.22-each stress level, rate of load application and removal, and actualexpected stress-frequency spectra determines the actual fatigue behaviorof materials and <strong>structures</strong>.The test data produces very wide scatter bands. Much of the scatteris typical <strong>for</strong> fatigue tests, and much is due to the many variables previouslydiscussed. In cases where the data was considerably higher orlower than the general trends, it was usually possible to at+mibute thisto variations in test procedure, specimen preparation, etc., in which casesthe data was not included. Statistical evaluation of the data is requiredin order to develop <strong>design</strong> fatigue curves <strong>for</strong> <strong>aluminum</strong> alloys. This typeof analysis is very important, since it appears that the scatter of <strong>aluminum</strong>fatigue data is greater than with steel, which could affect the selection ofsafety factors.It now becomes necessary to reduce the data shown on Figures 3 through9 to a set of <strong>design</strong> fatigue (S-N) curves <strong>for</strong> the various <strong>aluminum</strong> alloysunder consideration, which will be suitable <strong>for</strong> use in comparing the <strong>hull</strong>structure of a large <strong>aluminum</strong> <strong>hull</strong> with that of an equivalent mild steel<strong>hull</strong>. This process involves reduction of the variables presented inFigures 3 through 9 to obtain a single curve <strong>for</strong> each alloy, <strong>for</strong> comparisonto an equivalent steel S-N curve. Figure 10 contains such <strong>design</strong>curves, which are based upon the welded strength with bead on, using theaverage of R = o (zero to maximum stress) and R =-1 (complete reversa~)~and disregarding notch effects and salt water spray. The rationale <strong>for</strong>this approach follows.The choice between welded and unwelded values is fairly straight<strong>for</strong>ward,since the lower welded strengths would govern the <strong>design</strong> of atypical ship structure <strong>for</strong> both cyclic and short-term loading. This approachis somewhat conservative, in that the reduction in fatigue strength<strong>for</strong> <strong>aluminum</strong> due to welding is proportionally greater than that <strong>for</strong> steel.The fatigue strength of both <strong>aluminum</strong> and steel is improved by removingthe weld bead of full penetration butt welds. However, this representsan idealized condition which can not be economically achievedin ship construction. Cold working of fillet welds by peening willincrease their fatigue strength, but again this represents an unreasonablefabrication requirement. There<strong>for</strong>e it must be assumed that ‘tbeadonJlvalues are more appropriate <strong>for</strong> typical ship <strong>structures</strong>.For the idealized shipfs <strong>hull</strong> girder bending on a trochoidal wave, itwould be expected that fully reversed cyclic stress values (R = -1) wouldapply. However, as shown later, the actual life cycle-stress hi~togram ofa bulk carrier lies between the cases of R = -1 ad R = O (from zero stressto maximum tension or compression) because of the effects of the relativelyhigh still water bending moment. Similarly, local <strong>structures</strong> seldomexperience fully reversed stresses due to various combined loading conditions;i.e., bending plus compression or tension. Pending a more completeevaluation of this problem, it is proposed to use the average of thevalues <strong>for</strong> R = O and R = -1.The quantity of data on the effects of notches of various types onfatigue strength is far too limited to derive general <strong>design</strong> curves atthis time. In addition, it is not possible to relate the stress concentrationfactors prevalent in t~ical ship <strong>structures</strong> to the test data now‘available. The use of bead-on data reflects the notch problem; thus itis proposed to neglect additioml stress concentration effects.
-23-As noted previously, salt spray significantly reduces the fatiguestrength of steel and, to a greater degree, <strong>aluminum</strong>. Howe~er~ thiseffect is not being considered in this study <strong>for</strong> several reasons. First,the highly-stressed portions of the <strong>hull</strong> girder would be subjected to directsalt spray during relatively small percentage of their operating life. Thebottom, <strong>for</strong> example, is totally immersed, while the deck wofid experiencespray in the highly stressed midship portion only a small percentage ofthe time. Current salt spray fatigue data is based upon continuousexposure, and it is probable that the effects of salt sPraY are exponential~For a given reduction in exposure time, the reduction in strength degradationwould be far less. Secondly, the relative depth of surface pittingand loss of thickness of thin test samples <strong>for</strong> a given period of exposurewould be far greater than <strong>for</strong> the thick plates of a b~k carrier h~l~which may reduce the net section loss in area. In conclusion, itdoes not appear that the salt spray data in Figure 9 is applicable totypical ship <strong>structures</strong> in a normal life-cycle sea environment. However,it is not intended to minimize the problem. As shown in Figure 9, asufficient concentration of salt spray can effectively destroy the stresscarryingcapabilities of <strong>aluminum</strong> alloys at a large number of cycles.Thus this problem warrants considerable future attention.Figure 10 indicates that the S-N curves of the various <strong>aluminum</strong> alloyshave approximately the same shape, with initial strength corresponding tothe bead-on values of welded ultimate tensile strength of Table 4 reducingto between 6 and 9 ILSIat 108 cycles. Based upon the curves of Figure 10,the gross area under the S-N curves of the <strong>aluminum</strong> alloys relative tothat of mild steel are as follows:6050l,-!;!,,30 ~ ,Ii ...—- -’*-I10II (,,, !1 1 ,,, 11,! 1!’ 1 >,, 1 - ,—–L. .––.1! ‘- - .. . .. ..1’ II.. L-..—L. ...-.1 11, ‘. I l.. 1...-....11 .’..1 ...$. .. ..OJ102II ,, J; I 1,,,, ,,,, ) 1’ 1 ,1! I Ill, 11,.3’ :IOLI ‘ ‘ 105 ‘,“ .~06‘ 10’ 108L ‘,NUhBFEOF CYCLES(N)FIG. 10 Recommended S-N Fatigue Curves <strong>for</strong> Welded 5000 SeriesAluminum Alloys and Mild Steel <strong>for</strong>Design of <strong>Ship</strong> <strong>Structure</strong>
- Page 5: CONTENTSI.. II.III.Iv.v.VI ●VII.I
- Page 9 and 10: LIST OF FIGURES(Cent’d)FIGURE NO.
- Page 11 and 12: I. INTRODUCTIONThis report summariz
- Page 13: art in fabricating and maintaining
- Page 16 and 17: MONTEROSSO GRANA /17VALGRANA / CARA
- Page 18 and 19: -8-Numerous references have been re
- Page 20 and 21: .10.TABLE 2. Mechanical Properties
- Page 22 and 23: TABLE 2 Mechanical Properties of Al
- Page 24 and 25: TABLE 3 Mechanical Property Limits
- Page 26 and 27: -16-l?igures5, 6, 7 ati 8 present f
- Page 28 and 29: -18-ti-’”’-”-””””-L
- Page 30 and 31: -20-60 .r---.— ..,.— -——,L-
- Page 34 and 35: -24-!Z456-H321 = 0.485083-H321 = 0.
- Page 36 and 37: -26-(c)Members with partial or cont
- Page 38 and 39: -28-AllOyS 5083 and 54.56(~ content
- Page 40 and 41: -30-The previous paragraphs have de
- Page 42 and 43: -32-The problem of cargo hold abras
- Page 44 and 45: -34-The question of residual stress
- Page 46 and 47: .36-Each alloy was given a relative
- Page 48 and 49: -38-GENERAL OBSERVATIONSFYior to a
- Page 50 and 51: -40-The question of comparative imp
- Page 52 and 53: -42-(d)(e)Poor quality welds due to
- Page 54 and 55: -44-The ABS criteria noted above we
- Page 56 and 57: -46-DNV would consider fatigue in e
- Page 58 and 59: -48-is less, for the exposed side s
- Page 60 and 61: Equation (2):-50-Hu1l SMa~um = Hull
- Page 62 and 63: -52-Another aspect of vibrations wh
- Page 64 and 65: -54-000000000Bottom Shell PlateSide
- Page 66 and 67: -56-at the deck and keel. This stre
- Page 68 and 69: -58-AT is the change inUT= Thermal
- Page 70 and 71: -60-SUl@!ARYAll parties contacted f
- Page 72 and 73: -62-(c)(d)(e)(f)T~e exterior side o
- Page 74 and 75: TABLE 12 Aluminum Bulk Carrier - Su
- Page 76 and 77: .66-INSUT.ATION AND SHEATHINGShell8
- Page 78 and 79: -68-(b)(c)(d)(e)(f)(g)(h)(i)(j)At l
- Page 80 and 81: -70-IIF.INSTALLATION OF SYSTEMS AND
- Page 82 and 83:
Rudder Assembly -carrier should be
- Page 84 and 85:
-74-(b)MechanicalTensile Strength 6
- Page 86 and 87:
-76-(e)The steel piping must be of
- Page 88 and 89:
-78-Other Piping Systems and Valves
- Page 90 and 91:
-80-struetion for the aluminum hull
- Page 92 and 93:
-82-Large heavy type machine~ must
- Page 94 and 95:
suffers attack in an alkaline envir
- Page 96 and 97:
-86-REPAIRSObtaining proper repairs
- Page 98 and 99:
-88-The design of the midship s~cti
- Page 100 and 101:
-90-assuming the increase is applic
- Page 102 and 103:
LIGHT SHIP WEIGHT ESTIMATE-92-In or
- Page 104 and 105:
-94-TABLE 20 Aluminum Bulk Carrier
- Page 106 and 107:
TABLE 22 Trim and StabilityFull Loa
- Page 108 and 109:
-98-TABLE 24 Price of Steel Bulk Ca
- Page 110 and 111:
GaseNumber. . . -.,- .TABLE 27 Comp
- Page 112 and 113:
-1o2-TABLE 28CarriersComparison of
- Page 114 and 115:
12 ---n..T.[T7%l,=LEGS IU ORF=ErY
- Page 116 and 117:
-106-such as iron ore, on two of th
- Page 118 and 119:
-108-7)is,zg~ gg~5e mzz~E’4E!~K2j
- Page 120 and 121:
-11o-(a)(b)(c)(d)Inerting system fo
- Page 122 and 123:
-112-fatigue, particularly in the p
- Page 124 and 125:
-114-2k* Installation of Systems an
- Page 126 and 127:
-116-LIST OF REFERENCES(7)Fatigue P
- Page 128 and 129:
-11.8-LLST OF REFERENCES(Cent’d)(
- Page 130 and 131:
-120-ADDITIONAL SOURCES OF INFORMAT
- Page 132 and 133:
-122-redistribution of the still wa
- Page 134 and 135:
-124-APPENDIX BEXCERPTS FROMRULES A
- Page 136 and 137:
-126-92.07-10(d)(~) Interior stairs
- Page 138 and 139:
-128-~gE1+0102030- .. ..—405060
- Page 140 and 141:
ectintyclassification4KEYWORDSROLEL
- Page 142:
SHIP STRUCTURE COMMITTEE PUBLICATIO