You are on page 1of 7

Proceedings of OMAE’02

21st International Conference on Offshore Mechanics


and Arctic Engineering
June 23-28, 2002, Oslo, Norway

OMAE2002-28191

STRUCTURAL OPTIMIZATION ON TOPSIDE SUPPORTS FOR SHIP-SHAPED


HULLS

Machado, Christiane Lopes & Santos, Marco Antonio

PROJEMAR S.A. Estudos e Projetos de Engenharia


th
Av. Almirante Barroso, 63 – 29 floor
Rio de Janeiro – RJ, 20031-003, Brazil
projemar@projemar.com – www.projemar.com

ABSTRACT The conclusions present cost and benefit for each solution,
PROJEMAR S.A. , a naval and offshore design company and some discussions on the time acceleration for the
at Brasil has developed several conversion projects of FPSO’s conversion.
to operate at Campos Basin. This way, PROJEMAR faced some
interface difficulty in the structural design of the topside NOMENCALTURE
supports, mainly for the vessel reinforcement.
The structural analysis of the cargo region, when performed SCF - Stress Concentration Factor
together with the topside supports and loads, introduces LT - Length of the Topside (m)
significant buckling and fatigue problems in the main transverse BT - Breadth of the Topside (m)
and longitudinal structure of the vessel. The solution adopted in HT - Distance from the topside deck to the hull deck (m).
first instance is to reinforce the transverse structure and WT - Weight of the topside (tons)
longitudinal bulkheads of the cargo tanks just below the topside t - Thickness of the structural member in question (mm).
supports. This reinforcement is expensive and takes a lot of Lpp - Length of the vessel between perpendiculars (m)
time to be done, as hundreds of buckling bars should be B - Breadth of the vessel (m)
installed. D - Depth of the vessel (m)
In order to avoid this work inside the cargo region, T - Maximum draft of the vessel (m)
PROJEMAR evaluated this study to minimize the steelwork Ltank - Length of the cargo tanks of the vessel (m)
inside the cargo region, and to optimize the steel weight for
topside supports. PROJEMAR analyzed three different INTRODUCTION
concepts of topside supports: two transverse bulkheads During the last 15 years, the Brazilian offshore installations
supporting each topside module, two sets of strong brackets grew up in a significant rate. Almost all the units recently
supporting each topside module, and sets of pillars supporting installed were converted ones.
each topside modules. PROJEMAR has then experienced several different designs
The results present the amount of steelwork inside and for the Brazilian offshore basins, with a great number of them in
outside the cargo tanks, the total amount of steel reinforcement FPSO’s.
after the evaluation of stress, deflection, buckling and fatigue The critical point of the structural conversion of the old
verification. VLCC’s in FPSO’s is the topsides supports structural

1 Copyright © 2002 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 11/26/2014 Terms of Use: http://asme.org/terms


arrangement. The deck structures of tankers were originally
designed for no loads on deck besides the environmental
conditions. Then , they have to be strongly modified to support
the big additional loads from the topsides.
However, the free area of the hull deck gets very congested
when the topside facilities are installed. Then, it becomes a
challenge to optimize the structural impact of the topside
supports into the whole general arrangement.
PROJEMAR then defined a typical VLCC hull to study the Figure 2 – Truss design – Lateral view
optimum topside supports in FPSO conversions. The main
characteristics of the typical VLCC used as basis for this study The truss design has the following characteristics
are presented below: concerning the stress distribution and response for the several
Lpp = 320 m modes of failure:
B = 55 m 1. The truss transfers lower loads to the hull structure
D = 28 m through lots of points. The hull deck plating and under
T = 20 m deck main girders receives local concentrated loads
Ltank = 65 m. and high SCF’s on deck plating and girder web.
2. The truss members are strongly affected by the hull
The analysis of the hull structure should address stresses, girder bending moment and shear forces. Due to this,
buckling and fatigue. Depending on the adopted solution, one of they must be designed top support the elongation of
the structural failure modes will be the relevant.
the deck plating in a ratio of ∆l = 9,0 E − 4.LT . The
The stress and buckling criteria were based on ABS Rules
for Vessels and FPSO’s (see references 1 and 2). The fatigue local deflection of the deck transverse girder is another
analysis was carried out according ABS recommendations on critical effect for the truss design.
ABS Rules Part 5 (reference 1), and also guided by Fatigue 3. The buckling failure is not critical once the
Handbook (see reference 3). concentrated loads are acting on one unique panel,
Three different solutions are addressed in this study: which is easy to fix due to the relative reduced load
- the truss supports; magnitude.
- the stools supports; 4. The fatigue damage is another governing condition
- the partial bulkhead supports. that should be developed. The high SCF’s from the
Almost all the solutions above presented finished in truss connection to the hull generates high
the need of redesign the under-deck structure. probabilities of cracks on the deck structures. Here
again, the design must address the stochastic analysis
TRUSS SUPPORTS of all the truss connections to the hull, for a fatigue
The truss supports design consists on a truss structure safety of factor of at least 2.0.
supporting the topsides. The length of all members should be
similar. This way, the configuration of the truss is simple and Based on the main characteristics above described, the
should be placed at every frame of the hull. One example of the structural detail for connection between the hull and the truss is
truss analyzed on this study is presented on figure 1. the most important feature to be designed.
This connection comprises two brackets of thick plate
forming a cross-joint. The truss chords and braces ends
superposed on the tops of these brackets, which are welded
aligned with the deck longitudinals and transverses. The SCF
for this connection were calculated based on a fine mesh model
( t x t mesh size) , composed by plats and solids, as 4.0 for the
chords and braces connection with the plate supports and 1.8
for the welding of the plate supports on the hull structure..
Figure 1 – Truss design – Front view

2 Copyright © 2002 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 11/26/2014 Terms of Use: http://asme.org/terms


PARTIAL BULKHEAD SUPPORTS 4. The buckling is the governing mode of failure in this
design concept. As big vertical loads are transferred
The partial bulkhead supports conception consists on two from the topsides to the deck transverses, these ones
transverse bulkheads that supports the topside modules. In order need to be deeply investigated for buckling and
to reduce the moment on the connection between the topside correctly reinforced to avoid the plastic and elastic
module and the bulkhead due to the deck plating elongation, buckling. Sometimes, it gets necessary to provide
one of these connections will slide upon the bulkhead,. Then, inserts on the deck transverse webs to reduce the
the sliding support receives the force loads and horizontal vertical normal stress level and then the buckling usage
moments. factor.
Some holes are designed on the transverse bulkhead to 5. The fatigue failure is studied only for the transverse
accommodate the general arrangement of the piping and safety ends of the bulkheads and for the longitudinal ends of
systems through the deck of the vessel. Sketches of the the bulkheads tripping brackets. In fact, the design of
transverse bulkhead option are presented on figure 3. these ends are governed by a fine mesh FE model to
As the transverse bulkheads absorbs moments from the maintain the SCF less or equal to 1.5.
longitudinal acceleration of the module, they need to be
reinforced with major tripping brackets aligned with the topside STOOLS SUPPORTS
supports, in order to resist to the horizontal moment imposed on The stools supports conception consists on the reduction of
the bulkhead. For the sliding support, the same situation applies the closed area of the two transverse bulkheads so that the
to it, once the structure had to overcome the static and dynamic become two or three supports. One row of supports are sliding
friction loads to slide upon the topsides and then releases the ones similar to the bulkhead option.
structure from the deck elongation originated by the hull girder The stool option intends to reduce the steel installed above
bending moments. deck, reducing then the congestion on the deck area and also
the blast loads and safety equipment to be installed.
Sketches of the stool option are presented on figure 4.

Figure 3 – Bulkhead Supports – Front view

The transverse ends of the bulkhead supports, and the Figure 4 – Stools Supports – Front view
longitudinal ends of its tripping brackets needs to be correctly
designed so that the hot spot stresses generated from the lateral The relevant points on this design concept are stated below:
abrupt finish of the partial bulkhead on the deck structure does
not impose an inadmissible SCF to the deck transverses. 1. The stool design reduces the total amount of steel to be
Some important aspects of this support philosophy are installed above deck. Therefore, for lighter modules,
presented hereafter: where the deck transverse as originally designed can
1. The bulkheads transfers loads from topsides to the withstand the concentrated load from the stool, the
hull by two transverse supports. Therefore, the loads supports weights get optimized compared with the
are big, but distributed along the deck transverse. bulkhead supports. This option increase the number of
Therefore, only the ends of the transverse bulkheads local hot-spot stresses on the deck plating. the ends of
will face some hot spot stresses. the transverse bulkheads will face some hot spot
2. The idea of the bulkhead to support the topsides is to stresses.
reduce the relative displacement among the support 2. The stools conception allows the deck transverse to
points distributed in on frame, and also to eliminate the deflect a little , imposing some gap forces on the
racking displacements due to the bending of the hull module. Nevertheless, the hull will have reduced
girder that appears whenever there is more than two stresses installed on the deck transverse.
longitudinal supports. 3. The buckling is also the governing mode of failure in
3. The inclusion of bulkheads above deck, moves the this design concept, based on the significant vertical
neutral axis of the supports close to the deck plating. loads that are transferred from the topsides to the deck
Therefore, the additional local stress induced on the transverses.
deck plating due to the topsides are negligible.

3 Copyright © 2002 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 11/26/2014 Terms of Use: http://asme.org/terms


4. The fatigue analysis is performed for the longitudinal ORIGINAL STRESS DISTRIBUTION ON HULL
and transverse ends of the stools and also for the deck The first step on this work was to evaluate the stress
transverses just below the stool ends. distribution on the deck transverses for a typical VLCC tanker.
This analysis could prove the usage factor of these structures
The first evaluation on the stool design is the number regarding stresses and buckling. Based on the stress levels, it
of supports that would generate the lighter supports. Based can be foreseen some of the boundaries of the feasible design
on topside modules from 1000 t to 2000 t (operation options.
weight), some preliminary calculations were carried out, The structural analysis was based on ABS Rules for
considering static and dynamic loads for the topsides. Tankers (see reference 1). Therefore, the FE model analyzed
The dynamic inertia loads were generated based on has three tanks length, and the loadcases follows the SafeHull
regular wave approach and critical related periods as loading pattern.
described on reference 4. The hull transverses presents stress levels beyond 20% of
On figure 5, the horizontal axis presents the number of the admissible ones, and reaches the maximum of 60% at the
transverse supports considered; the vertical axis presents connection between the cross ties and the longitudinal bulkhead
the ration of the total amount of steel design as topside transverse. Figure 6 shows the von Mises usage factor on the
supports divided by the topside dry weight; and it was deck transverse at mid-tank.
analyzed from 2 rows of supports per topside modules (one
at each end of the module) up to 6 rows of supports (1 row
of supports at each hull frame – approx. 5.0m for VLCC). 0.87 0.755 0.64 0.526 0.411 0.296 0.182

From that, it can be concluded that the optimum


configuration for the stool option is 2 rows of supports and
3 supports in each row.

Total Weight Distribution (Above + Below Main Deck)

600

550
#5 Frames

500
Y
#4 Frames
450 Z X
Braced Options #2 Frames Output Set: ENVELOPE_MAXIMUM
Contour: vonMises_Usage_factor
400

#3 Frames
Figure 6 – von Mises Stress Distribution on tanker
350

Therefore, the deck transverses can easily withstand the


300 stress distribution imposed by the tanker loads.
0 5 10 15 20 25 30

Number of Supports
Moreover, the buckling failure was investigated analyzing
the normal and shear stress levels .Figures 7 to 9 below points
Figure 5 – Optimization of Steel Weight on Stools Design out the compressive stress distribution on the hull frame.

TYPICAL TOPSIDES DEFINITION


The optimization of the topside supports must be based on
the topside module weight. As this optimization is a numeric
flow calculation, the following data was taken as the typical
topside module according to PROJEMAR experience:
LT = 5 frames spacing ~ 25.0 m;
BT = 40% of the hull’s breadth ~ 22.0 m;
HT ~ 4.0 m;
WT = 1000 tons – module 1;
1400 tons – module 2;
1800 tons – module 3;
2200 tons – module 4.

Some different weights had to be analyzed in order to


establish deadlines for usage of each design conception. Figure 7 – X-normal buckling factor

4 Copyright © 2002 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 11/26/2014 Terms of Use: http://asme.org/terms


model. As the cargo tank model is a three tank length model,
with stress analysis available for the mid-tank, two bays of
0.992 0.827 0.662 0.496 0.331 0.166 0.00082
topside modules were modeled. as beams for the primary
structure and as non-structural distributed and concentrated
masses for all the other items.
The structural analysis of the FPSO hull followed then ABS
Rules for Floating Production Installations (see reference 2).

STRESS DISTRIBUTION ON HULL FOR TRUSS


SUPPORTS
The truss concept distributes the loads on several local
points of the deck plating. Due to this, the deck transverses
Y stress distribution remains similar to the tanker condition.
Z X Figure 10 below presents the von Mises stress results on the
Output Set: ENVELOPE_MAXIMUM
Contour: Y-Normal_Buckling deck transverse for the Topside Module 2200 tons. The stresses
Figure 8 – Y-normal buckling factor are normalized by the admissible criteria.
V1
L24
C1 1.03 1. 0.838 0.676 0.514 0.352 0.19
0.992 0.827 0.662 0.496 0.331 0.166 0.00082
G23

Y Y

Z X Z X
Output Set: ENVELOPE_MAXIMUM
Output Set: ENVELOPE_MAXIMUM
Contour: vonMises_Usage_factor
Contour: XY-shear_Buckling

Figure 9 – XY-shear buckling factor Figure 10 – von Mises Stress for Truss Supports

It must be clarified that the X-normal stress aligns with the However, the deck plating were loaded with hot-spot
vertical direction, and the Y-normal stress aligns with the stresses from the truss connections. Then, the critical stress
transverse direction. level raised from 80% to 100% of the admissible criteria.
Analyzing the buckling criteria, it is clear that the deck Considering also that the truss seats induced SCF about 2.0, all
transverse and cross ties ends can buckle during the buckling the welds will have to be ground to increase the fatigue
life. It is known that several 70’s VLCC’s had the cross-ties expected life of the connections.
buckled during the tanker life. Therefore, the tanker structural
analysis correctly predicted the possibility of buckling on the STRESS DISTRIBUTION ON HULL FOR STOOLS
cross ties ends. SUPPORTS
Then, the topside supports should can safely transfer The stool concept distributes some concentrated loads on
vertical loads (X-Normal Stress), which has a buckling usage two frames. Due to this, the deck transverses stress distribution
around 0.5. However, some reinforcement for buckling should receives great vertical loads , and the buckling limits is usually
be designed for the deck transverse ends , so that the girder can achieved only with some web girder. Figure 11 below presents
absorb the moment due to the topside load acting on its mid- von Mises stress results for the deck transverse for the topside
length. module 2200 tons. Here, the deck transverses should be
replaced by thicker ones.
FPSO STRUCTURAL ANALYSIS OF CARGO REGION
The FPSO structural analysis consists on the add of the
topsides modules and supports on the VLCC FEM. The tanker
model was altered to incorporate the topside module grillage

5 Copyright © 2002 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 11/26/2014 Terms of Use: http://asme.org/terms


V1
C1
G37 0.787 0.669 0.551 0.434 0.316 0.198 0.0804

V1
C1
G37 0.623 0.535 0.447 0.359 0.271 0.183 0.0947

Z X
Output Set: ENVELOPE_MAXIMUM
Y
Contour: Max_vonMises_Usage_factor

Figure 11 – von Mises Stress for Stools Supports Z X


Output Set: ENVELOPE
Contour: Max_vonMises_Usage_factor

Otherwise, the deck plating was loaded with hot-spot Figure 12 – von Mises Stress for Partial Bulkhead Supports
stresses from stools in only few places, reaching about 90% of
the admissible criteria for nominal stresses. Then, the fine mesh The deck plating has hot-spot stresses from bulkheads ends
analysis concluded on SCF’s for the stools connections to the lower than the truss and stools concept. The conclusion from
deck plating around 1.5, which is considerably lower than the this analysis is that only the longitudinal ends of the stools
truss concept. The conclusion from this analysis is that the should have the welds to the deck ground on an extension of
longitudinal and transverse ends (about 300 mm extension) of about 300 mm.
the stools should have the welds to the deck ground.
FINAL RESULTS AND COST COMAPRISON FOR
STRESS DISTRIBUTION ON HULL FOR PARTIAL DIFFERENT STUCTURAL CONCEPTS
BULKHEAD SUPPORTS
The partial bulkhead concept distributes the concentrated The added steel to support the topside modules are
loads on each of the two frames, throughout all the deck length. presented on table 1. It was considered that the topside support
Besides, the partial bulkhead act as extension on the deck must be adequately designed to maintain the topside module
transverse height, increasing the inertia and reducing the stress structure.
level on the deck transverses. The total amount of steel to be added to the hull is not the
Therefore, as shown on figure 12, the deck transverses most efficient parameter to be compared when deciding the
stress distribution receives great vertical loads and the buckling topside support concept to be adopted. The structural
limits are usually achieved only with some web girder. The von replacement on the deck transverses requires the work above
Mises stresses on the deck transverse for the topside module min deck to be delayed until all the cuts on the deck gets
2200 tons are considerably less than the stools and truss option. finished. This delay can also generates delays on the final
Here, the deck transverses must not be replaced. schedule for the conversion, raising the price to be paid on the
end of the job. Due to this, some taxes were obtained as
averages among some huge yards.

6 Copyright © 2002 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 11/26/2014 Terms of Use: http://asme.org/terms


TOPSIDE - 1000 tons TRUSS STOOL BHD TRUSS STOOL BHD
Topside Weight (tons)
Add above Main Deck 75.0 47.0 75.0 (tons) (tons) (tons)
Add below Main Deck 2.0 12.0 12.0 1000 tons - Topside 81.5 69.0 96.0
Replace below Main Deck 0.0 0.0 0.0 1400 tons – Topside 118.5 102.0 111.0
Weld Grinding 150.0 48.0 0.0 1800 tons – Topside 159.0 145.0 118.0
TOTAL 77.0 59.0 87.0 2200 tons – Topside 215.0 210.0 136.0
TOPSIDE – 1400 tons TRUSS STOOL BHD Table 2 – Final Cost Comparison
Add above Main Deck 105.0 55.0 80.0
CONCLUSIONS
Add below Main Deck 2.0 15.0 12.0
The final conclusion from this study is that the truss is
Replace below Main Deck 0.0 8.0 0.0 always a bad choice for heavy topsides (+ 1000 tons).
Weld Grinding 200.0 48.0 0.0 For topside modules up to 1500 tons of operation weight,
TOTAL 125.0 78.0 92.0 the best solution is the stools option. And, for topside modules
TOPSIDE – 1800 tons TRUSS STOOL BHD above 1500 tons , the best option is the partial bulkhead option.
Add above Main Deck 150.0 63.0 91.0 It must be stated here, that the partial bulkhead option
Add below Main Deck 2.5 21.0 15.0 generates the need for additional safety and fire-fighting
Replace below Main Deck 0.0 18.0 0.0 equipment. However, this cost is usually small when compared
Weld Grinding 250.0 48.0 24.0 to the time required for the under-deck structural replacement,
which differs from each yard.
TOTAL 185.0 102.0 106.0
The conclusion on the cost comparison was based on
TOPSIDE – 2200 tons TRUSS STOOL BHD theoretical yard cost for steel replacement, estimating man-
Add above Main Deck 187.0 73.0 109.0 hours and other costs for different levels of erection and
Add below Main Deck 13.0 24.0 15.0 welding works .
Replace below Main Deck 0.0 38.0 0.0 Some additional work is required to account for the
Weld Grinding (m) 250.0 48.0 24.0 equipment costs and erection costs for a couple of shipyards to
TOTAL 200.0 135.0 124.0 achieve final cost comparison.
Table 1 – Total Steel Reinforcement for 1 Topside Module
AKNOWLEDGMENTS
The estimates on the cost for the final built-up net steel on This study has been supported by PROJEMAR, with the
each region of the hull were normalized by the cost of the deep assistance of ABS Rio de Janeiro Office, regarding ABS
added steel above main deck. Rules and recommendations.
Added steel above main deck : 1
Added steel under main deck : 1.75
Replaced steel under main deck : 2.5
Nevertheless, the weld grinding is a controlled procedure,
which requires very specialized personnel to execute it. Then,
the high costs for this procedure were included in the cost
parameter as 2% of the steel added above main deck for each
meter of weld grinding.
REFERENCES
Therefore, the parameter used to compare the fabrication
[1] ABS 2001 Rules for Building and Classing Steel
costs for each option was the total estimated cost of the topside
Vessels, June 2001
support divided by the cost of one tons of steel added above
[2] ABS Guidelines for Building and Floating Production
main deck. The results are presented on table 2:
Installations, June 2000
[3] Almar-Naess, A., Fatigue Handbook – Offshore Steel
Structures, Tapir Publishers, 1995
[4] Faltinsen, O. M., Sea Loads on Ships and Offshore
Structures, Cambridge University Press, 1990

7 Copyright © 2002 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 11/26/2014 Terms of Use: http://asme.org/terms

You might also like