ABS TECHNICAL PAPERS 2008

SLAMMING IMPACT DESIGN LOADS ON LARGE HIGH SPEED NAVAL CRAFT Sungeun (Peter) Kim, Derek Novak American Bureau of Shipping, Houston, Texas, USA

Kenneth M. Weems Science Application International Corporation, Bowie, Maryland, USA

Hamn-Ching Chen Texas A&M University, College Station, Texas, USA International Conference on innovative approaches to further increase speed of fast marine vehicles, moving above, under and in water surface, SuperFAST’2008, July 2-4, 2008, Saint-Petersburg, Russia

ABSTRACT This paper presents the recent developments at ABS to revise the requirements for slamming impact loads on high speed naval craft. According to the ABS Guide for Building and Classing High Speed Naval Craft (HSNC 2007), slamming impact load is one of the most critical factors for the scantling design of hull structures. As modern naval craft design requires higher ship speed with increasing ship size, ABS is continuously putting efforts to refine prescriptive rules, analysis procedures and numerical tools. Recently, ABS investigated large high speed naval craft designs and proposed new requirements for slamming impact loads. This paper is mainly focused on the refinement of prescriptive rules for bottom slamming design pressure on mono-hulls and wetdeck slamming design pressure on the cross-structure of multi-hulls. Extensive numerical simulations were carried out using the nonlinear time domain seakeeping program LAMP. Vertical acceleration, impact forces and slamming pressures were calculated and compared with available model test data and design practices. This paper also presents ABS’s on-going efforts for the development and validation of computational fluid dynamics (CFD) code as an alternative numerical tool to analyze the extremely violent nonlinear free-surface flows such as sloshing, slamming and green water impact problem. Some of the most recent CFD simulation results are presented including the wetdeck slamming of a catamaran using the level-set Finite-Analytic Navier-Stokes (FANS) code. 1. Design Conditions INTRODUCTION As the new generation of high speed naval craft becomes larger and faster, slamming impact loads on these vessels are a critical design concern. Currently available rules for slamming design pressure were mostly developed for small planing hulls based on experimental and theoretical work undertaken in the 60’s and 70’s [2, 3]. To revise the current design criteria specified in the ABS Guide for Building and Classing High Speed Nacal Craft (HSNC) [1], ABS recently carried out extensive numerical analysis for the new designs of high speed naval craft As testing vessels, high speed naval craft of large semi-planing mono-hull, small planing mono-hull, displacement mono-hull, and wave-piercing catamaran are considered, and the state-of-the-art nonlinear seakeeping program LAMP is used for numerical simulation.

Slamming Impact Design Loads on Large High Speed Naval Craft

1.1. Design sea states and ship speeds For the determination of design loads on high speed naval craft, design sea states are to be defined by significant wave height, wave modal period and ship speed. Table 1 shows an example of the sea states and significant wave heights typically being used for the design of naval craft operating in the North Atlantic. Vm denotes the maximum desgin speed. Table 1: Sea sates in North Atalntic Sea State Hs (m) V(knots) 2 0.5 Vm 3 1.25 Vm 4 2.5 Vm 5 4 Vm 6 6 10 7 9 10 8 14 10

207

ABS TECHNICAL PAPERS 2008 Table 2 shows the design sea states and ship speeds defined in ABS HSNC guides (3-2-2/Table 1). Note that the significant wave height of survival condition is not to be taken less than L/12. Vm denotes the maximum speed for the craft in the design condition, and the ship speed of 10 knots in survival condition is to be verified by the Naval Administration. Table 2: Design sea sates and ship speed Operational Survival Condition Condition h1/3 V h1/3 V Naval Craft 4 m Vm 6 m 10 knots Coastal 2.5 m Vm 4 m 10 knots Naval Craft Riverine 0.5 m Vm 1.25 m 10 knots Naval Craft 1.2. Tested vessels This study is mainly focused on the slamming impact loads on high speed naval craft. In this study, two semiplaning mono-hulls and one displacement mono-hull and one planing mono-hull are considered for bottom slamming design pressure. Also a wave-piercing high speed catamaran is considered for wetdeck slamming design pressure.

⎡12h ⎤ V 2 Bw2 ncg = N 2 ⎢ 1/ 3 + 1.0⎥τ 50 − β cg Δ ⎣ Bw ⎦

[

]

(1)

where ncg average of the 1/100 highest vertical acceleration at LCG in g’s N2 0.0078 h1/3 1/3 highest significant wave height, in m, as given in Table 2 Bw maximum waterline beam, in m βcg deadrise angle at LCG, in degrees, not to be taken less than 10o nor more than 30o V design speed in operation and survival conditions, in knots, as given in Table 2 Δ displacement in kg τ running trim angle at V, in degrees, Note that the vertical acceleration in equation (1) is a function of running trim angle of the hull. At the early stages of design, however, the running trim angle is not known a priori. Furthermore, the running trim angle of the vessel in a seaway varies in time and is not much relevant to the vertical acceleration of the vessel operating in severe sea states. Based on the numerical study presented in Section 3, the empirical formula for vertical acceleration is revised as follows.

CAT-2 MONO-1 CAT-1 MONO-2 MONO-4

MONO-3

Fig.1: Typical high speed naval vessels

2. BOTTOM SLAMMING DESIGN PRESSURE FOR MONO-HULLS 2.1. Vertical acceleration According to the Heller and Jasper [2], bottom slamming design pressure on planing hull can be expressed in terms of vertical acceleration of the craft. The vertical acceleration is to be determined by a model test or theoretical computation. If this information is not readily available during the early stages of design, the following formula may be used. Based on the experimental study of seakeeping performance of planing boats [4, 5], Savitsky and Brown presented an empirical formula for the vertical acceleration of planing hull in a seaway [3].

208

⎡12h ⎤ VB w2 n cg = 35 N 2 CV ⎢ 1 / 3 + 1.0⎥ 50 − β cg Δ ⎣ Bw ⎦ where

[

]

(2)

ncg average of the 1/100 highest vertical acceleration at LCG in g’s CV = 0.657 L − 2.5 , not to be taken less than 1.5 nor more than 5 The vertical acceleration at any section of the hull along the ship length may be expressed as below.

n xx = ncg K v

(3)

where nxx KV

average of the 1/100 highest vertical acceleration at any section, in g’s vertical acceleration distribution factor

A revision of the vertical acceleration distribution factor KV is proposed, as given in Fig. 2. The factor has been significantly increased for survival condition, based on the numerical simulation results presented in Section 3.

Slamming Impact Design Loads on Large High Speed Naval Craft

ABS TECHNICAL PAPERS 2008

Vertical Acceleration Distribution Factor Kv

Longitudinal Pressure Distribution Factor F_L

8

1.2

Current Survival Condition

0.8 Fv

Kv

1

Operation Condition

6

4

0.6 0.4

2

0.2 0

0 0

0.2

0.4

0.6

0.8

1

X/L from AP

Fig. 2: Vertical acceleration distribution factor 2.2. Bottom Slamming Pressure In the early 1960s, for small planing craft, Heller and Jasper suggested a slamming pressure formula in a very concise form[3]. Based on the numerical analysis in this study, the current slamming design pressure given in HSNC has been validated first and then a revised slamming design pressure is proposed as below. For semi-planing hull, the slamming design presurre can be expressed as:

p bxx =

70 − β bx N 1Δ [1 + n xx ][ ]FL Lw Bw 70 − β cg

(4)

where pbxx bottom design pressure at any section, in kN/m2 N1 0.01 Δ displacement in kg Lw craft length on the waterline, in m Bw maximum waterline beam, in m βbx deadrise angle at any section, in degrees, not to be taken less than 10o nor more than 30o FL longitudinal pressure distribution factor In the proposed formula, longitudinal pressure distribution factor FL is introduced to consider 3D flow effects near the bow and stern area, as shown in Fig. 3. For semi-planing hull, the slamming design presurre may be simplified as:

p bxx =

N 1Δ [1 + n cg ]FV Lw Bw

0

0.2

0.4 0.6 X/L from AP

0.8

1

Fig. 3: Longitudinal pressure distribution factor 3. BOTTOM SLAMMING SIMULATION FOR MONO-HULLS 3.1. LAMP method The Large Motion Amplitude Program (LAMP) is a 3D panel method developed by SAIC for the time-domain simulation of nonlinear ship motions and wave loads in extreme wave conditions. LAMP development started as a DARPA project in 1988 and has been supported by the US Navy, USCG, SAIC/MIT and ABS [12]. 3.2 MONO-1 hull MONO-1 is a large high speed semi-planing naval craft, recently built to ABS Class in accordance with the ABS Guides for Building and Classing High Speed Naval Craft (HSNC). The overall length of the craft is more than 110m and loading conditions are as given in Table 3. Table 3: Loading conditions of MONO-1 Loading Condition Displacement Speed (tons) (knots) Full Load Departure 3100 38 Full Load Arrival 2900 41 Full Load Minimum 2800 42 Operation Full Load Survival 3100 10 Fig. 4 shows the LAMP nonlinear geometry model of MONO-1 that includes the hull geometry above the mean waterline. Nonlinear hydrostatic restoring and FroudeKrylov forces acting on the instantaneous wetted hull surface are calculated over the nonlinear geometry model.

(5)

where FV

vertical acceleration distribution factor, defined in HSNC 3-2-2/Figure 8 Fig. 4: LAMP nonlinear geometry model Fig. 5 shows the LAMP linear hydro panel model of MONO-1. A large number of quadrilateral or triangular panels are distributed on the hull surface below the mean waterline as well as on the truncated free surface. This

Slamming Impact Design Loads on Large High Speed Naval Craft

209

ABS TECHNICAL PAPERS 2008 model is used to calculate radiation and diffraction forces due to linear wave-body interaction. Once the geometry models were prepared, nonlinear time–domain seakeeping analysis was carried out using LAMP. Fig. 6 shows the time history of vertical acceleration calculated at x=90m from AP.

design factor for high speed craft. The figure presents the actual design pressures that were actually used for the scantling check of bottom plates. The actual design pressures were determined by the vertical acceleration at LCG directly measured from model tests. The proposed slamming pressure in this study is very close to the actual design pressure of MONO-1. 1/100th Vertical Accel at Full Load Departure 2 Current LAMP Model Test Proposed

VACC (g)

1.5

1

0.5

Fig. 5: LAMP linear hydro panel model

0 0

0.2

0.4

0.6

0.8

1

X from AP

Vertical Acc. at x=90m from AP 15

Fig. 7: Vertical acceleration at operational condition: full departure with V=38knots and Hs=4m

VCAA(m/s^2)

10 5 0 -5

1/100th Vertical Acc. of LCS at Full Load Survival

-10 -15

2 0

200

400

600

800

1000

1200

Current

t(s)

Fig. 8 shows the 1/100 highest vertical acceleration of MONO-1 at survival condition: full load departure with ship speed V=10knots and significant wave height Hs=9m. The proposed vertical acceleration shows much better agreement with LAMP results. Fig. 9 shows the bottom slamming design pressure distribution along the ship length of MONO-1 at operational condition. The main difference between the current and proposed design pressure is coming from vertical acceleration, which is known as the most critical

210

Proposed

1

0.5

0 0

0.2

0.4

0.6

0.8

1

X from AP

Fig. 8: Vertical acceleration at survival condition: full load departure with V=10knots and Hs=10m Bottom Slamming Pressure at Full Load Departure 500 400 P (kPa)

Fig 7 shows the 1/100 highest vertical acceleration of MONO-1 at operational condition: full load departure with ship speed V=38knots and significant wave height Hs=4m. Compared with model test measurements (green dot) and LAMP simulation results (pink quare), the proposed vertical acceleration (red line with triangle) has been significantly improved.

VACC (g)

Fig. 6: LAMP prediction of time history of vertical acceleration at x=90m from AP. As given in Eq. (2), the slamming design pressure is expressed in terms of the average of 1/100 highest vertical acceleration in a design sea state under consideration. Once the time series of vertical acceleration are obtained from seakeeping analysis in time domain, peak analysis is to be performed to estimate the 1/100 highest vertical acceleration.

LAMP

1.5

300 200 100 0 0

0.2

0.4

0.6

0.8

1

X from AP Current

Proposed

Design

Fig. 9: Bottom slamming design pressure at operational condition: full load departure Fig. 10 shows the slamming design pressure on MONO-1 at survival condition. It can be seen that the slamming

Slamming Impact Design Loads on Large High Speed Naval Craft

ABS TECHNICAL PAPERS 2008 pressure has been significantly increased by the proposed formula in better agreement with actual design pressure.

response, the first 1/5 of time series is ignored in the peak counting. x=20 from AP

Bottom Slamming Pressure at Survival condition 1500

500

LAMP Current Proposed

w(kN/m)

P (kPa)

400 300 200

1000

500

100 0 0

0.2

0.4

0.6

0.8

0

1

0

X from AP Current

Proposed

600 t(s)

800

1000

1200

800

1000

1200

w B/6

800

1000

1200

x=50 from AP 1500 LAMP Current Proposed

w(kN/m)

Numerical calculation of slamming impact pressure using the 2D boundary element method can be a very challenging and time-consuming task. Instead, direct calculation of vertical impact forces on each sectional cuts using wedge approximation is considered in this study, which is a more efficient and reliable method. As a design practice, the slamming pressure may be estimated from sectional impact force as follows:

1000

500

0 0

200

(6)

Fig. 11 shows the time history of sectional impact forces simulated by LAMP/LMPOUND, calculated at three representative sections x= 20m, 50m and 90m from AP at operational condition. The current and proposed rule values are also compared in the figure. Note that the relative ship motion and acceleration is expected high at x=90m from AP, but the impact force was actually low because of the high deadrise angle of that section. In general, the proposed impact forces have been significantly increased because of the increased vertical acceleration. Peak analysis is required for the representation of statistical properties of the response in time series, either calculated by numerical simulation or measured by model tests. In this study, only a highest peak is counted between zero-crossings. No intermediate peaks are counted as independent events. To eliminate the noise in time series, only those peaks exceeding a threshold value are counted. In this studt, 10% of the average of the 1/100 highest peak is used as threshold value. To avoid transient

Slamming Impact Design Loads on Large High Speed Naval Craft

600 t(s)

x=90 from AP LAMP Current Proposed

w(kN/m)

slamming pressure, in N/m2 sectional impact force, in N/m maximum beam, in m

400

1500

where p w B

400

Design

Fig. 10: Bottom slamming design pressure at survival condition: full load departure

p=

200

1000

500

0 0

200

400

600 t(s)

Fig. 11: Vertical impact force at operational condition Those counted peaks are to be represented by relevant probability distribution functions. In this study, twoparameter Weibull distribution is considered as follows:

Qj ( x > x ) = exp(−

xβ

θβ

)

(7)

Fig. 12 shows the peak analysis results of vertical impact forces using Weibul distribution. Note that the response at a exceedance probability level of 0.001 is a typical target value that corresponds to the most probable short-term extreme value out of 1000 encountered wave cycles.

211

ABS TECHNICAL PAPERS 2008

Fig. 13 shows the LAMP geometry model of MONO-2 that includes the hull geometry above the mean waterline.

x=20m from AP 1 LAMP Weibull Current Proposed

Q

0.1 0.01 0.001 0.0001 0

500

1000

1500

Fig. 13: LAMP geometry model

w(kN/m)

Fig. 14 shows the vertical acceleration of MONO-2 in operational condition. Compared with LAMP results, the proposed vertical acceleration has been increased in the stern area. Fig. 15 shows the vertical acceleration in survival condition. It can be seen that the proposed vertical acceleration has been significantly increased.

x=50m from AP LAMP

1

Weibull Current

0.1 Q

Proposed

0.01 0.001

1/100th Vertical Acc. at Operational Condition 2

0.0001

Current

0

500

1000

1500

VACC (g)

w(kN/m)

x=90m from AP 1

Proposed

1

0.5

LAMP Weibull Current Proposed

0.1 Q

LAMP

1.5

0 0

0.2

0.01

0.4

0.6

0.8

1

X from AP

Fig. 14: Vertical acceleration at operational condition: full load at V=35knots and Hs=4m

0.001 0.0001 0

500

1000

1500

1/100th Vertical Acc. at Survival Condition

w(kN/m) 2

Fig. 12: Peak analysis of vertical impact force at operational condition

MONO-2 is a medium-size high speed semi-planing naval craft recently built to ABS Class in accordance with the ABS Guides for Building and Classing High Speed Naval Craft (HSNC). The overall length of the craft is about 60m and loading conditions are as given in Table 4 Table 4: Loading conditions of MONO-2 Loading Condition Displacement Speed (tons) (knots) Full Load Departure 750 35 Full Load Survival 750 10

212

VACC (g)

3.3. MONO-2 hull

Current LAMP

1.5

Proposed

1

0.5

0 0

0.2

0.4

0.6

0.8

1

X from AP

Fig. 15: Vertical acceleration at survival condition: full load at V=10knots and Hs=6m

Fig. 16 shows the slamming design pressure of MONO-2 at operational condition. Note that the proposed design pressure has been significantly increased in the bow and stern areas.

Slamming Impact Design Loads on Large High Speed Naval Craft

ABS TECHNICAL PAPERS 2008

Bottom Pressure at Operational Condition

1/100th Vertical Acc. at Survival Condition

400

2 Current LAMP

1.5

VACC (g)

P (kPa)

300 200 100

Proposed

1

0.5 0 0

0.2

0.4

0.6

0.8

1

0

X from AP Current

0

Proposed

0.2

Design

0.4

0.6

0.8

1

X from AP

Fig. 16: Bottom slamming design pressure at operational condition: full load departure

MONO-3 is a displacement-type naval craft recently built to ABS Class. The overall length of the craft is about 100m and loading conditions are as given in Table 5. Table 5: Loading conditions of MONO-3 Loading Condition Displacement Speed (tons) (knots) Full Load Departure 2200 25 Full Load Survival 2200 10

Fig. 20 and 21 shows the slamming design pressure of MONO-3 at operational and survival condition, respectively. Note that the proposed design pressure has been consistently increased along the ship length. Bottom Pressure at Operational Condition 500 400 P (kPa)

3.4. MONO-3 hull

Fig. 19: Vertical acceleration at survival condition

300 200 100

Fig. 17 shows the LAMP geometry model of MONO-3 that includes the hull geometry above the mean waterline.

0 0

0.2

0.4

0.6

0.8

1

X from AP

Current

Proposed

Fig. 20: Bottom slamming design pressure at operational condition Bottom Slamming Pressure at Survival condition 500

Fig. 17: LAMP geometry model

Fig. 18 shows the vertical acceleration of MONO-3 in operational condition. Fig. 19 shows the vertical acceleration in survival condition. Note that MONO-3 is a displacement vessesl and has smaller vertical acceleration than the semi-planing vessels of MONO-1 and MONO-2. The proposed vertical acceleration shows good agreement with LAMP simulation results.

P (kPa)

400 300 200 100 0 0

0.2

0.4

0.6

0.8

1

X from AP Current

Proposed

Fig. 21: Bottom slamming design pressure at survival condition

1/100th Vertical Acc. at Operational Condition 2

Current

VACC (g)

3.5. MONO-4 hull

LAMP

1.5

Proposed

MONO-4 is a typical high speed planing naval craft. The overall length of the craft is about 25m and loading conditions are as given in Table 6.

1

0.5

0 0

0.2

0.4

0.6

0.8

1

X from AP

Fig. 18: Vertical acceleration at operational condition

Slamming Impact Design Loads on Large High Speed Naval Craft

Table 6: Loading conditions of MONO-4 Loading Condition Displacement Speed (tons) (knots) Full Load Departure 75 30 Full Load Survival 75 10

213

ABS TECHNICAL PAPERS 2008

at operational condition

Fig. 24 shows the slamming design pressure on planing hull of MONO-4 at operational condition, calculated by eq. (4). Note that the proposed design pressure has been reduced due to the reduction of vertical acceleration in operational condition, as given in Fig.22. Fig. 25 shows the slamming design pressure on planing hull of MONO-4 at survival condition. 1/100th Vertical Acc. at Operational Condition

Current Proposed

VACC (g)

4

2

0 0.2

0.4

0.6

0.8

1

X from AP

Fig. 22: Vertical acceleration at operational condition 1/100th Vertical Acc. at Survival Condition 6 Current Proposed

VACC (g)

300 200 100 0 0

0.2

0.4

0.6

0.8

1

X from AP Current

Porposed

Fig. 25: Bottom slamming design pressure at survival condition 4. WETDECK SLAMMING DESIGN PRESSURE FOR MULTI-HULLS

The underside of wet deck or cross structure of multihulls is likely subject to significant slamming impact loads in severe sea states. In this study, the current wetdeck slamming design pressure given in HSNC is investigated for a latest high speed catamaran.

6

0

Bottom Pressure at Survival Condition 400

P (kPa)

Fig. 22 shows the vertical acceleration of MONO-4 in operational condition. Note that MONO-4 is a small planing hull and higher vertical acceleration is expected, but the proposed vertical acceleraction is slightly reduced compared to the current vertical acceleration which is believed to be overpredicted. Fig. 23 shows the vertical acceleration in survival condition. The prosposed vertical acceleration has been significantly increased.

4

Numerical simulation of wetdeck slamming impact loads can be a very challenging and time-consuming task. Recently, a very efficient wetdeck slamming module has been implemented into the LAMP system based on a 2D longitudinal cut model [10]. LAMP wetdeck module is used for the prediction of wetdeck slamming pressure of a catamaran operating in design operational condition and survival condition. According to the LAMP simulation results, a revised wetdeck slamming design pressure is proposed as follows:

p wd = 30 FI VV I (1 − 0.5h a / h1 / 3 )

2

(8)

where 0 0

0.2

0.4

0.6

0.8

1

X from AP

Fig. 23: Vertical acceleration at survival condition Bottom Pressure at Operational Condition

FI wet deck pressure distribution factor as given in Fig 26 V design speed in operation and survival conditions, in knots, as given in Table 2 VI relative impact velocity as given below

=

400

P (kPa)

300

ha

200

h1/3

100

4h1 / 3

+ 1 (m / s) L vertical distance, in m, from waterline to underside of wetdeck significant wave height, in m, as given in Table 2

0 0

0.2

0.4

0.6

0.8

1

X from AP Current

Proposed

Fig. 24: Bottom slamming design pressure

214

Slamming Impact Design Loads on Large High Speed Naval Craft

ABS TECHNICAL PAPERS 2008

In the comparison to the slamming model test data for CAT-1, it was found that additional pitch damping is needed for an accurate LAMP simulation of the pitch motion in the most severe slamming conditions. In order to match with measured pitch motion, the supplementary pitch damping model is considered in LAMP simulations as follows:

Wet Deck Pressure Distribution Factor F_I 2 Current Proposed: Operational Proposed: Survival

F_I

1.5

1

0.5

2

ME5 = −v5 * KL5 − v5 *

0 0

0.2

0.4

0.6

0.8

1

v5 * KQ5 v5

where

X from AP

Fig. 26: Vertical acceleration at operational

5. WETDECK SLAMMING SIMULATION FOR CATAMARAN 5.1 Test Vessel: CAT-1

CAT-1 is an aluminum wave-piercing high speed catamaran, which is a US Navy research vessel built to ABS class. Recently wetdeck slamming events were reported during full-scale trials.

v5 KL5 KQ5

pitch velocity linear pitch moment coefficient quadratic pitch moment coefficient

Using this added pitch damping model with coefficient tuned based on the pitch response near resonance, a very good agreement between the predicted and measured ship motions was achieved Fig. 28 shows the sensor locations of pressure patches used in the slamming model test of CAT-1 performed by DTMB [11].

The overall length of vessel is about 70m and loading conditions are given in Table 7. Table 7 Loading conditions Loading Displacement Condition (tons) Full Load 960 Departure Full Load 960 Survival

P1 Speed (knots) 40 10

5.2 LAMP Simulations A number of LAMP studies have been made for the CAT1 catamaran, including validation studies versus data from both model tests and full-scale trials [11].

Fig. 27 shows the LAMP geometry model of CAT-1. For the evaluation of wetdeck slamming in the simulations, the underdeck surface is discretized by quadrilateral panels along with 8 longitudinal cuts where the impact loads are calculated.

Fig. 27: LAMP geometry model of CAT-1

Slamming Impact Design Loads on Large High Speed Naval Craft

Fig. 28: Sensor location of presure patches used in model test of CAT-1 Fig. 29 shows a typical wetdeck slamming impact pressure at four sensor locations of CAT-1 in a model test condition wih regular waves. The wetdeck impact pressure consists of the first sharp peak with short duration due to slamming impact followed by a second round peak with longer duration due to nonlinear hydrostatic and Froude-Krylov forces.

Fig. 30 shows the time history of wetdeck slamming impact pressure of CAT-1 calculated at a pressure location P1, as shown in Fig. 28. The vessel is operating in operational condition with ship speed V=40knots and significant wave height Hs=4m. A significant increase of wetdeck slamming presure is proposed based on LAMP simulation results. Fig. 31 shows the peak analysis results of wetdeck slamming pressure at P1 using Weibul distribution, comparing with current and proposed wetdeck slamming design pressure. Note that the exceedance probability level of 0.001 corresponds to the typical most probable short-term extreme value.

215

ABS TECHNICAL PAPERS 2008

Wet-Deck Slamming Pressure at P1 200000

150000

p(Pa)

Fig. 32 shows the time history of wetdeck slamming impact pressure at the pressure location P1 in survival condition with ship speed V=10knots and significant wave height Hs=9m. Note that, in this study of the CAT1 vessel, the wetdeck slamming pressure of survival condition is found to be more severe than that of operational condition, which is likely due to the higher chance of underdeck wetness in the severe sea state of Hs=9m.

LAMP LAMP

100000

Current Proposed

50000

0 0

100

200

300

400

500

600

700

800

900

t(s)

In 2006, full-scale trials measurements of bow accelerations of CAT-1 have been made in severe sea states. Futher study will be carried out for the validation of LAMP wetdeck simulations using full-scale measured data.

Fig. 32: Wetdeck slamming pressure at location P1 in survival condition Wet-Deck Slamming Pressure at P1 1.0000

0.1000

LAMP

Q

Fig. 33 shows the peak analysis results of wetdeck slamming pressure at P1 in survival condition.

Weibull

0.0100

Current Proposed

0.0010

Wet-Deck Slamming Pressure 40000

0.0001

P1

30000

0

P2

P(Pa)

100000

150000

200000

Fig. 33: Wetdeck slamming pressure at location P1 in survival condition

P4

20000

50000

p(Pa)

P3

10000 0 7

8

8

9

9

10

10

t(s)

Fig. 29: Typical wetdeck slamming pressure of CAT-1 in regular waves Wet-Deck Slamming Pressure at P1 200000

150000

LAMP

p(Pa)

LAMP 100000

The level-set Finite-Analytic Navier-Stokes (FANS) code is a CFD two-phase flow solver with multi-block oversetgrid scheme, developed for highly nonlinear wave flows around ships and offshore structures [7, 8].

Current Proposed

50000

0 0

100

200

300

400

500

600

700

800

900

t(s)

Fig. 30: Wetdeck slamming pressure at location P1 in operational condition Wet-Deck Slamming Pressure at P1 1.0000

0.1000

LAMP

Q

5.3. CFD Simulations ABS is putting significant efforts for the development and validation of computational fluid dynamics (CFD) code as an alternative numerical tool to deal with extremely violent nonlinear wave flows such as sloshing, slamming and green water impact problem.

Weibull

0.0100

Recently, the level-set FANS method has been successfully validated for the prediction of sloshing impact pressure on tank boundary by comparing with sloshing model test data [9]. Fig. 34 shows an example of FANS simulated flow inside a LNG tank at a low filling condition with a prescribed transverse motion. Currently, ABS is working with Texas A&M to implement a compressible gas model into FANS code in order to more accurately predict sloshing and slamming impact pressure considering air cushion effect on trapped air. ABS is also building up a high performance computing system using PC cluster for CFD simulation.

Current Proposed

0.0010

0.0001 0

50000

100000

150000

200000

p(Pa)

Fig. 31: Wetdeck slamming pressure at location P1 in operational condition

216

Slamming Impact Design Loads on Large High Speed Naval Craft

ABS TECHNICAL PAPERS 2008

Fig. 34: Sloshing of a low filling tank in a transverse motion

FANS code was also used for the wetdeck slamming simulation of the catamaran CAT-1. Fig. 35 shows the overset moving grid system of FANS geometry model with 25 blocks, 2.2 million nodes on half domain (y>0). The simulations were performed using 16 processors on a Linux cluster

Fig. 35: FANS geometry model of CAT-1

Figures 36-38 show an example of a wetdeck slamming event of CAT-1 in regular waves. The catamaran is towed at a constant forward speed with Froude number Fr=0.3. It is allowed to heave and pitch freely in waves as shown in Figure 36. The incident wave length to ship length ratio (λ/L) is 1.0 and the wave amplitude to ship length ratio is H/L = 0.04. The heave displacement and pitch angle of the catamaran were obtained by solving the following two degree-of-freedom motion equations using the 4th-order Runge-Kutta method: ⎧⎪mz&& = Fz ⎨ && 2 ⎪⎩ I yyθ y = M y ; I yy = mryy

It is seen from the free surface patterns in Fig. 36 and the wave elevation contours in Fig. 37 that the level-set FANS method has successfully predicted highly nonlinear waves in front of the bow and along the CAT-1 hulls. Due to the large amplitude heave and pitch motions, wet deck slamming (denoted by the blue color on the lower deck in Fig. 38) was observed when the wave height between the catamaran hulls exceeded the lower deck clearance. A detailed examination of the animation movie indicates that the ship bow was lifted upward by the incident wave with the bulbous bow partially emerged out of the water prior to the slamming impact. Wet deck slamming was observed around the wave peak region when the bow plunges into the water under combined heave and pitch motions.

Fig. 36: Heave and pitch motions of CAT-1 in regular waves

Fig. 37: Wave elevation contours

Slamming Impact Design Loads on Large High Speed Naval Craft

217

ABS TECHNICAL PAPERS 2008

REFERENCES

1. ABS, Rules for Building and Classing High Speed Naval Craft, 2007 2. D. Savitsky and P. W. Brown, Procedures for hydrodynamic evaluation of planing hulls in smooth and rough water, Marine Technology, Vol. 13, No. 4, 1976 3. S. R. Heller and N.H. Jasper, On the structural design of planing craft, 1960 4. G. Fridsma, A systematic study of the rough-water performance of planing boats, Davidson Laboratory Report 1945, Stevensen Institute of Technology, 1969

Fig. 38: Wetdeck slamming event of CAT-1 in regular waves CONCLUSIONS

In this paper, slamming impact loads have been studied for the high speed naval craft recently built to ABS class. Extensive nonlinear seakeeping analyses were carried out to calculate vertical accelerations and sectional impact forces along the ship length. Based on LAMP simulation results, validated by model test measurements and design practice, new requirements for bottom slamming design pressure on mono-hulls and wetdeck slamming design pressure on multi-hulls were proposed. This paper also presents ABS’s on-going efforts for the development and validation of CFD methods to predict fully nonlinear 3D impact loads. The level-set FANS method successfully demonstrates the nonlinear free surface capturing capability for the benchmarking cases of sloshing and wetdeck slamming analyses. To minimize computational time, we are considering the coupling of LAMP and FANS codes. FANS can be used for nonlinear viscous flow near the ship while LAMP is used for linear potential flow away from the ship. This coupled analysis method will be challenging, but also very promising for the simulations of slamming and green water impact loads, sloshing impact loads coupled with ship motion and damaged stability of vessels in waves.

5. G. Fridsma, A systematic study of the rough-water performance of planing boats, irregular waves, Part II, Davidson Laboratory Report 1945, Stevensen Institute of Technology, 1971 6. R. G. Allen and R. R. Jones, A simplified method for determining structural design-limit pressures on high performance marine vehicles, AIAA, 1978 7. H-C. Chen and K. Yu, Numerical simulation of wave runup and greenwater on offshore structures by a level-set RANS method, 16th ISOPE, Vol III, pp. 185-192, San Francisco, 2006 8. H-C Chen and K. Yu, CFD simulations of wavecurrent-body interactions including green water and wetdeck slamming, FLUCOME, Florida, 2007 9. H-C. Chen and K. Yu, Chimera FANS simulation of sloshing impact pressure inside a LNG tank: Benchmark test and code validation, ABS Report, 2007 10. W. M. Lin, H-C. Chen and S. Zhang, Development and evaluation of nonlinear numerical methods for wetdeck slamming of a high speed catamaran, Procedding of international conference on violent flows, 2007

ACKNOWLEDGMENTS

11. W. M. Lin, S. Zhang, K.M. Weems, M.J. Meinholds, B. Metcalf, and A.M. Powers, Numerical simulation and validation study of wetdeck slamming on high speed catamaran, Proceedings of 9th International Conferecne on Numerical Ship Hydrodynamics, Ann Arbor, Michigan, 2007

Since the 1988 DARPA project, the LAMP system has been developed for the advanced ship motion simulation under the sponsorships of the US Navy, the US Coast Guard, and the American Bureau of Shipping. The authors would like to thank Dr. L. Patrick Purtell of ONR for his continuous supports for LAMP development.

12. Shin, Y.-S., Belenky, V. L., Lin, W.-M., Weems, K. M., Belknap, W. F., & Engle, A. H. ‘Nonlinear Time Domain Simulation Technology for Seakeeping and Wave Load Analysis for Modern Ship Design.’ Transactions of the Society of Naval Architects and 111, pp. 557-578, 2003 Marine Engineers

218

Slamming Impact Design Loads on Large High Speed Naval Craft