cranfield university mahadi abd murad an integrated structural ...

More documents

Recommendations

Info

influenced by biaxial effects in the structural response. Overall, the main findings of the Chapter 3 can be concluded as; Any increasing pressure in the elastic working region has no influence on the increment of stress concentration values. The longer the arc defect length, the higher the hoop stress concentration will be. Poisson‘s ratio has a lesser influence on stress concentration as the length of the arc shaped defect is increased. The hoop stress is directly proportional to the length of the defect and inversely proportional to the thickness of the repair. However, axial stress is inversely proportional to both defect length and repair thickness. Stress concentration factor is not the function of the composite repair length neither thickness but the length of the defect instead. However, from the simulation works, it is clear that the stress concentration factor at the defect in the pipe with the composite repair is lower than at the defect without the repair. The optimum length of composite repair for the defect width of 40 mm is 160 mm as compared against the recommended length given by the manufacturer which is 240 mm. With this PRI study, nearly 20% of stress loading has been transferred in the circumferential direction and nearly 38% in the longitudinal direction. As discussed in the earlier chapters, the FEA results that have been presented in Chapter 3 have to be validated experimentally. In order to do that, the experiment itself had to be carried out based on proper guidelines and standards so that the experimental results are correct and anticipate fewer errors. The selection and installation of sensors must be correct so that data acquisition can produce sensible results. The technique of installing the composite is also crucial because often the pipeline repairs are of the wet type rather than the cured type. In other words, the installation work usually relies on the skill of the applicator in ensuring the effectiveness of the fibre reinforcement after it 194
has been fully cured. Poor surface preparation and failing to follow the manufacturers‘ guidelines may result in failures such as delamination etc. All the above requirements have been fulfilled using the right methodology. The important findings of Chapter 4 can be summarized as: Yielding as a demarcation point in the study of bonding integrity between the reinforced material (steel substrate) and reinforcing material (composite) by previous researchers is arguable. Most of the repaired pipelines do not yield and there is a lack of analytical information about how well load is being transferred from the defect steel into the composite material in the elastic working region. Therefore, Chapter 4 has proved that it is possible to carry out this study provided all the above requirements have been followed correctly. The technique of installing strain gauges at the arc-shaped defect area is important because different orientation has produced different result. From this study, it is recommended to adopt the orientation technique that is similar to that of S3 since this technique has produced the most steady reduction of strain readings. About 14.25% load is transferred in the hoop and 23.5% in axial directions. All the manual calculation, numerical and experimental hoop stresses and strains and axial stresses show a very good agreement at the defect and nominal areas. However, axial strain correlation is much less apparent. This is attributed to mass and pressure loading differences between numerical and experimental system and lack of better complex analytical expressions in the manual calculation. In terms of the load transfer in the composite repair, the highest hoop load transfer between the composite layers occurs at the defect area but for the axial load transfer, it occurs at the nominal area. The error of 4.17% between the numerical and experimental hoop strains at the defect area could have been further improved had the accurate elastic constant values been obtained. This 195
Page 1:
CRANFIELD UNIVERSITY MAHADI ABD MUR
Page 4 and 5:
ABSTRACT One of the most common pro
Page 7 and 8:
TABLE OF CONTENTS ABSTRACT ........
Page 9 and 10:
3.7.1 Introduction ................
Page 11:
5.6 Summary and Conclusions .......
Page 14 and 15:
Figure 3.1: Schematic diagram of th
Page 16 and 17:
Figure 5.4: The average temperature
Page 18 and 19:
NOMENCLATURES ACFM Alternating Curr
Page 21 and 22:
1 INTRODUCTION One of the most comm
Page 23 and 24:
1.1 Objectives Knowing this backgro
Page 25 and 26:
The methodology of this integrated
Page 27 and 28:
Chapter 6 contains the summary and
Page 29 and 30:
period of the major part of the exi
Page 31 and 32:
Figure 2.2: Reported interest in de
Page 33 and 34:
Figure 2.4: Dent damage caused by v
Page 35 and 36:
presented current practices, recent
Page 37 and 38:
Energy Workforce Sdn Bhd (2011) cla
Page 39 and 40:
available provides real time detect
Page 41 and 42:
2.3.3 Reliability Analysis of Infor
Page 43 and 44:
Figure 2.9: Pipeline failure probab
Page 45 and 46:
2.3.3.1 NDT Reliability and POD cur
Page 47 and 48:
increase in the yield limit of stee
Page 49 and 50:
the SCC assessment in their integri
Page 51 and 52:
Manufacturers such as Walkers Techn
Page 53 and 54:
Company L.P, 2005); Aquawrap ® pro
Page 55 and 56:
monitoring, inspection, and damage
Page 57 and 58:
2.6.2 Aims of Structural Health Mon
Page 59 and 60:
Figure 2.16: Application fields and
Page 61 and 62:
egistered by the piezoelectric sens
Page 63 and 64:
discretization error. Round-off err
Page 65 and 66:
notches and Neuber‘s is better fo
Page 67 and 68:
maximum local stress at discontinui
Page 69 and 70:
De Carvalho (2005) concludes that t
Page 71 and 72:
accurate simulation of loading espe
Page 73 and 74:
Based on the above facts, the opera
Page 75 and 76:
3 FINITE ELEMENT ANALYSIS (FEA) IN
Page 77 and 78:
Figure 3.2: Schematic diagram of th
Page 79 and 80:
In this modelling work (i.e. compos
Page 81 and 82:
This partition toolset also helps u
Page 83 and 84:
Figure 3.6: Mesh control using Swee
Page 85 and 86:
Figure 3.7: Schematic representatio
Page 87 and 88:
discretization. Therefore, in this
Page 89 and 90:
According to Staten et al. (2010b),
Page 91 and 92:
3.5 The influence of pressure on th
Page 93 and 94:
each pressure value remains constan
Page 95 and 96:
Axial Stress, σa (S22) was referre
Page 97 and 98:
Figure 3.14 shows a graph of analyt
Page 99 and 100:
the notch defect is found to be in
Page 101 and 102:
One of the contributions of this st
Page 103 and 104:
Table 3.4: Convergence tests on var
Page 105 and 106:
Figures 3.16 and 3.17 show that the
Page 107 and 108:
Figure 3.20 shows that as the relat
Page 109 and 110:
3.7.4 Concluding Remarks A biaxial
Page 111 and 112:
yarns change and lateral contact be
Page 113 and 114:
For predicting E2 and Gl2, a number
Page 115 and 116:
Where: Ar = weight of one sheet of
Page 117 and 118:
Where; The Stiffness matrix is trad
Page 119 and 120:
simulations for the hoop strain at
Page 121 and 122:
Figure 3.24: Finding the hoop stres
Page 123 and 124:
Table 3.8: The effect of repair len
Page 125 and 126:
four up to 18 layers. In this study
Page 127 and 128:
hoop and axial stress concentration
Page 129 and 130:
Table 3.11: Predicted Elastic Const
Page 131 and 132:
Table 3.13: Axial Stress v/s L/W ra
Page 133 and 134:
Hoop Stress (MPa) Figure 3.3: Hoop
Page 135 and 136:
However, Frost (2008) and Alexander
Page 137 and 138:
Table 3.17: Ratio of Axial Stress N
Page 139 and 140:
axial load (e.g. buried pipelines t
Page 141 and 142:
3.10 Numerical Strain Analysis of C
Page 143 and 144:
Figure 3.36 shows the contours befo
Page 145 and 146:
Stress Concentration Factor 2 1.8 1
Page 147 and 148:
In fact, for Notch 40 with 18 plies
Page 149 and 150:
this second approach (i.e. the expe
Page 151 and 152:
4.2.2 Strain Gauge Selection and Cr
Page 153 and 154:
Finally, a method used to calculate
Page 155 and 156:
Where, and are the maximum and mini
Page 157 and 158:
should be between 20 and 28 bar and
Page 159 and 160:
welded weld neck flange at both end
Page 161 and 162:
Figure 4.7: Sensor and heating elem
Page 163 and 164: Figure 4.9: A summary of data acqui
Page 165 and 166: One of the objectives of doing this
Page 167 and 168: Figure 4.14: A schematic diagram of
Page 169 and 170: Table 4.1: Strain gauge readings at
Page 171 and 172: microstrain 3.00E-04 2.50E-04 2.00E
Page 173 and 174: However, the axial load transfer is
Page 175 and 176: Mircostrain 500 450 400 350 300 250
Page 177 and 178: Since the defect is not a flat notc
Page 179 and 180: Again, the epoxy resin plays a sign
Page 181 and 182: Table 4.4: Study at the defect area
Page 187 and 188: 14.25% in hoop strain and 23.5% in
Page 189 and 190: consideration for the selection of
Page 191 and 192: 5.4 Methodology of Thermal Expansio
Page 193 and 194: Figure 5.1: Some preventive measure
Page 195 and 196: 5.4.1 The Fundamentals of the Therm
Page 197 and 198: 5.5 Analysis of Results 5.5.1 The t
Page 199 and 200: Thermal Output (strain ) 0.0002 0.0
Page 201 and 202: (Strain ,ε) Figure 5.12: Coefficie
Page 203 and 204: (Strain ,ε) 0.0002 0.00018 0.00016
Page 205 and 206: Table 5.1: The strain readings afte
Page 207 and 208: maximum shear strain increase as th
Page 209 and 210: The above results show that shear s
Page 211 and 212: shear strains becomes more stable a
Page 213: for future safe design and operatio
Page 217 and 218: 6.3 Primary PhD Achievements The ea
Page 219 and 220: In Chapter 5, the CTE was determine
Page 221 and 222: ASM International (2002), "Thermal
Page 223 and 224: Bucinell, R. B. (2001), Calculating
Page 225 and 226: Proceedings of OMAE'01 20th Interna
Page 227 and 228: Greenwood, R. (2002), UKOPA Pipelin
Page 229 and 230: Kotrechko, S. A., Krasovskii, A. Y.
Page 231 and 232: Mousavi, S. E., Xiao, H. and Sukuma
Page 233 and 234: Raju, I. S. and Newman, J. C. (1986
Page 235 and 236: Tennyson, R. C. and Mufti, A. A. (2
Page 237 and 238: Xue, L., Widera, G. E. O. and Sang,
Page 239 and 240: General Technical Specifications: T
Page 241 and 242: Photos of the Design and Build of t
Page 243 and 244: Composite Installation Process usin
Page 245 and 246: APPENDIX 3 - STRAIN GAUGE INSTALLAT
Page 247 and 248: Figure 7: Mylar tape was stuck onto
Page 249 and 250: Figure 19: Repeating the above proc
Page 251: 2) Electrical Insulation Test Resul
show all

cranfield university mahadi abd murad an integrated structural ...

Create successful ePaper yourself

Delete template?

Save as template?