SESAM TM FOR SUBSEA EQUIPMENT

CASE STUDY SESAM TM FOR SUBSEA EQUIPMENT DOP analysis a case study In the design of subsea structures, dropped object protection (DOP) is crucial in order to prevent severe damage to oil and gas pipelines and equipment. SAFER, SMARTER, GREENER

Reference to part of this report which may lead to misinterpretation is not permissible. No. Date Reason for Issue Prepared by Verified by Approved by Draft 22 April 2015 First issue Mo Dessoukey 22 April 2015 Prepared by DNV GL - Software DNV GL AS. All rights reserved This publication or parts thereof may not be reproduced or transmitted in any form or by any means, including copying or recording, without the prior written consent of DNV GL AS.

Table of contents 1 INTRODUCTION... 1 1.1 Background 1 1.2 Scope of work 1 2 PLASTIC DESIGN... 2 3 IMPACT LOADS... 2 4 THE SOFTWARE... 2 5 THE CASE... 3 5.1 Geometry of the structural model 3 5.2 Description of dropped object scenario 3 5.3 Method 1: Principle and modelling 4 5.4 Method 2: Principle and modelling 5 6 RESULTS... 6 6.1 Method 1 6 6.2 Method 2 11 7 COMPARISON CHARTS... 16 8 CONCLUSION... 18 9 REFERENCES... 19 SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page i

1 INTRODUCTION 1.1 Background In the design of subsea structures providing sufficient dropped object protection (DOP) is of the utmost importance in order to prevent damage to critical equipment, and to protect oil and gas pipelines from severe damage. Dropped object analysis is considered an accidental case and therefore the structure can be taken into the plastic range, this is to avoid excessive weight, increased cost and over-design. The main objective of this case study is to demonstrate the simplicity of using the same Sesam model as for in-place design in a non-linear analysis simulating dropped objects. Furthermore, the case study demonstrates the capabilities of non-linear analysis in Sesam. 1.2 Scope of work The objective of this case study is to compare the two most common ways of modelling a dropped object scenario in Sesam.The first method is a model of the dropped object, where a non-linear spring element is introduced to establish contact between a falling object and a stiffened plate panel. In the second method, the node representing the point of impact is given a mass and an initial velocity equal to the velocity at impact for the same load scenario as in the first method. The structure being impacted in this case is a protection panel located on a subsea manifold, which must withstand an impact energy of 150 kj, and a total deflection of no more than 25 cm, because if the panel deflects more than 25 cm it will impact oil-gas pipes that are located below the protection panel, and risk an oil leak. Design for dropped object protection (DOP) often requires that the steel be taken into the plastic range and the large displacements accounted for, which means that nonlinear analysis methods must be used. The accidental limit state (ALS) design check for this panel is outside the primary focus of this case study, which is primarily to compare two different ways of analysing DOP on a stiffened panel in USFOS. This case study evaluates one load scenario only for both methods, and is therefore not a complete design check. For a complete design check more places of impact and loading scenarios should be considered. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 1

2 PLASTIC DESIGN Most design is based upon the supposition that maximum capacity of the structural element is reached at yield strength or initial buckling (also known as linear design). In other words, exceeding either will lead to the failure of that element. But since DOP design often requires us to take the structural element into the plastic region, elasto-plastic behaviour of the element must be considered. Elasto-plastic or nonlinear analysis allows for redistribution of stresses in an element over its cross section when some parts start to yield, and thus it allows for the absorption of most of the impact energy from a dynamic load through plastic non-elastic deformation. 3 IMPACT LOADS In the design against impact loads (where the eigenperiod of the load is smaller than the eigenperiod of the structure), the dynamic effects on the structure are of importance. The dynamic load is applied incrementally in time-steps, and the structural stiffness is solved at every step based on the structural configuration, nodal coordinates, and element forces for that step. Equilibrium must be achieved at every step through the use of iterations. 4 THE SOFTWARE Usfos (part of Sesam) is a finite element program for non-linear static and dynamic analysis of frame structures. The structure may be exposed to external loads, acceleration fields or temperature fields. It accounts for nonlinear effects such as large structural motions and inelastic deformations. The theoretical basis of the computer code is continuum mechanics and the finite element method with the basic idea that only one finite element represents one physical element in the structure. Usfos also operates on element stress resultants, i.e. forces and moments. Material non-linearities are modelled by plastic hinges at element mid span and at element ends. The element formulation is based on the exact solution to the differential equation for a beam subjected to end forces. Effects of large displacements and coupling between lateral deflection and axial strain are included by using non-linear strain relations (Green strain) instead of conventional linear strain (engineering strain). This gives a very accurate representation of the element behaviour, including membrane effects and column buckling. In order to run an analysis in Usfos, two files are needed: one model file, where all the information about the structure is stored (nodal co-ordinates, beam names, sizes, etc.) and one control file that consists of the analysis command lines and control parameters, and is created by the engineer. The model file is typically converted from GeniE, and if minor modifications are required, the use of a text editor can be used to modify the model file. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 2

5 THE CASE 5.1 Geometry of the structural model In our case study the model is a protection panel located on the top of the Subsea manifold; it consists of stiffened plates and some gapes and holes. The panel was modelled in GeniE as a concept model, from which a full FE mesh was created. The FE mesh was then imported into Usfos using the utility tool Fem2UFO (a freeware conversion tool). The panel is approximately 6 m in length and 3 m in width. The plate thickness of the panel is 5 cm. All welds are full penetration welds. The panel consists of steel of material quality S 355, with yield strength of 355 N/mm 2 and an elasticity modulus of E = 205 kn/mm 2. 5.2 Description of dropped object scenario Since the design impact energy is 150 kj, this impact energy was achieved by dropping an object of 5000 kg from a height of 3 m. Giving us: Impact velocity = 7.67 m/s Total Kinetic Energy = ½ mv² = 147kJ 150kJ The dropped object was considered infinitely rigid, so no energy is absorbed by the deformation of the dropped object, as the effects on the dropped object itself is outside the scope of this case study. It is of importance to note that this case study does not represent the worst case impact scenario and that other more governing cases are possible, and might cause other failure modes such as buckling or rupture. It is also of note that the water dampening effects on the panel were not accounted for, even though it is possible to model this in Usfos. It is more conservative to ignore the water dampening effects. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 3

5.3 Method 1: Principle and modelling To model Method 1, we require four elements: the dropped object itself, which weighs 5000 kgs, a nonlinear spring to define the impact location, a dummy member, whose sole purpose is to keep the dropped object in its initial position until we decide to initialize the fall, and finally we require the stiffened protection panel which is being analysed. At the start of the analysis, a Usfos command is used to break the dummy beam, leaving the dropped object free to fall unto the panel. The command used is USERFRAC and this breaks the dummy member in two, with the lower half of the dummy member falling with the dropped object. The weight of the dummy member was negligible and thus no adjustment to the weight of the dropped object was required. The non-linear spring stiffness is defined using a P-δ curve. The stiffness of the spring is very important, as a very stiff spring will not allow the dropped object to impact the structure, and if the stiffness is too low the object will simply go thru the structure without impacting it. For this case, the spring has zero stiffness from the height of 3 m to 0.1 m, and very high stiffness right before the point of impact. This is to allow the dropped object to act on the structure. P δ 0 4 3,5 3 2,5 2 1,5 1 0,5 0 0,5 1 1,5 0,2 0,2 P (N) δ (m) 0,4 0,6 0,8 1 1,2 SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 4

5.4 Method 2: Principle and modelling The modelling principle for Method 2 is quite simple. A mass of 5000 kg is attached to the location of impact and an initial velocity of 7.67 m/s (which is the impact velocity of a 5000 kg object falling from a height of 3 m) is applied to this node at the start of the analysis. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 5

6 RESULTS 6.1 Method 1 Method 1 at t=0s Initial positions before the start of the analysis. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 6

Method 1 at t= 20.83s 20.83 seconds after the start of the analysis Notice all elements in red have gone plastic; also the damage to the panel is quite extensive. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 7

Method 1 Results i. Maximum displacement Secondary Impact From the above graph, we note that at impact the panel deflects to 21 cm before rebounding to a permanent plastic deformation of 16 cm. It is also important to note that the initial impact is followed by a secondary impact (at t=21.41 s), which happens when the dropped object bounces off and then reimpacts the panel. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 8

Method 1 Results ii. Nodal Velocity at impact From the graph above, we can see that the impact velocity at the impact location is only 7.5 m/s, which is lower than the 7.67 m/s expected value. The difference is due to the non-linear spring stiffness, which absorbed some of the impact energy. The stiffness of the non-linear spring can be adjusted to match the expected value. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 9

Method 1 Results iii. Kinetic Energy of impact Secondary Impact The Impact energy above is also lower than the expected 150kJ impact, due to the spring stiffness absorbing some of the impact energy. To overcome this, we can add more mass to the falling object, increase the falling height or reduce the spring stiffness. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 10

6.2 Method 2 Method 2 at t=0s Initial positions before the start of the analysis. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 11

Method 2 at t=20.78s 20.78 seconds after the start of the analysis SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 12

Method 2 Results i. Maximum displacement For method 2, note the initial deformation of 0.05 m due to the node mass acting on the structure. At t=20.79 s the initial impact happens and the panel is deformed by 22 cm. Then the panel rebounds to leave a total plastic deformation of 16 cm. Note the sinusoidal vibrations on the panel due to the node mass being stuck to the node and following the panel up and down. There is no secondary impact in this case since the impact mass is not detached as was the case in the first method. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 13

Method 2 Results ii. Nodal Velocity at impact The graph shows that the impact velocity at the impact location is 7.67 m/s which is exactly as expected, since we defined the velocity. Again the sinusoidal vibrations become dominant after impact. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 14

Method 2 Results iii. Kinetic Energy of impact Impact energy of exactly 147 kj as expected. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 15

7 COMPARISON CHARTS 4 2 Nodal Velocity (m/s) at node 650 0 19,5 20 20,5 21 21,5 22 2 4 6 8 Node Mass Method Dropped Object Method 10 Time (s) Note the more realistic behaviour of the first method. Also note the secondary impact. 1,60E+05 1,40E+05 146796 139835 Kinetic Energy (kj) 1,20E+05 1,00E+05 8,00E+04 6,00E+04 4,00E+04 Node mass method Dropped Object Method 2,00E+04 0,00E+00 19,5 20 20,5 21 21,5 22 Time (s) Kinetic energy of impact is higher for the node mass method where there is no energy absorption due to the spring stiffness, as in the dropped object method. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 16

0 19,5 20 20,5 21 21,5 22 0,05 Global Deflection (m) 0,1 0,15 Dropped Object Method Node Mass Method 0,2 0.211397 0.217891 0,25 Time (s) More realistic global deflection in the dropped object method as compared to the node mass method, which exhibits signs of sinusoidal vibrations. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 17

8 CONCLUSION The conclusion from this case study is that the protection panel gives enough protection for this design scenario. Both methods give very similar results, i.e. a total displacement of 21cm for method 1 and 22 cm for method 2, both of which are below the 25 cm design requirement. The complete design of this panel for all loading scenarios was outside the scope of this case study. In method 1 the dropped object is acting upon the structure, giving us more realistic energy dissipation due to the object detachment. Also, secondary impact can be analysed. In this particular case the secondary impact is of little importance since there is no energy input into the system, i.e. the most damage will happen on initial impact. This might not be the case if the dropped object is actively pushed down by an outside source. Method 1 is less conservative because no mass is attached to the structure, but this method involves much more complex modelling. In method 2 the dropped object is attached to the structure. It is more conservative because a permanent mass is stuck to the structure. After initial impact the structure goes into sinusoidal vibrations, due to the attached mass moving up and down with the structure. This makes energy dissipation in the model less accurate and the study of secondary impacts impossible. Method 2 is much less time consuming with regards to the modelling. For this particular design scenario, method 2 would have been sufficient, given the fact that secondary impact is of no importance. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 18

9 REFERENCES /1/ SINTEF GROUP (2001). USFOS Getting Started. Structural Engineering, Marintek, SINTEF GROUP. Light version of the USFOS Theory Manual. /2/ USFOS User s Manual, modelling (1999) /3/ USFOS Getting Started (2001) Light version of theory manual including course material Structural Engineering, Marintek, SINTEF GROPUP /4/ USFOS User s Manual, Manual Commands (1999) http://www.usfos.com /5/ Singelstad, Anne (2009) Master Thesis - Nonlinear analysis of a space frame subjected to loading from dropped objects. University of Stavanger, Norway. /6/ Fem2UFO http://bitbucket.org/svortega/fem2ufo/downloads. fem2ufo is a Python library to convert FE structural models generated by DNV GL's GeniE-Sesam to USFOS format. SESAM FOR SUBSEA Sesam www.dnvgl.com/software Page 19

ABOUT DNV GL Driven by our purpose of safeguarding life, property and the environment, DNV GL enables organizations to advance the safety and sustainability of their business. We provide classification and technical assurance along with software and independent expert advisory services to the maritime, oil and gas, and energy industries. We also provide certification services to customers across a wide range of industries. Operating in more than 100 countries, our 16,000 professionals are dedicated to helping our customers make the world safer, smarter and greener. SOFTWARE DNV GL is the world-leading provider of software for a safer, smarter and greener future in the energy, process and maritime industries. Our solutions support a variety of business critical activities including design and engineering, risk assessment, asset integrity and optimization, QHSE, and ship management. Our worldwide presence facilitates a strong customer focus and efficient sharing of industry best practice and standards.