Beruflich Dokumente
Kultur Dokumente
Residual Stress on Ultimate Strength Article Type: Research Paper Keywords: Plates, Stiffened Panels, Ultimate Strength, Geometric Distortion, Residual Stress Corresponding Author: Dr. Imtaz Ali Khan, B.Tech (IIT Madras), PhD Corresponding Author's Institution: First Author: Imtaz Khan, PhD Order of Authors: Imtaz Khan, PhD; Imtaz Ali Khan, B.Tech (IIT Madras), PhD Abstract: During the welding of steel plated structures, geometric distortions and residual stresses are developed. The influence of welding induced geometric distortions and residual stresses on the compressive ultimate strength of plates and stiffened panels were investigated using nonlinear finite element analyses considering a range of plate thicknesses with various levels of residual stresses. The results and insights observed from the present study are presented.
Manuscript Click here to download Manuscript: Residual stress effect on ultimate strength_071009.doc
1 Introduction
Understanding the behaviour of unstiffened and stiffened plates under predominantly compressive loads is very important for design of ship structures. Often the hull girder strength depends very much on the collapse strength of the deck. The compressive strength of unstiffened plates and stiffened panels is mainly governed by its slenderness (Faulkner, 1975). However the influence of geometric deformations has long been recognised and proposals have been made to incorporate these deformations in the design process. Steel plated structures, such as ships, offshore platform decks and box girder bridges, are welded as assemblies of individual steel plate elements. When stiffening members are welded to the plate, the welding temperatures take on such extreme values that considerable residual stress and geometric distortions are developed. A number of studies related to effect of welding-induced geometric deflection magnitude on the collapse behaviour of steel plates have been previously carried out, i.e. Dwight & Moxham (1969), Faulkner et al. (1973), Carlsen & Czujko (1978), Masubuchi (1980), Ueda (1999), Masoaka & Mansour (2004) and Paik et al. (2004). While the shape of welding-induced geometric distortion in steel plate elements is normally very complex (Antoniou 1980, Antoniou et al. 1984, Kmiecik et al 1995), it has also been recognised that the shape of geometric distortions can significantly effect the behaviour of plate elements up to and after the ultimate strength is reached. In this regard, considerable research efforts have been devoted to investigate the effect of welding induced geometric distortion shape on plate collapse behaviour (e.g. Ueda & Yao 1985, Paik & Pedersen 1996, Guedes Soares 2005). The insights developed by these studies are quite useful to accommodate geometric distortion shape into the ultimate limit state design of steel plate elements as a parameter of influence. However most previous studies on this regard were concerned with the effect of the geometric distortion alone rather than the combined effect of residual stress and geometric distortions. It has been noted that the ultimate compressive strength of plates and stiffened panels decreases due to the presence of welding induced residual stress and geometric distortions. The aim of this study is to investigate the influence of welding induced residual stress on the collapse behaviour of steel plate elements with a fixed realistic shape of geometric distortion. The amount of residual stress has been varied to categorise its effect on the ultimate compressive strength. Corresponding author: Dr. Imtaz Khan Lloyds Register, 71 Fenchurch Street, London EC3M 4BS
w z = w p cos
m x y sin a b
where
m =5
(1) where m is the number of half-waves in the longitudinal direction of the plate. a is the longitudinal frame spacing and b is the spacing between two stiffeners. The x-axis is along the longitudinal direction and the y-axis is along the transverse direction of the plate as shown in the Fig. 1. wp is the maximum amplitude of deflection in the plate and is equal to b/200. Simply supported boundary conditions have been considered for the plate and are shown in Fig. 1.
Fig 1: The shape of geometric distortions and the boundary condition for the plate.
For stiffened panel, two types of geometric distortions have been considered and they are: Local mode of distortion Global mode of distortion
The local mode of distortions is experienced in the plate and the web (Fig. 2a, b). In local mode of distortion, the web and plate have a sinusoidal mode along the longitudinal and transverse direction. The exact shape of the distorted model can be visualised from the Fig. 2a and 2b. The global mode of distortion is considered for the plate and a tilt in the stiffener is incorporated as can be seen from Fig. 2 c, d. The global mode of distortion also has sinusoidal shape with different phase difference. The local and global geometric distortions have been superimposed to create the desired
d ei pl p A d a o L h s a u q S
d e x Fi 3 R U , 1 R U , 3 U
U 1 , U 3 , U R 1 , U R 3 Fi x e d
distortion shape as shown in Fig. 3 (left). The scantling and material property of the stiffened panel is shown in Table 1.
w w1 = ww cos
nz mx sin hw a
sin
sin
Fig. 2: The shape of geometric distortion in the stiffened panel model. (a) local mode of distortion in the plate and web for the panel section at x=0 along the transverse direction; (b) the shape of the stiffened panel plate along the longitudinal direction; (c) the shape of the global mode of distortion along the longitudinal direction; (d) shape of global mode of distortion in the plate, web and flange.
Fig 3: Left, welding induced geometric distortion in the stiffened panel considering local and global modes (magnified by 30 times). Right, applied boundary condition for the stiffened panel. The plate along the line AB (transverse axis) has been fixed in vertical direction.
, where
= sin 1
ws hw
w g1 = w g cos
w p1 = w p cos
mx
ny
Table 1: Scantlings of the stiffened panel with 4 stiffeners. a (mm) b (mm) tp (mm) hw (mm) tw (mm) bf (mm) tf (mm) Young's Modulus (E) Poisson ratio( ) Yield Stress ( y )
4300 815 17.8 463 8 172 17 205.8 GPa 0.3 315 MPa where a = frame spacing along longitudinal direction; b = spacing between two stiffeners; tp =plate thickness; h w = web height; tw = web thickness; bf = breadth of flange; t f= flange thickness
2.2 Welding Induced Residual Stress Fig. 4 represents a typical welding induced residual stress pattern in a welded box member. Typical residual stress distributions in welded steel members represent the tensile residual stresses developed in the heat-affected zones and the compressive residual stresses that then must also coexist to achieve an equilibrium condition in the plane of the member. For practical design purposes, the welding-induced residual stress distributions of a plate element between support members may be idealised to be composed of tensile and compressive stress blocks, as shown in Fig. 5. The equilibrium requirement for the tensile and compressive stresses provides a relationship between the magnitude of compressive residual stress, r in the plating and the widths .t of the tension zones each side of the weld (Faulkner, 1975):
r 2 = 0 ( b / t ) 2
(2)
0=402
Fig 5: Idealised residual stress pattern in a plate and a web with edge welds (definition of ). To maintain equilibrium, the applied stress area multiplied by the applied compressive residual stress should be equal to the area multiplied by the tensile residual stress.
The magnitude of residual compressive stress is related to (Eq. 2). Generally for welded ships is considered to have values between 4.5 and 6, but allowing shakedown, values between 3 and 4.5 are considered more appropriate. In this study three levels of compressive residual stress have been considered, namely low (0.05 y), medium (0.15 y) and high (0.25 y). The values of for these magnitudes of residual stress are 2.45, 4.99 and 6.4 respectively. The values for this study are derived based on the amount of residual compressive stress applied to the chosen scantling of the plate.
thick the effect of residual stress is marginal, but for moderately thick plate ( =1.5, 1.79 in this case) the effect of residual stress is very prominent. Fig. 6 represents a plate with applied residual stress before the nonlinear analysis has been performed. The compressive residual stress block is surrounded by two tensile residual stress blocks. Fig. 7 shows the load shortening curves for the plate with thickness 9.1 mm. The load shortening curve for the imperfect plates without residual stress is observed to decline sharply after it has achieved the ultimate strength, and the ultimate strength is observed to be marginally higher compared to the other cases considering the presence of welding induced residual stresses. With the increase of residual stress magnitude the ultimate strength decreases marginally for this plate, but corresponding strain at the ultimate strength also increases with an increase of residual stress. It could be argued that the presence of tensile residual stresses induces certain degrees of hardenings; hence the ultimate strengths are achieved at higher strain values.
Fig. 6. Meshed imperfect plate model showing the welding induced residual stress distribution (a x b = 4300 x 815 mm).
0.5
0.4
0.1
0.5
1.5
Fig. 7: Load shortening curve of the 9.1 mm thick plate with varying magnitudes of compressive residual stress.
Ultimate compressive strength of the 12.8 mm thick plate
0.7
0.6
0.5
0.4
0 0 0.5 1
1.5
y
Fig.8: Load shortening curve of the 12.8 mm thick plate with varying magnitudes of compressive residual stress.
u/
0.3
wopl=b/200, =(b/t)(
u/ Y
0.3
0.2
Compressive residual stress = 0 a x b x t = 4300 x 815 x 9.1 mm E=205.8 GPa, Y=315 MPa
1/2 Y /E) =
3.50
2.5
2.5
0.8
0.7
0.6
0.5
0.2
0.1
/1 y
1.2
1.4
1.6
Fig.9: Load shortening curve of the 17.8 mm thick plate with varying magnitudes of compressive residual stress.
Ultimate compressive strength of the 21.3 mm thick plate
1
0.9
0.8
0.7
0.6
u/ Y
0.5
0.4
0.1
0.5
1.5
Fig. 10: Load shortening curve of the 21.3 mm thick plate with varying magnitudes of compressive residual stress.
0.3
wopl=b/200, =(b/t)(
wopl=b/200, =(b/t)(
u/
0.4
= 1.79
1.8
1.50
2.5
0.8
0.2
0.5
1.5
Fig. 11: Load shortening curves of the 32 mm thick plate with varying magnitudes of compressive residual stress.
wopl=b/200, =(b/t)(
0.6
u/
0.4
= 1.00
2.5
Table 2: Ultimate compressive stress of the plate (a x b x t = 4300 x 815 x t mm) with geometric distortions and welding induced residual stress.
Thickness
% of Comp.
% of Tensile
9.1 =3.50
0 5 15 25 0 5 15 25 0 5 15 25 0 5 15 25 0 5 15 25
0.533 0.5317 0.5254 0.5185 0.6443 0.6429 0.6353 0.6253 0.8109 0.7824 0.7427 0.7319 0.8966 0.8673 0.8221 0.796 0.986 0.9779 0.9724 0.9705
12.8 =2.49
17.8 =1.79
21.3 =1.5
32 =1.00
Fig. 8 presents the load shortening curves for the plate with 12.8mm thickness. The curve with zero residual stress shows smooth decline and change of slope. The curves with welding induced residual stress show a different pattern. The load shortening curve with 0.05 y of welding induced compressive residual stress demonstrates a smooth decline in the post collapse region, where as the load shortening curves with 0.15 y and 0.25 y residual compressive stress show sudden decline in the post collapse region. Fig. 9 represents the load shortening curve for the plate with 17.8 mm thickness. The load shortening curves show different behaviour compared to the previously discussed 9.1 mm and 12.8mm thick plates. When the plate doesnt have any residual stress the change in the slope of load shortening curve after the collapse is higher compared to the plates with residual stress. Furthermore with an increase in the magnitude of residual stress, the strain at collapse also increases. This phenomenon may be due to the presence of the residual stress. Similar behaviour is observed for the 21.3mm and 32 mm thick plates as shown in Figs. 10 and 11. From Table 2, the change in the ultimate strength of plates due to presence of residual stress can be observed. It can be noticed that with an increase in the residual stress magnitude, the ultimate strength decreases. On the other hand the margin of decrease in the ultimate strength is dependent on the slenderness ratios or the thickness of the plates. Fig. 12 represents the decrease in the ultimate strength of plates with respect to the slenderness ratios ( ) of the plates. For very thick plates the decrease in the ultimate strength due to an increase in residual stress is small and similar trends are also observed for thin plates. The plates with slenderness ratios 1.49 and 1.79 show a large decrease in the ultimate strength with increased welding induced residual stress.
10
(mm)
Residual stress
Residual stress
u/ y
( u / y )r ( u / y )0 1 * 100
-0.244 -1.426 -2.720 -0.217 -1.397 -2.949 -3.515 -8.410 -9.742 -3.268 -8.309 -11.220 -0.822 -1.379 -1.572
% Difference
12
0 0.5 1 1.5 2
2.5
Slenderness Ratio ( )
3.5
Fig. 12: Decrease in the ultimate compressive strength of plate with different slenderness ratios with varying amount of welding induced residual stress.
11
Fig 13: Meshed imperfect stiffened panel model showing the residual stress distribution in the plate and the webs.
Ultimate strength of the stiffened panel with initial imperfections and residual stress, plate thickness (tp) =12.8mm 0.8
0.7
0.6
0.5
0.1
Fig 14: Load shortening curves of stiffened panel with 12.8 mm thick plate considering different values of compressive residual stress
12
0.5
1.5
=(b/tp)(
Y /E)
u/ y
0.4
0.3
0.2
a x (b x tp + hw x tw + bf x tf) = 4300 x( 815 x 12.8 + 463 x 8 + 172 x 17) mm E = 205.8 GPa, Y = 315 MPa wos = woc = a/1000, wow=hw/200, wopl =b/200,
1/2
= 2.49
2.5
Ultimate strength of the stiffened panel with initial imperfections and residual stress, plate thickness (tp) =17.8mm 0.9
0.8
0.7
0.6
0.5
0.1
Fig. 15: Load shortening curves of stiffened panel with 17.8 mm thick plate considering different values of compressive residual stress
Ultimate strength of the stiffened panel with initial imperfections and residual stress, plate thickness (tp) =21.3 mm
0.9
0.8
0.7
0.6
0.5
0.1
0 0 0.5 1
1.5 /
y
Fig. 16: Load shortening curves of stiffened panel with 21.3 mm thick plate considering different values of compressive residual stress
=(b/tp)(
1/2 Y /E) =
u/ y
0.4
0.3
a x (b x tp + hw x tw + bf x tf) = 4300 x( 815 x 21.3 + 463 x 8 + 172 x 17) mm E = 205.8 GPa, Y = 315 MPa wos = woc = a/1000, wow=hw/200, wopl=b/200, 1.50
0.2
0.5
1.5
=(b/tp)(
1/2 Y /E) =
u/ y
0.4
0.3 a x (b x tp + hw x tw + bf x tf) = 4300 x( 815 x 17.8 + 463 x 8 + 172 x 17) mm E = 205.8 GPa, Y = 315 MPa wos = woc = a/1000, wow=hw/200, wopl=b/200, 1.79
0.2
2.5
2.5
13
Fig. 17: Typical load shortening curve showing the no load shedding zone.
Table 3: Ultimate compressive stress of the stiffened panel with geometric distortions and welding induced residual stress (a x b x t = 4300 x 815 x t mm) Thickness (t) % of Comp. % of Tensile
u/ y
% Difference
1 * 100
Residual stress 0 5 15 25 0 5 15 25 0 5 15 25
Residual stress 0 60 80 100 0 60 80 100 0 60 80 100 0.7053 0.6925 0.6679 0.6331 0.7985 0.7785 0.7470 0.7000 0.8545 0.8313 0.7916 0.7442
17.8 =1.79
21.3 =1.5
14
( u / y )r ( u / y )0
14%
12%
10%
8%
6%
4%
2%
5% 15% 25%
0% 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3 Plate Slenderness Ratio ( )
Fig. 17: Decrease in the ultimate compressive strength of stiffened plate with different plate slenderness ratio with varying welding induced residual stress.
The decrease in the ultimate compressive strength of the stiffened panels with a presence of welding induced residual stress is presented in Fig. 17. It is observed from the figure that with an increase in the residual stress, the ultimate compressive stress decreases further. The stiffened panel with a slenderness ratio of 1.5 has a higher rate of change compared to the stiffened panel with a slenderness ratio of 2.49.
15
c)
Due to the presence of welding induced residual stress, the plates and stiffened panels reach their ultimate strength at higher strains. The presence of tensile residual stresses induces certain degree of hardening effect; hence the ultimate strengths are achieved at higher strains. For the cases considered the effect of welding induced residual stress is significant for plates with slenderness ratios around 1.5 and 1.8. Where as for the stiffened panels the effect of residual stress becomes significant when high value of compressive residual stress is applied. In the post collapse region of the plates and stiffened panels with welding induced residual stress, the load shortening curves show higher values of stress compared to that without any residual stress, so the structural capacity of the plates and stiffened panel are higher with the presence of residual stress in the elements.
d)
e)
This study was carried out on a limited number of structural models and a few levels of residual stresses. More studies would be beneficial on a range of structural parameters and various levels of residual stress levels to obtain inclusive conclusions.
Acknowledgement:
The authors would like to thank Mr. Tim Hall and their colleagues in Lloyd's Register for their support and comments. The views expressed in this paper are those of the authors and are not necessarily those of Lloyd's Register. REFERENCE:
ABAQUS User manual. Release 6.7 Antoniou AC, Lavidas M & Karvounis G. 1984. On the shape of post-welding deformations of plate panels in newly built ships, Journal of Ship Research, Vol. 28, No. 1, pp 110. Antoniou AC. 1980. On the maximum deflection of plating in newly built ships, Journal of Ship Research, Vol. 24, No 1, pp 3139. Carlsen CA & Czujko J. 1978. The specification of post-welding distortion tolerances for stiffened plates in compression, The Structural Engineers, 56A, 5, 133141. Dwight JB & Moxham KE. 1969 Welded steel plates in compression, The Structural Engineers, 47, 2, 4966. Faulkner D, Adamchak JC, Snyder GJ & Vetter MF. 1973. Synthesis of welded grillages to withstand compression and normal loads, Computers and Structures, 3, 221246. Faulkner D. 1975. A Review of Effective Plating for the Analysis of Stiffened Plating in Bending and Compression. Journal of Ship Research, Vol. 19, No.1, pp 1-17. Guedes Soares C, Teixeira AP, Luis RM, Quesnel T, Nikolov PI, Steen E, Khan IA, Toderan C, Olaru VD, Bollero A & Taczala M. 2005. Effect of the shape of localized imperfections on the collapse strength of plates. Maritime Transportation and Exploitation of Ocean and Coastal Resources, Guedes Soares C, Garbatov Y and Fonseca N (Eds.), pp 429-438. Kmiecik M, Jastrrebski T & Kuzniar J. 1995. Statistics of ship plating distortions, Marine Structures, 8, 119132. Kmiecik M. 1970. The Load Carrying Capacity of Axially Loaded Longitudinally Stiffened Plates Having Initial Deformation, Ship Research Institute, Technical Report No. R80, Trondheim, Norway. Masaoka K & Mansour AE. 2004. Ultimate Compressive Strength of Imperfect Unstiffened Plates: Simple Design Equations. Journal of Ship Research, Vol. 48, No. 3, pp 191-201. Masubuchi K. 1980. Analysis of Welded Structures, Pergamon Press, Oxford, UK.
16
Paik JK & Pedersen PT. 1996. A simplified method for predicting ultimate compressive strength of ship panels, International Shipbuilding Progress, 434, 139157. Paik JK & Thayamballi AK. 2003. Ultimate Limit State Design of Steel-plated Structures, John Wiley and Sons, Chichester, UK. Paik JK, Thayamballi AK & Lee JM. 2004. Effect of Initial Deflection Shape on the Ultimate Strength Behaviour of Welded Steel Plates under Biaxial Compressive Loads. Journal of Ship research, Vol. 48, No. 1, pp 45-60. Smith CS, Davidson PC, Chapman JC & Dowling PJ. 1988. Strength and stiffness of ships plating under in-plane compression and tension, The Royal Institution of Naval Architects Transactions, 130, 277296. Ueda Y & Yao T. 1985. The influence of complex initial deflection modes on the behaviour and ultimate strength of rectangular plates in compression, Journal of Constructional Steel Research, 5, 265302. Ueda Y. 1999. Computational Welding Mechanics, in commemoration of his retirement from Osaka University, Osaka, Japan.
Lloyd's Register, its affiliates and subsidiaries and their respective officers, employees or agents are, individually and collectively, referred to in this clause as the Lloyd's Register Group. The Lloyd's Register Group assumes no responsibility and shall not be liable to any person for any loss, damage or expense caused by reliance on the information or advice in this document or howsoever provided, unless that person has signed a contract with the relevant Lloyd's Register Group entity for the provision of this information or advice and in that case any responsibility or liability is exclusively on the terms and conditions set out in that contract.
17