficient of Therma 1 0.2% proof UTS

point kg/dm3expansIOn conductivity £1% stress

mm/mm/QC W/moC (MNm'2) MNm-21350°C 8.02 lO'6x18.5 At 1050°C 30 8, 6" 250 450

Archive of SID

www.SID.ir

Page 3: reformer tube life evaluation

Archi

ve o

f SID

Archi

ve o

f SID

~ ~ ~ ~ -- ~ ~ - - --- --- ~ .. = ~-

1000000Cl...en::;0

.£:::0

8ci0or-

EE-z

£CD::;i5.. 10002.8U)enCD

~EJE.c~

25 Cr/35 Ni + NbHP-Mod

~27.2 ~

0.£:::

8ci

13.60 .-.!:~~

6.8 C.2.8<n<n

"3.4 ~. 00500

200

E::3E.c

1.36 ~°C

Figure 2. Minimum creep rupture stress versus temperature for a period of 100,000 h [5].

4

.or

HP (25Cr-35Ni-1.8Si-4.2W-1.9Co)

000 11 .2 3

Distance from bore surface, mm

...

Figure 3. Carbon concentration profiles of IlK and liP alloys exposed at lor 1200h in 3CII~ - 40lhO - 37N~L5].

In retrofit or revamp design, the HP-40 NbMod. allows two possibilities. First, with theoutsidediameter fixed because of existinginlet and

outlet manifolds, make the tubes thinner. Thisimproves the rate of heat transfer and increases30% to 40% production capacity. With the thinner

IJE Transactions B: Applications V01. 17, No. 2, July 2004 - 185

s

E.....=

I (25Cr-35Ni-1.35Si-1.33Nb)

0U....CS

oD.....

U1

Archive of SID

www.SID.ir

Page 4: reformer tube life evaluation

Archi

ve o

f SID

Archi

ve o

f SIDFigure 4. A view of exchange of the damaged tubes and the

catalyst.

wall, the cost per foot of catalyst tubing iscomparable, and in some cases less than HK-40,even though the cost of HP-40 Nb Mod. alloytubes seem to be more because of their perceivedexpense [5]. The second option is to keep the sametube diameter and thickness and raise the designtube metal temperature by 60°C [5].

The HP-Mod. alloys are said [5] to copesubstantially better with thermal stresses. Figure Ishows the effect of wall thickness on the life ofHP-40 Mod. and HK-40. It also shows thereduction or increase in life due to high or lownumber of thermal cycles and the reduction in lifecaused by operating 28°C above normal operatingtemperature. Note, that for the same wall thickness,the relative life of HP-40 Nb Mod. is about three

times that of HK-40, while if the alloys aredesigned for the strength, only twice the life can beexpected. In addition, the good ductility andweldability of HP-40 Nb Mod. deservementioning. It has, in the as-cast condition, about8% elongation that after prolonged service drops toabout 4%.

f field repairs are required on HP-40 Nb Mod. asolution anneal for about two hours isrecommended, which will restore much of theelongation [5].

Figure 2 shows minimum creep rupture stressversus temperature for a period of 100,000 h [5].

It has been found that modification of HK-40

186 - Vol. 17, No. 2, July 2004

alloys by Nb improves resistance to creep ruptureand carburization. Figure 3 illustrates thecarburization resistance of some of these modifiedalloys compared to HK-40 [6].

A more recent development is the HP micro-alloys, which traces of titanium, zirconium andrare earths are added to the alloys during casting.The micro-alloys enhanced carburization resistanceand improved high- temperature creep-ruptureresistance [3].

3. MATERIALS AND EXPERIMENTS

Due to carbon formation and catalyst break up in aHP40 Nb modified tube in Razi PetrochemicalComplex [7], it is deduced that the consume of theheat flux was locally stopped. Therefore, localoverheating occurs which caused the failure of thetube after 12350 hr of its operation. Due to thisoverheating the tube and all of the catalyst werereplaced, see Figure 4.

Many test samples were prepared for creep lifeassessment and metallographic analysis from aregion far from the failed part of the tube (fourmeter). The creep rupture test was preformedaccording to ASTM-E 139-83. The results areshown in Table 3.

For micro-structural evaluation, standardmetallographic preparation techniques were used.The metallographic specimens were etched inglyceregia (40 parts glycerol, 40 part hydrochloricacid and 20 part nitric acid). The microstructurewas examined using optical and scanning electronlTIlCrOscopy.

The phases observed were analyzed forchemical composition by using an energydispersive x-ray analyzer system (EDX) inconjunction with a SEM.

The damage assessment of reformer furnacetube is based on Larson Miller parameter (LMP)and on the metallographic analysis of a servicedindividual tube. Although this tube is not the realrepresentative of the complete furnace tubes butit is important note that destroying a largenumber of the tubes is not practical, in additionthis tube is the most damaged one. Therefore theextent of damage is the highest and its remaininglife is the lowest.

IJE Transactions B: Applications

...

Archive of SID

www.SID.ir

Page 5: reformer tube life evaluation

Archi

ve o

f SID

Archi

ve o

f SID

TABLE 3. The Results of Creep Rupture Tests Under Different Stresses and Temperatnres.

35

30

23.607

25

~ 20QJ

~ 15

10

5

0

23 24

. 24.294

25

LMP

25.559 26 27

Figure 5. Rupture stress versus Larson and Miller parameter (master curve).

3.1 Damage Assessment The most importantmechanism for damage and life extinction of thetubes is creep. The extent of the damage can becalculated by using temperature compensating timeparameters or by evaluating the range of micro-structural deterioration [2].

3.2 Larson- Miller Parameter The standard

method used to interpret creep stress -to- rupturedata is the parametric expressions such as the onewhich developed by Larson and Miller [8-10] thatis defined by the following equation:

TLMP =- (logt + C)

1000

where; T, is the service temperature (K), t, time to

IJE Transactions B: Applications

rupture (hr), and C is a constant.To find the maximum value of the constant C in

the Larson - Miller equation, or in other words, toassess the minimum remaining life of the tubes thefollowing data are used:

LMP = T (log t + C) W-3 at a constant stress

LMP=constant

LMP(l) = (938 + 273)(logl08.3 + C)

LMP(2) = (995 + 273)(log14.4 + C)

At stress 30Mpa LMP( 1) = LMP(2)

(938 + 273)(log108.3 + C) =

(995 + 273)(log14.4+ C) ---> C =17.459

Vo!. 17, No. 2, July 2004 - 187

Specimen No. Temperature (°C) Stress (MPa) Rupture time (hr)

1 938 30 108.3

2 995 30 14.4

3 1000 30 42.1

4 1041 20 25.2

5 1016 20 71.3

Archive of SID

www.SID.ir

Page 6: reformer tube life evaluation

Archi

ve o

f SID

Archi

ve o

f SID(a)

01mm"I ;0

- , "

.. !~

, 5,'" f' 11"

't

(b)

't

,4

',C<'

(c)

Figure6. (a)SEMmicrographshowsmassiveprimarycarbidesinanaustenitic matrix and tine secondarycarbideswithin the austenite grain upon exposure to elevated temperature, (b) SEM micrograph shows the damage part of thetube which due to high temperature, secondary carbides were reduced and the inter-dendritic carbides had undergone

signiticant agglomeration and coarsening and (c) light micrograph shows random creep voids.

The stress state in the material is complex, but itis clear that stress due to intemal pressure plays aleading role in damage accumulation. As theinternal pressure varies little along the tubes,damage concentrates in the hotter section [11,12].Therefore, in the present work the average stress inthe tubes or the effective stress was used forcalculation.

188 - Vol. 17, No. 2, July 2004

The following data and equations were used forfinding the effective stress in the tubes, [13].

rj= 41 mm, 1'0= 54 mm,

Pi = 33.35 bar = 3.34 MPa

(5r= -3.34, MPa = (53

IJE Transactions B: Applications

j

Archive of SID

www.SID.ir

Page 7: reformer tube life evaluation

Archi

ve o

f SID

Archi

ve o

f SID

crosssectionof tlIC tube

75%

damagelevel

113 1/2 213

Figure 7. Classitication of the damagein a reformer furnacetub~, as indicatedafter metallographic preparation [15].

aH=12.43 MPa = a 2

a = 4.5MPa= a i ' by using VonMises equation;(IX

I

- I

[

(a I -(2)2 +(al -(3)2 +

1

2

a=-I,.J2 2 '

(a2 -(3)

iY = 13.4MPa

By using data in Table 3 and the highest valueof the constant C according to the creep data, the(LMP)s values were determined for variousstresses and temperatures as below;

LMP(l) = (930 + 273)(log108.3 + 17.459)W3 =23.607

LMP(2) = (995 + 273)(logI4.4 + 17.459) 10-3=23.606

LMP(3) = (1000 + 273)(log42.l + 17.459) 10-3=24.294 .

LMP(4) = (1041 + 273)(log25.2 + 17.459) 10-3=24.481LMP(5) = (1016 + 273)(log71.3 + 17.459) 10-3=24.893

Using the results of the creep rupture tests(Table 3), the plot of the stress versus Larson

IJE Transactions B: Applications

Miller parameter (master curve) was drawn inFigure 5.

By extrapolating the master curve to thecalculated effective stress experienced in the tubes(13.4 MPa), the Larson-Miller parameter at theeffective tube stress was deduced to be 25.56. Byusing this value of (LMP = 25.56), the remaininglife of the tubes at the service temperature (870°C)was calculated as follow.

From Figure 5, the line equation is:

Stress = -8. 505 LMP + 230.88

At stress 13.4 MPa LMP = 25.559

LMP = T (log t + C) 10-3

25.559 = (870 + 273)(log t + 17.459) 10-3

t = 79860 hr = 9.2 year

3.3 Micro-Structural Deterioration Creepdamage starts in the wall as round voids randomlydistributed on dendritic boundaries figure 6 Theirpreferred fonnation is on boundaries perpendicularto the tensile stress.

The structure consists of massive primarycarbides in an austenitic matrix (Figure 6a) inaddition; fine secondary carbides were precipitatedwithin the austenite grain upon exposure toelevated temperatures. In serviced materialremoved from the damaged part of the tube (Figure6b), temperature was high and the number of

Vol. 17, No. 2, July 2004 - 189

Archive of SID

www.SID.ir

Page 8: reformer tube life evaluation

Archi

ve o

f SID

Archi

ve o

f SID

secondary carbides was reduced and the inter-dendritic carbides had undergone a significantagglomeration and coarsening. EDX analysis ofthese inter-granular carbides shows that they arechromium and niobium rich carbides [6].

Figure 6c shows the result of the lightmetallurgical survey along the tube illustratingrandom distribution of voids.

With regard to references [11 and 14] and byfive level damage characterization-approach of1. Le May [15] (Figure 7), in which damage wasclassified as level A or having no detectablevoids, level B as displaying isolated cavities,level C having oriented cavities, level D havingmicro-cracks and level E having macro-cracks,the damage was revealed to be in the cnd rangeof level H. This means the approximateremained life is something around seventypercent of the designed life (which is generally100,000 hr).

4. CONCLUSIONS

1. The remaining life of primary reformer tubeswas predicted to be about 9 years. This wasdone by utilization stress and temperatureassisted acceleration that involves the use of

Larson - Miller parameter.Calculations based on the stress-rupture testresults showed that the Larson-MillerConstant (C) for 12350 hI' serviced HP-Nbmodified reformer steel tube was about25.56.

On the basis of the metallographic, damagethat have been obseryed in the microstructureof the serviced tube and its comparison withthe classification of the damage in thereformer tubes it was indicated that around

seventy percent of the designed lifc (-70,000hr = 8.1 years)) was remained.

2.

".).

5. ACKNOWLEDGEMENTS

The authors wish to acknowledge with gratitudethe supply of materials and assistance provided by

190 - Vo!. 17, No. 2, July 2004

Razi Petrochemical Corporation.

6. REFERENCES

I. Viswanathan, R. and Foulds, J., "Accelerated StressRupture Testing For creep Lite Prediction - Its Value and

Limitation", Journal of PresSllre Vessel Technology,Vol. 120, (May] 998), 105-]] 5.

2. !3haumik, S. K., Rangaraju, R. and Parameswara, M. A.,"Failure of Reformer Tube of an Ammonia Plant",Engineering Failure Analysis, 9, (2002), 553-561.

3. !3lair and Stevens, T. L. (Eds.), "Heat-Resistant HighAlloy Steels, Steel Castings lIandbook", 61hEd., St;elfounders Society of America and ASM International,( 1995), 22( I )-22( 13).

4. I'aral!oy, 1139WM, "lligh Strength, High TemperatureAlloy t':Jr Steam Cracking Furnaces and Stemn ReformerFurnaces", Paralloy House, Nuttield Raod, !3illingham,Cleveland. England TS23 4DA.

5. Schillmoller, C. M.. "Ilp-Mod.ified Furnace Tube forSteam Reforms and Steam Crackers", NiDI TechnicalSeries, No. 10058.

6. ASM Handbook, "lleat-Resistant Materials", MaterialsPark, ASM International, (1999).

7. Jahromi,S. A. J. and NaghiKhani,M., " FailureAnalysisof HP40 -Nb Modified Primary Reformer Tube ofAmmonia Plant", Iranian Journal of Science andTecllllology, Transactions B, Vu!. 28, Bo. B2, (April2004),269-271.

8. Viswanathan, R., "Damage Mechanisms and LifeAssessment of High-Temperature Components", MetalsPark, Ohio 44073, ASM, ( 1995).

9. ASM, Ilandbook, "Heat Resistant Materials", MetalsPark, Ohio 44073, ASM, International, (1999).

10. Komei Kasahara, "A Method t()r Estimating ResidualCreep Lives of Cast HK-40 Reformer Tubings", TokyoGas COll/pany Ltd., R&D InMitufe, 1-16 Shihaura,Tokyo, Japan C 235/89, C Mech. E., (1980),249-264.

11. Da Silveira, T. L. and Lemag, I., "Damage Assessmentand Management in Reformer Furnaces" Joul'I/al (~rPressure Vessel Technology Now, Vol. 119, (1997),423-427.

12. Gong Jian Shan- Tung Tu and Kee !3ong Yoon, "DamageAssessment and Maintenance Strategy of IIydrogenRet(mner Furnace Tubes", Engineering FailureAnalysis, 6 (1999), 143-153.

13. Goerge ElIwood Dieter, "Mechanical Metallurgy", 3Ed..New York, MagGraw Hill Company, (1986).

14. C. W, Thomas, Stevens, K. J. and Ryan, M. J."Microstructure and Properties of Alloy 11P50NbComparison of As Cast and Service Exposed Materials",Materials Science and Technology, Vol. 12, (1996), 469-475.

15. May, I, Le., Da Silveira, T. L. and Vianna, C. 11.,"Criteria tor the Evaluation of Damage and Remains Lifein Reformer Furnace Tubes", lilt. J. Pres. Ves, 8 Piping66, (1996),233-241.

IJE Transactions 13:Applications

Archive of SID

www.SID.ir