Вы находитесь на странице: 1из 14

Non-linear FEA of Combined Joint of Metal and Nonmetal Structure Components

Anderson, Eric A., Henry, Robert F., Shkolnikov, Vladimir M.


Art Anderson Associates, Bremerton, WA, USA Combined bonding/fastening joints of metal and non-metal structure components are quite beneficial for heterogeneous structural systems to maximize their capacity/weight efficiency. Correct design analysis of such joints is, however, a challenging task that requires sophisticated non-linear 3D structure simulation. Paper reflects results of ABAQUS FE analysis of such a combined joint utilized in composite module of floating lighter system.

1.

Introduction

The paper is related to design and production of large heterogeneous load-bearing systems such as ships and marine floating platforms. The joints of metal and non-metal structural components are an integral part of such constructions. The joined components could be either composite sections of metal ships, for instance composite superstructures of a metal surface vessels, outboard structural structures (sail, stabilizer, nose dome etc) of a submarine, or metal connectors and other installations for a composite module of a floating lighter platform. The module is being developed by Art Anderson Association (AAA) in the frameworks of an engineering investigation aimed at creation of Navy Shore Entry Amphibious and Barge Offloading System (SEABOS). The module is an elementary unit of the deployable floating platform for discharging cargo from ships and moving that to shore in the event a port is denied, degraded or not available [1, 2]. This is a relatively large composite structure, the overall dimensions of which are 80248 = 24.47.32.4m. The principal modules distinction from the conventional marine structures is that this is outfitted with rigid and flex connectors for joining with other modules to be deployed as various floating platform configurations. The platforms are subjected to high sea state that brings about severe operational conditions for the modules. Specifically, the paper is dedicated to a FE analysis of joint of the metal rigid connector with the basic composite structure of the module. The FE analysis is performed regarding a non-uniform non-linear structural model constructed of 3D solid elements with utilization of ABAQUS software.

2.

Problem identification

Three typical techniques are commonly used for the joining of heterogeneous load-bearing structures: 1) mechanical fastening with utilization of bolts, screws, or rivets; 2) bonding using a polymer resin or an adhesive; 3) combined bonding/fastening. The primary benefit of mechanically fastening is easy disassembly, inspection, and repairing of the joined components. The major drawback is the stress concentration nearby fastener holes and non-uniform stress distribution between the holes. That leads to high failure probability of the

2002 ABAQUS Users Conference

entire joint when one fastener fails and to lowering the joint actual efficiency depending on individual capability of the fasteners. In addition, the hole inner surface is subject of intensive corrosion due to the seawater exposure. Watertightness of the mechanically fastened joint is also a significant issue pertinent to most marine engineering structures. Bonding joints, applicable for comparatively low-loaded heterogeneous structures, are configured so that their strength potential is limited by adhered failure outside of the joint boundary. Nevertheless, the peel stresses known to significantly reduce joint strength are the primary cause of failure of the bonding joints. Particularly, minimization of the peel stresses is a tricky task, when a standard non-tapered structural profile (angle) is employed as a joining agent. The combination of bonding and fastening substantially reduces the shortcomings pointed-out above. The greatest benefit is reduction of stress concentration nearby the holes and smoothing of the non-uniform stress distribution. In addition, the watertight integrity and corrosion protection are both maintained in a natural way by filling up all gaps and cavities of the combined joints with a polymer resin/adhesive. That includes areas inaccessible for inspection and application of protective anticorrosion coating. All that considerably increases serviceability and reliability of the structure and lessens its weight, production cost, and maintenance expenses. The known results of previous investigations and collected design and operation experience testifies convincingly for considering a combined joint as a consolidated structural unit [3, 4]. While the benefits of the combined bonding/fastening are recognized, it is difficult to achieve them executing the conventional structural design procedure. Stress determination and strength reconciliation of the combined joints both are not trivial computational tasks. Particularly, this is only a reason for the existing conservative requirement that each component of the combined joint, if applied, is designed to carry the entire load, thereby ensuring full structural redundancy. Obviously, there are penalties in weight, cost, and lowered structural efficiency as a result of the redundancy. To avoid that, it is necessary to perform a detail FE simulation to obtain analytically verified data on determination and optimization of the joint design parameters. Interaction of the bonding and fastening brings about the mutually supportive effect, which significantly increases load-bearing capability of a combined joint. Apparently, while the bonding adhesive film exists the fasteners can not fail. Concurrently, while fasteners are unbroken the adhesive film can not be destroyed. A fastener could achieve its yield point continuingly bearing its partial load and allowing underloaded fasteners to take higher load. Overall, that smoothes the load distribution and heightens the joint capability. This ideal mental picture of a combined joint behavior might be reflected at a simplified expression relevant to the overall joint bearing capability Q :
Q= c
j

fi

Y f A fi + cb S b Ab j

(1)

where: Y is the yield strength of a metal fastener, S is the allowed strength of an adhesive film, A is the cross-section area of each a joining component, c is a reduction coefficient relevant to partial involvement of an individual joining agent due to non-uniformity of stress distribution over the entire joint structure, f and b are indexes denoting fasteners and bonding film respectfully, i and j are count indexes of the individual fasteners and contacting surfaces, if more than one is designed.

2002 ABAQUS Users Conference

The expression (1) is simple however quantification of the c coefficients is quite problematical in simplified modeling. The coefficients c are variable in range 0 c 1 depending on actual structure configuration and loading conditions. Only way to determine them analytically is accurate simulation of the combined joints with a non-uniform computational model in its nonlinear deforming. Presumably, such the analysis will result in measurable labor, cycle time and/or material cost savings, design improvements, and maintenance cost reduction regarding heavy-loaded heterogeneous structures. So far, the adequate and accurate simulation of the combined joints under operational exposure is the pressing task with respect to the development of efficient heterogeneous structural systems. At the present, the task is resolvable throughout a non-linear FE analysis with application of advanced commercially available FE software.

3.

Computational model

Such a FE analysis has been carried out based on ABAQUS software regarding a combined joint of a metal rigid connector with a basic composite module structure of a SEABOS floatable platform [1, 2]. This is a representative example to illustrate the structural behavior of a combined joint as well as the efficiency of the applied FE computational model and software. ABAQUS FE software is quite suitable for the above-described problem. This can model the processes of non-linear deforming of multi-laminate composite systems under arbitrary loading and enables evaluation of stress-strain state, strength, elastic stability, vibration and dynamic resistance pertinent to the problem under consideration. The structure is modeled as a section of the module and reflects its basic design parameters. The model comprises of a composite (GRP) structure, a steel housing structure of the rigid connector, a backing GRP rib cage, a polymer adhesive film, and steel barrel-nuts with fastening bolts. The model undergoes global bending relevant to the required sea state 5 conditions in the PiersonMoskowitz spectrum [1, 2]. The bolts are loaded with additional pretension load. In particular, the composite part is modeled as a structure 48 tall by 27 wide with a 5/8 thick top deck and a 1 thick end wall. The stiffeners of the GRP rib cage are thick plates spaced 3 apart. The connector housing is assembled of two welded steel plates. The bolts fastening the steel and GRP sections via barrel-nuts are modeled as steel round rods. The barrelnuts themselves are modeled as 4(OD)3/8 steel pipes. The adhesive is represented as 0.01 epoxy film. The parts are assembled and tied by contact surfaces of the GRP and steel plates with adhesive film, by the barrel to GRP ribs, the bolts to GRP and to barrel, respectfully. The bolt heads are tied to the steel plates. Restraints are applied to the top deck and to the vertical stiffeners and bottom face of the exterior skin. All these features, the whole model assembly, and its boundary-loading conditions are represented on Figure 1. The GRP applied for the module skins is a fiber glass epoxy resin quadraxial (0/+45/-45/90) stitch-bonded composition based on Glass = P4MA to BFG spec ES0024 and a toughened epoxy = Formulation B to BFG spec ES0023 system. The glass package includes a small amount of a mat. Mechanical properties of the applied GRP composition as well as of other utilized materials are summarized in Tables 1 and 2. The individual plastic properties of the utilized steels are approximated as shown on Figure 2 based on the given properties data [5]. The adhesive is considered as a linearly elastic material. Two runs have been completed on this model. One is with a bolt pretension of 28 kips per bolt and the global tension load of 582kips. Another has been done with a bolt pretension load of 1.5 kips

2002 ABAQUS Users Conference

and the same global load. The 28 kips is based on the AISC standard bolt pretension for a diameter bolt and the 1.5 kips has been based on 5% of the AISC standard.

4.

Results analysis and discussion

In a result of the performed FE simulation, all stress components of the structure on varied load steps are gained. Overall, the character of the structure behavior completely agrees with the preliminary constructed mental picture: the initial zones of high stress are getting non-linear widening along with the linear loading increase thus smoothing the stress distribution among the fasteners. Figures 3 - 7 illustrate the computational results regarding the bolt pretension load of 1.5 kips at the final loading conditions, when the applied tension has approached its maximum. The shown Moir pictures reflect Von Misses stresses of the joint structural components. Specific conditions of the operation, particular design solutions, selected materials, and manufacturing processes, all are factors affecting the joint failure mode. For instance, the lay up technology commonly used for marine structures is associated more with the initiation and propagation of embedded cracks or flaws than with in-plane fracturing. The failure modes typically inherent to joints of heterogeneous structures are the following: Failure of fiber reinforcement of composite structural part (mode 1), causing loss of in-plane shear load carrying capability due to extensive matrix damage and, for laminated structures, delamination due to transverse normal and/or shear stresses (interlaminar separation). Matrix damage, i.e. intralaminar cracking (mode 2), resulting in delaminations through the thickness of the laminated shell structures. Debonding of the joined structures as result of adhesive film failure (mode 2a). Failure of joined metal part (mode 3). Breakage of mechanical fasteners (mode 4). Changeability of the severe operational conditions that affect marine structures, dictate that each failure mechanism be evaluated in the structural analysis. That could be done conventionally via comparison of the acting stresses with allowed stresses which are determined as a function of the ultimate strength taking into account knockdown and safety factors. Those are reflecting specified conditions of operation and individual properties of an applied material composition. The stress disturbance inherent to a structural joint due to its complicated configuration brings about distinct 3D stress state which requires, in general, 3D failure analysis of applied materials. For metal parts, that could be done via strength reconciliation applying the Huber-Mises formula convenient for checking the effect of applied stresses on the yielding of metal plates:
2 cr = x x y + 2 y + 3 xy

(2)

If a certain combination of normal x , y and shear xy stresses results in a critical stress equal to the yield strength this combination of the stresses is assumed to just produce the yielding. Correspondingly to possible failure modes of composite parts, 2D and 3D failure criteria should be applied to characterize the material stress level. The common used quadratic 3D criterion relevant to the mode 1 failure of a structural orthotropic composite is expressed as [9]:

2002 ABAQUS Users Conference

~
k =1 l =1

kl

=H

(3)

where: H is a parameter summarizing actual stress level in a particular structure area; the ~ are components of characteristic stresses: situation, when H = 1 , answers to the failure onset; kl
~ = kl kl S kl

(4)

about the axes of orthotropy k , l = 1, 2, 3 ; kl are acting stress components; S k l are ultimate stresses of a utilized GRP. For design reconciliation purposes, S k l imply design allowable stresses conventionally reflecting design safety margin and strength knockdowns covering fatigue, temperature, moisturizing, and UV reducing influences. Depending on the sign of the normal stresses kk resulting in tension or compression, S k k are determined as
Skk Skk = S kk

( )t ( )c
m

for k k 0 for k k < 0

, k = 1, 2, 3

(5)

To assess the initiation (or propagation) of a GRP delamination and/or debonds of an adhesive film the criterion (3) is simplified to 2D failure consideration:
~
k =1 k3 2

2 for 33 0 =H, m= 3 for 33 > 0

(6)

This assumes that, regardless of the nature of loading, edge delamination is driven by interlaminar stresses that exist near the free edges of a laminate. All structural and joining components have been verified in their bearing capability based on the above represented criteria. The structure under consideration has been designed so that the metal section is the weakest link. The performed analytical work validated legitimacy and workability of this design concept.

5.

Conclusions

The computational results have clearly demonstrated that the applied FE model and software ABAQUS both are satisfactory to reflect structural behavior of the combined bonding/fastening joint. The performed simulation is quite useful to verify analytically the preliminary engineering assumption on the interaction of the structural components. The FE analysis has validated the sufficient strength of the combined joint with regards to the particular design solution to meet the given operational requirements. The analysis performed with the 28 kip bolt pretension results in the bolts yielding while the analysis performed with the 1.5 kip bolt pretension does not. So that, the 5% value is recommended to provide effective interaction of the bonding and fastening elements of the combined joint.

2002 ABAQUS Users Conference

The performed analysis is helpful in handling a factual stress situation and optimizing the joint configuration. The optimization enables one to predetermine the weakest link of the joint and to control the mode of its failure onset. As the performed detailed FE analysis with 3D failure evaluations has demonstrated, the engineering problem associated with the structural design of reliable and cost effective large metal to composite joints structures goes beyond routine structural design practices. The development and running time of the detail models and assessment of the results are quite time consuming. On the other hand, this complex analysis is currently necessary in order to create reliable, serviceable, and cost effective composite structures for military and commercial marine applications. This situation necessitates the development of a specialized, problem-oriented computational tool covering both conventional and innovative design solutions. Another appropriate approach to resolve the problem related to the combine joint design is collecting of relevant computational data and construction on this base calculation instructions and diagrams regarding typical design solutions for a wide range of applications. Those approaches could be implemented in parallel to cover all practical design cases. The commercially available FE code ABAQUS is well suited for such the problem resolution and could be considered as a basis for its further development.

Acknowledgement
The work reported in the paper is a part of the authors research carried out at Art Anderson Association under the Navy sponsored research and development contract, #N00014-01-C-0081. The authors wish to express their gratitude for the financial support as well as for the contributions that have been made by other organizations and individuals involved in the project.

References
1. 2. 3. Operational Requirements Document for Joint Modular Lighter System (JMLS), Version 7, US Navy, JMLS Program Management, 2/13/2001. Shkolnikov, V.M., Convertible Floating Dock JMLS-SEABOS. Guidance for Conceptual Structural Design, AAA, Bremerton, 2001, 24p. Shkolnikov, V.M., Pressing Problems of Composites Application for Submarine Outboard Structures.- Shipbuilding Industry.- ser. Design of Ships.-v.174, Leningrad, Shipbuilding, 1988, pp. 136-162. Nikolaev, L.S., Shkolnikov, V.M. Outer Composite Hulls of Submarines. Design and Strength of Joints. - Shipbuilding Standard B.5.1067-83, Leningrad, Shipbuilding, 1983. 1991 Annual Book of ASTM Standards, Section 1. Iron and steel Product, v. 01.04, Steel Structural, reinforcing, Pressure Vessel, Railway Camanho, P.P., Matthews, F.L., Stress analysis and strength prediction of mechanically fastened joint in FRP: a Revieu.- Composites: Part A, 28A (1997) 529 - 554. Camanho, P.P, Matthews, F.L., Delamination Onset in Mechanically Fastened Joints in Composite Laminates, Journal of Composite Materials, Vol.33, No.10, pp. 906-927, 1999. Shkolnikov, V.M., Service Life of Composite Structures Undergoing Variable Operational Exposure.- Questions of Material-Science, v. 3, St. Petersburg (Russia): CSRI CM

4. 5. 6. 7. 8.

2002 ABAQUS Users Conference

"Prometey", 1995, p.30-38. 9. Chen, W.H., and Lee, S.S., Numerical and Experimental Failure Analysis of Composite Laminates with Bolted Joints under Bending Loads, Journal of Composite Materials, Vol. 29, No.1, pp.15 1995.

Tables

Table 1. Mechanical Properties of Isotropic Components


Ultimate strength, ksi Yield strength, ksi Modulus of elasticity, ksi Poissons ratio Steel Connector Housing (A572) 65 50 Steel (A500) Barrel 45 33 Steel (A325) Bolts 120 92 Polymer Adhesive 2.75 285 0.3 29,000 0.3 29,000 0.3 30,000 0.29

Table 2. Mechanical Properties of GRP


Ultimate/Yield strength, Ksi Flatwise tension In-plane tension In-plane shear In-plane compression Interlaminar shear Modulus of elasticity, Ksi Throughthickness shear In-plane tension/compre ssion Flatwise tension/compre ssion In-plane shear Poissons ratio

47.3

39.9

60.4

20.1

4.5

5.1

2,400

1,000

720

370

0.27

2002 ABAQUS Users Conference

Throughthickness 0.1

In-plane

Bearing

Figures

Figure 1. Full assembly. Boundary & loading conditions

2002 ABAQUS Users Conference

Figure 2. Stress-strain diagrams of applied steels

2002 ABAQUS Users Conference

Figure 3. Steel connector section

10

2002 ABAQUS Users Conference

Figure 4. GRP section (front)

2002 ABAQUS Users Conference

11

Figure 5. Adhesive film

12

2002 ABAQUS Users Conference

Figure 6. Fastening bolts

2002 ABAQUS Users Conference

13

Figure 7. Barrel nuts

14

2002 ABAQUS Users Conference

Вам также может понравиться