fem modeling and analysis in prevention of the waterway dredgers

1、ReviewFEM modeling and analysis in prevention of the waterway dredgers crane serviceability failureDu?an Kovac ˇevic ´ a,?, Igor Budak a, Aco Antic ´ a, Ale? Nagode b, Borut Kosec ba University of Novi Sad, Fac

2、ulty of Technical Sciences, Trg Dositeja Obradovic ´a 6, 21000 Novi Sad, Serbia b University of Ljubljana, Faculty of Natural Sciences and Engineering, A?kerc ˇeva c. 12, 1000 Ljubljana, Sloveniaa r t i c l e i n f

3、oArticle history:Received 6 July 2012Received in revised form 27 September 2012Accepted 6 October 2012Available online 5 November 2012Keywords:Waterway dredgersJib craneServiceability failureFEM modelingLink FEa b s t r

4、a c tThis paper gives a proposal of one suitable concept for FEM structural modeling and anal-ysis of cranes which are a common part of the waterway dredger facilities. These dredgerstypically contain two cranes: main ji

5、b crane between the catamaran-like pontoons andauxiliary crane in the bow zone. Discussion here is dedicated towards advantages andweaknesses of various FE models for the crane structures, in order to avoid of frequently

6、failures of serviceability state of a main crane. Basic goal is to emphasize the benefits ofsophisticated but practical numerical model in comparison with the simpler and inade-quate models for prediction of a real struc

7、tural behavior of jib cranes. In that sense resultsof our research could be sort of template for FE modeling phase in an efficient design ofsuch structures.? 2012 Elsevier Ltd. All rights reserved.1. IntroductionIn the c

8、lassification of the waterway dredgers for exploitation of material under the water surface, a large group is bucket dredgers with the buckets on the continual chain supported by a jib crane structure [1], as is shown in

9、 Fig. 1. The excavation is performed by moving the buckets which plunges into the material at the water bed. Excavation conti- nuity depends on the bucket size and distance between them, as well as on the speed and lengt

10、h of chain. The most frequent failure in function (i.e. serviceability failures) is consequence of halt in chain free movement, because of large lateral defor- mation of jib crane. Inadequate jib’s design as a consequenc

11、e of improper numerical model for jib crane is a usual reason for appearance of this type of failure. This paper considers the proper FE model of structures of the jib crane as a girder for the dredger’s working tool and

12、 bow crane as facility for manipulation of jib crane in the service and repair circumstances. The papers dealing with the FEM analysis of such similar type of structure [2–5] are relatively rare. Their main feature is ap

13、plication of complex models, but only to analyze the critical structural parts. Another group of papers recommend sim- plest models which are not adequate in simulation of jib’s real behavior. These circumstances have ad

14、ditionally motivated the authors to research the alternatives, i.e. the possibilities for advanced modeling of the jib structure as a whole. Due to exploitation conditions, the jib crane structure of this type of dredger

15、s should satisfy several opposite demands and its design should be, in the positive engineering sense, a compromise solution. Jib bearing capacity is the main and man- datory performance. Stress condition in all structur

16、al elements has to be within the limits which exclude the failure state [6]. Jib structure stability (i.e. buckling resistance) is another requirement for the structural integrity. Global buckling is ques- tionable, whil

17、e the appearance of the local stability loss is possible because this is a thin walled structure.1350-6307/$ - see front matter ? 2012 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.engfailanal.2012.10.009

18、? Corresponding author. Address: Laboratory for Testing of Structures, Department for Civil Engineering and Geodesy, Faculty of Technical Sciences,University of Novi Sad, Trg Dositeja Obradovica 6, 21000 Novi Sad, Serbia

19、. Tel.: +381 214852646, mobile: +381 246123008; fax: +381 21459925.E-mail address: dusan@uns.ac.rs (D. Kovac ˇevic ´).Engineering Failure Analysis 28 (2013) 328–339Contents lists available at SciVerse ScienceDirectE

20、ngineering Failure Analysisjournal homepage: www.elsevier.com/locate/engfailanalThe analysis is performed for the uniformly distributed load (weight of the chain with the buckets). Fig. 2 shows principal stresses and ver

21、tical displacements in characteristic points as well as the lowest natural frequencies for 1D (top) and 2D (bot- tom) models. 1D model is formed from the beam FE (2400/1000 mm box cross-section, wall thickness t = 20 mm)

22、 and in the topological sense it is completely identical to the 2D model with the rectangular shell (isoparametric, nine-node, heterosis, thickness t = 20 mm). Boundary conditions are identical for both models and adjust

23、ed to the real jib’s supports conditions. It is evident that the principal stresses (S1 and S2) in the more accurate 2D model are 1.75–11.75 times greater than in a simpler 1D model. The state with vertical displacements

24、 (Dz) is similar. Here the factors are from 1.54 to 1.80. Furthermore, the lowest natural frequency (f1) in a 1D model is almost 1.8 times lower than in the 2D model. It is necessary to note that both models have the sim

25、ilar (laterally curved) shape of lowest natural frequency. These differences in the response for the same action indicate the necessity of application of the more sophisticated mod- el with 2D shell FE in analysis of the

26、 jib. This test shows very clearly that apparently similar models can obtain very diverse data about the structure, and sometimes also a very wrong impression about the bearing capacity, serviceability and risk of failur

27、e, thus definitely confirming the demand for applying more complex models. The computing time for proposed 2D model is comparatively short and therefore it is affordable model in engineering sense. Making of 2D model nee

28、ds more time and designer’s competency, but benefits are several. In this sense, all further considerations will apply the numerical model with 2D shell FE.3. Recommended FE model of the jib structureBecause of new produ

29、ction demands jib crane will be extended from initial length (35.5 m) to new length L = 49.0 m, by addition of new segments marked on the Fig. 3. It is important to emphasize problems of serviceability failures in earlie

30、r service-life of dredger, before strengthening redesign and new length reconstruction and confirming of these failures by use of described 2D shell FE model contrary to 1D beam FE model. These failures were a consequenc

31、e of incorrect initial de- sign and use of inadequate model in initial design numerical analysis. The increase of the jib length has the following consequences: increase of axial, flexural and torsional flexibility and i

32、ner- tia increase, i.e. natural frequencies decrease what could be possible origin of more frequent serviceability failures. In this sense, adequate plan of structural stiffening were performed. In modeling, the FE softw

33、are AxisVM? 11.2c has been used. AxisVM? enables linear and nonlinear, static and dynamic analysis of structural behavior for various types of actions [11]. For the geometric modeling of the jib structure, the non-automa

34、tic approach for FE meshing has been selected. The pro- cedure has the following steps:– designing stiffeners (only one symmetric side) of diverse type and dimensions, Fig. 4, – placing stiffeners at appropriate position

35、, i.e. forming the structure skeleton and connecting stiffeners with the cover plates (symmetric side), Fig. 5, – elaborating the support details – for example, zone of the axle around which the jib rotates in the vertic

36、al plane, Fig. 6 and – adding another symmetric side, Fig. 7.Such an approach leads to the model with ‘‘near-to-minimal’’ FE mesh size (important for the computation efficiency) and ‘‘a(chǎn)lmost-always’’ rectangular FEs, wit

37、hout significant shape distortion, which is the most favorable solution for the minimi- zation of numerical errors in computation.Fig. 2. 1D and 2D models: displacements, principal stresses and the lowest natural frequen