Sunteți pe pagina 1din 8

Marine Structures 16 (2003) 175182

A comparison of hydrodynamic impacts prediction methods with two dimensional drop test data
Allen Engle*, Richard Lewis
Naval Surface Warfare Center, Carderock Division, USA

Abstract The effect of hydrodynamic impacts on a ships bow structure is an important design consideration. In addition, to possible failure to local structure, such impacts can generate slam-induced whipping loads, the magnitude of which when combined with ordinary wave induced loads, could lead to rst passage failure of a ships hull girder. This paper presents the results of a recent study that compares select methods of predicting hydrodynamic impacts with available test data of 2-D wedge shaped section shapes. Results based on peak pressures at different impact velocities are presented. The results are compared with traditional prediction theories of Wagner and Chuang. Crown Copyright r 2002 Published by Elsevier Science Ltd. All rights reserved.

1. Introduction Traditional ship structural design criteria utilizes a standard wave for determining primary stresses. This standard wave approach determines the design bending moment by statically balancing the ship on a trichoidal wave whose length and height is typically equal to some percentage of the length of the ship being designed. The stresses derived from this bending moment are then compared with allowable stresses and adjusted on a trial-and-error basis to reect past experiences with ships already in operation. While this approach has served the shipbuilding industry well, this methodology does not explicitly account for the effects of slam-induced whipping of the hull girder or hydrodynamic impacts on local structure. Furthermore, the available design procedures do not explicitly account for the effects of fatigue or the role that
*Corresponding author. E-mail address: webmaster@crane.navy.mil (A. Engle). 0951-8339/03/$ - see front matter Crown Copyright r 2002 Published by Elsevier Science Ltd. All rights reserved. PII: S 0 9 5 1 - 8 3 3 9 ( 0 2 ) 0 0 0 2 6 - 6

176

A. Engle, R. Lewis / Marine Structures 16 (2003) 175182

impulsive loads have on the longitudinal distribution of bending moments. In addition, lateral loads, torsion and associated effects are not addressed. As a result of these uncertainties, designers are typically forced to apply generous design margins, particularly at stations forward of midships to account for the effects of slamming. With the advent of new ship design concepts such as SWATHs and trimarans as well as the application of new materials, the shipbuilding industry is moving towards a design process that relies on rst principles and reliability-based design procedures. However, in order to support such a transition, designers must be able to rigorously predict the hydrodynamic loads of a ship operating in sea conditions ranging from mild to extreme seas. Over the past decade, advances in computational ship hydrodynamics have resulted in increasingly capable and accurate computer codes for the prediction of ship motions and loads. The application of these codes has been accelerated over the past few years by the ever-increasing power of modern computers, so that some of these advanced numerical calculations may now be done within design time scales. As a result of these advances, a new level of computational capability is now emerging for the prediction of the non-linear ship motions and wave loads for severe sea conditions. For example, non-linear time domain analysis programs, such as the Large Amplitude and Motions Program (LAMP) [10,11] or Ship Wave ANalysis (SWAN) [12], have the potential to deliver a rigorous method of analysis by satisfying the body boundary condition on the instantaneous wetted surface at each time step. However, since these codes by themselves address only the ordinary wave-induced portion of the possible total load, a separate computation for hydrodynamic impacts must be performed. Initial efforts to predict such hydrodynamic impacts were examined by Von Karman [13] and Wagner. However, these approaches aimed at asymptotic global solutions. As such these methods do not provide the required level of detail to fully describe the ow eld associated with impacts. More recent theories, such as [1] or [2] are also limited in that they require ow similarity. This results in so-called expanding wedge theories as the wedge deadrise angle decreases the predicted impact pressure becomes innite. Consequently, these theories have been found to be limited when applied to engineering applications. Thus, what is needed is a practical prediction method that is rigorous enough in its treatment that it can support general engineering applications. In this paper, an assessment of select methods for predicting such hydrodynamic impact loads has been performed. This assessment is applicable to hydrodynamic pressures on rigid structures only; local structural exibility is not addressed. The results are compared with traditional prediction theories of Wagner and Chuang.

2. Hydrodynamic impact study The basic test setup and ground rules for the comparison were as follows: Model: Two models were used for comparisons. One wedge had a 201 deadrise angle, the other a 101 angle. Both models were to be considered totally rigid. Small

A. Engle, R. Lewis / Marine Structures 16 (2003) 175182

177

diaphragm pressure gauges were installed on one deadrise face of each wedge (see Fig. 1). Symmetry was assumed across the centerline with two pressure gauges providing data as a check. Test xture: The wedges were dropped in test xture that acted as a guide up to and through the impact event (see Fig. 2). Wedge velocity and acceleration were measured during each drop. Acceleration in the vertical plane was measured by accelerometers located at the center of roll. Velocity was measured by a sonic transceiver placed on the wedges upper surface. A at rigid piece of metal was mounted to the drop test xture that was used as the xed reference for the sonic transceiver. Run conditions: Pressures were requested to be determined at six locations along the longitudinal centerline of the wedge. The locations are 1.5, 4.5, 10.5 and 13.5 in measured from the point of the keel. Impact velocities were to include a range from 0100 in/s. A recommended set of impact velocities include 49.1, 55.6, 62.2, 65.2, 68.1, 73.5, 76.1, 78.6, 83.4, 87.9 and 92.2 in/s. Items to be compared: (1) Largest peak impact pressure vs. impact velocity irrespective of wedge location. In order to properly compare results with available model tests, all submittals were requested to be in in/pounds units (pressure in psi, impact velocity in in/s). Computational results were received from a total of ve participants. The basic theories for each of these computational methods are: Method A: Computations were run using the program USAERO, a 3-D panel code [3]. For most problems, USAERO generally has a xed panel format (i.e., once the panels are congured, they stay in that position throughout the computation). However, for intersecting free surface problems, where the intersection moves across

Fig. 1. Wedge design.

178

A. Engle, R. Lewis / Marine Structures 16 (2003) 175182

Fig. 2. General test arrangement.

the body, USAERO must re-panel the surface automatically each time step in order to keep the free surface intersection and panel edge aligned (that is to say, as the body heaves the panels must adjust them in order to maintain a clean solution). Method B: Applied a 2-D BEM code. According to its denomination 2-D BEM is a two-dimensional boundary element method that solves the water-entry potential ow problem for an arbitrary (non-symmetric) cross section. A time-varying drop speed is accepted. The free-surface is assumed to be equi-potential. Wagners approach [4] is used for the free-surface, in the sense that the water pile-up at the two edges is accounted for by discretising the free-surface at each time step. At the intersection between the body and the water, a jet-ow boundary condition is imposed. Method C: The calculation method is based on a formulation developed by Takemoto [5] with further consideration given to account for both deacceleration and the number of time steps in the calculation being >100. Method D: A nite element technique has been applied to solve the slamming problem of a 2-D wedge penetrating a at-water surface. In the FEM analysis, the effects of splash and free surface pile up are considered. Impact velocity is assumed as a constant value during the slamming. In other words, the deceleration of the wedge due to the slamming load is not considered. Incompressibility and nonviscosity of uid are assumed. The effect of gravitational force is neglected and the hydrostatic pressure is not included.

A. Engle, R. Lewis / Marine Structures 16 (2003) 175182

179

Method E: The method that was used employs ideal, incompressible ow. The hydrodynamics are, however, fully non-linear, but the non-linear boundary conditions are satised on the horizontal axis, consistent with the assumption of small deadrise angle. Unlike Wagners method and its variants, velocity, pressure, and displacement continuity conditions are fully satised so that the jet velocity and the jet head pressure peak are well dened.

3. Discussion of results The maximum pressure from each drop test was estimated by using the largest peak pressure measured by the ve pressure gauges. The largest pressure was plotted for both the 101 and 201 wedges. The selected computations along with predictions made based on the theories of Wagner and Chuang are shown in Figs. 3 and 4. Wagners prediction is based on the formula:
2 2 2 Pmax 1 2 rV 1 p =4 cot y;

where Pmax is the maximum impact pressure, V the initial vertical impact velocity, y the wedge deadrise angle, as measured from the horizontal axis, and r the density of water. Chuangs prediction [8,9] is based on relationships he developed based on previous tests he performed. For the 201 wedge both Wagner and Chuang make the same prediction for maximum pressure. Of the computational methods examined in this study, only Method C overpredicts the maximum pressure seen in the tests while Method B appears to underpredict the measured results. For the 101 wedge, the maximum predictions by Chuang match up well whereas those by Wagner are high and serve better as an envelope describing the highest possible maximum pressure vs. impact

Test Data Method B Method D Chuang and Wa gner 60000 50000

Method A Method C Method E

Pascals

40000 30000 20000 10000 0 0 0.5 1 1.5 2 2.5

Impact Velocity (m/s)

Fig. 3. Maximum wedge face pressure vs. predictions 201 wedge.

180

A. Engle, R. Lewis / Marine Structures 16 (2003) 175182


Test Data Method D 250000 200000 Method A Method E Method B Chuang Method C Wagner

Pascals

150000 100000 50000 0 0.00

0.50

1.00 1.50 Impact Velocity

2.00

2.50

Fig. 4. Maximum wedge face pressure vs. predictions 101 wedge.

for a wedge. As was seen in the 201 wedge test, Method C again is shown to overpredict the measured results by the greatest amount. Of all the methods examined in this study, the approaches based on the theories of Maskew et al. [3] and Vorus [6] seem to provide the most consistent results. Better agreement between the various methods and model tests were obtained for the 201 wedge then for the 101 test condition. As the drop velocity for the two test sections were the same for both wedges, it appears that some of the theoretical approaches which have been examined will not be as applicable, in a general sense, as others might be. The study contained herein looked at wedge angles of 101 and 201 and assumed rigid body response. However, it is important to point out that depending upon the deadrise angle being examined, the effects of hydroelasticity may also need to be considered. Specically, Haugen [7] has determined that for relative angles of 61 or less (i.e., the angle between structural beams and the free surface), hydroelasticity can result in reducing impact pressures. Since the wedges utilized in this study were rigid and the angles which were tested are in excess of the critical 61 angle, excluding the effects of hydroelasticity in this study should not be considered important.

4. Conclusions The development of a rst principles approach to predicting hydrodynamic loads requires that designers have the tools available to predict both ordinary waveinduced loads and slam-induced whipping loads that result from hydrodynamic impacts. The initial studies contained herein show encouraging results for the prediction of hydrodynamic impacts. Once a comprehensive validation of these tools is completed, it is believed that these methods will have the potential to provide designers with a practical and advanced design assessment capability.

A. Engle, R. Lewis / Marine Structures 16 (2003) 175182

181

The following are some more specic conclusions and recommendations with regard to the applicability and limitations of the particular codes discussed in this paper. Predictions for hydrodynamic impact forces associated with a slam induced whipping response can be reasonably predicted by 2 and 3-D methods presented in this study. However, some of the rigorous treatments that are available require substantial CPU time. Since uncertainty analysis methods were not used to assess experimental uncertainty, it is unclear whether the degree of variation between experiment and simulation is solely due to limitations within available theories or the result of errors within the experiments themselves. Presently, only vertical impact loads have been examined included. However, a review of the literature suggests that horizontal whipping can be extremely important for ships with ne bow sections. Therefore, the planned inclusion of horizontal impact loads will hopefully improve the prediction of the total high-frequency whipping loads that can be important for ultimate strength and fatigue estimates.

Acknowledgements The authors wish to thank the people and organizations that participated in this study. Professor Bill Vorus and Dr. Rick Royce from the University of New Orleans, Mr. Jim Fraser of AMI, East, Dr. Toyoma of Mitsui Engineering, CENTENA, Dr. Shen Jin-Wei of the China Ship Scientic Research Center.

References
[1] Armand JL, Cointe R. Hydrodynamic impact analysis of a circular cylinder. 5th OMAE Symposium Proceedings, Tokyo, Japan, 1986. [2] Howison SD, Ockendon JR, Wilson SK. Incompressible water entry problems at small deadrise angles. J Fluid Mech 1991;21530. [3] Maskew B, Wang M-L, Troesch AW. Comparison of calculated and measured loads on a ared body oscillating in a free surface. Presented at the 20th Symposium on Naval Hydrodynamics, Santa Barbara, CA, August 1994. [4] Wagner H. Landing of sea planes. NACA Tech Memo 622, SITDL TR 2101, May 1931. [5] Takemoto H. J Soc Nav Archit Japan 1984;156:31422. [6] Vorus W. A at cylinder theory for vessel impact and steady planning resistance. J Ship Res, June 1996. [7] Haugen EM. Hydroelastic Analysis of slamming on stiffened plates with application to catamaran wetdecks. Dr Ing Thesis, NTNU, Trondheim, 1999. [8] Chaung, Sheng-Lun. Slamming of rigid wedge shaped bodies with various deadrise angles. DTRC report no. 2268, October 1966. [9] Chaung, Sheng-Lun. Investigation of impact of rigid and elastic bodies with water. NSRDC report no. 3248, February 1970. [10] Engle A, Lin WM, Salvesen N, Shin Y-S. Application of 3-D nonlinear wave-load and structuralresponse simulations in naval ship design. Nav Eng J 1997;109(3):25368.

182

A. Engle, R. Lewis / Marine Structures 16 (2003) 175182

[11] Lin WM, Shin Y-S, Chung J-S, Zhang S, Salvesen N. Nonlinear predictions of ship motions and wave loads for structural analysis. Proceedings of the 16th International Conference on Offshore Mechanics and Arctic Engineering (OMAE97) I: Part A, 1997, p. 24350. [12] Sclavounos P, Kring D, Huang Y, Mantzaris D, Kim S, Kim Y. A ship ow computational method as an advanced CFD tool of design. SNAME Trans 1997;37598. [13] Von Karman T. Calculation of the pressure distribution on airship hulls. US Experimental Model Basin, SITDL, TR 1856, March 1930.

S-ar putea să vă placă și