Selection and impact of parameters in composite materials designing

Transcription

1 Selection and impact of parameters in composite materials designing S. Żółkiewski* Silesian University of Technology Gliwice, Poland Abstract The paper considers the problem of selection of parameters in designing of composite materials reinforced with fiberglass, carbon and Kevlar fibers. Reinforcements the matrix of epoxy resin Epidian 6 and polyester resin Polimal 1094 were used. Testing of composite materials was done using the HBM measuring device. The results were acquired by the CatmanEasy software and the laboratory amplifier MGCplus. The laboratory stand for delamination and the fatigue stand were exploited. Obtained results were compared with results obtained in Unigraphics NX 7.0. There is presented impact on composite materials, such properties as a sort of resin, a sort of reinforcement, a number of composite layers, a kind of connection between a laminate and a steel plate. 1 Keywords: composite materials, displacement, fatigue tests I Introduction Mechanical and physical properties of fibrous composite materials are very good in comparison with other constructional materials. The major advantages of this type materials are low weight and high strength [1, 2]. Therefore machines and mechanisms built up with composite materials have much less weight than traditional constructions. In this work there are presented results of testing fibrous composite materials connected in bolt joints and shape joints and adhesive joints. Composite materials reinforced with fiberglass (fig. 1), carbon (fig. 2) and Kevlar (fig. 3) fibers are considered. Fig. 2. Picture of the carbon composite material Considered composites are the hybrid ones, because of connection between laminates and steel plates. There are presented results combing properties of composites such as a sort of resin, a sort of reinforcement, a number of composite layers, a kind of connection between a laminate and a steel plate. Fig. 1. Picture of the fiberglass composite material * Slawomir.Zolkiewski@polsl.pl 1 13th World Congress in Mechanism and Machine Science, Guanajuato, México, June, Fig. 3. Picture of the Kevlar composite material Presented results can facilitate the designing of mechanisms consisted with such a type elements. Mechanical properties of composite materials depend on reinforcement types and the volume participation of reinforcement in the warp. The Catman software and the MGCplus laboratory amplifier were used for acquiring data. Obtained experimental results were compared with numerical simulation in Unigraphics NX environment. Numerical simulations are reusable and can facilitate the designing

2 process, but in many cases results obtained from this method are not precise. II. Analyzed composite materials and laboratory stands Chemohardenable resins are the most common products used for manufacturing of composite materials. Generally accepted resins are polyester, phenolic, epoxide and silicone ones. In table 1 strength parameters of analyzed composite materials are presented. Fiber type Mass density ρ [kg/m3] Strength limit Rm [MPa] The Young modulus E [GPa] Fiberglass t. E Carbon HS Carbon HM Carbon UHM Kevlar ,6 Kevlar TABLE 1. Strength parameters of exemplary composite materials [1] Important strength parameters of popular resins (mass density, tensile strength, the Young modulus) are presented in table 2. components and concordance of layers and fibers. Especially handmade samples have some faults with inhomogeneity. Properties\resins Epidian 6 Polimal 1094 Mass density in 20 C [g/cm 3 ] 1,17 1,20 Viscocity in 25 C [mpa s] Flexural strength [MPa] Tenasity [MPa] Time of gelation (100g \ 20 C) [minutes] TABLE 3. Properties of resins [1] To facilitate the identification of analyzed samples the code system is assumed. The code system describes major features of samples. First letter is the type of reinforcement fibers. S means fiberglass composites, W carbon ones and K Kevlar ones. On second position the basis weight is marked. On third position the number of laminate s layers. Letter E means epoxide resin, letter P polyester resin. The last position marks the system of connection between laminates and steel plates (K adhesive and collar joint, S screw joint). For example the sample S-1000-x-3-E-S means composite material folded with three layers of 1000g/m 2 basis weight fiberglass laminates hardened by epoxide and connected by screws. Chosen analyzed samples are presented in table 4. Sort of resins Mass density ρ [kg/m 3 ] Strength limit R m [MPa] The Young modulus E [GPa] Code of sample Fabric G.S.M. [g/m 2 ] No. of lays. Resin Conne ction type Polyester ,5-4,5 Epoxide ,4-96 1,9-4,9 Phenolic ,1-6,8 Silicone ,8-34 1,5-3,7 TABLE 2. The strength parameters of resins [1] Epidian 6 resin is often used for manufacturing different types of epoxide compositions, especially for epoxidefiberglass laminates and for electro-insulating and constructional applications. Polimal 1094 is constructional resin on the average flexible, accelerating agent with low vinyl benzene emission [3]. As curing agent for Epidian 6, the PAC hardener is used and for Polimal 1094 the Butanox M50 is used. In table 3 properties of resins (mass density, viscosity, flexural strength, tenacity and time of gelation) are presented. Properties of laminates depends on properties of S-1000-x-3-E-K Fiberglass Epidian 6 Collar S-1000-x-3-E-S Fiberglass Epidian 6 Screw S-1000-x-3-P-K Fiberglass Polimal Collar S-1000-x-3-P-S Fiberglass Polimal Screw W-600-x-5-E-K Carbon Epidian 6 Collar W-600-x-5-E-S Carbon Epidian 6 Screw W-600-x-6-E-S Carbon Epidian 6 Screw K-320-x-6-P-S Kevlar Polimal Screw K-320-x-10-P-S Kevlar Polimal Screw S-450-x-4-E-S Fiberglass Epidian 6 Screw S-450-x-6-E-S Fiberglass Epidian 6 Screw S-450-x-10-E-S Fiberglass Epidian 6 Screw TABLE 4. Analyzed samples of composite materials 2

3 In figure 4 a picture of exemplary sample (S-1000-x-3-E- S) is shown. There are presented parameters of exemplary tested sample in table 5. Individual layers (the laminate and steel plate) are connected by screws. Fig. 5. Scheme of the collar shape joint: 1 the screw M6 with washer, 2 the steel plate, 3 the laminate plate Fig. 4. Picture of the composite material S-1000-x-3-E-S The plates have dimensions 355mm x 355mm and thickness depending on the type of composite. The plates have handmade in laboratory terms. Another way of connection is the collar shape one, the scheme of such a type joint is presented in figure 5 and the photo in figure 6. Code of the sample S-1000-x-3-E-S Reinforcement type Fabric from fiberglass roving G.S.M. of reinforcement 1000g/m 2 Weave Linen Number of reinforcement layers 3 Type of resin Epidian 6 Amount of curing agent for 100g of resin Mass of reinforcement Mass of resins Mass of laminate Laminate s thickness Mass of the steel plate Steel plate s thickness Type of connection Mass of joints Overall mass 80 g 370 g 290 g 660 g 3 mm 680 g 0,7 mm Screws 8xM6 80 g 1420 g Fig. 6. Picture of the collar shape joint In figure 7 a scheme of the screw joint is presented. This solution can be disconnected and disassemble and permit for exchanging worn out element. Assumed strength parameters for reinforcement (Unigraphics NX 7.0) Tensile strength The Young modulus R m = 2900 MPa E = Mpa TABLE 5. Properties of sample S-1000-x-3-E-S Fig. 7. Scheme of the screw joint: 1 the screw M6 with washer, 2 the steel plate, 3 the laminate plate 3

4 In figure 8 the screw joint between the steel plate and the carbon laminate is shown. The sample is connected with eight screws, this type of connection involve drilling holes in connected elements. Dimensions of samples and arrangement of holes is determinate by laboratory stands. Very important is precision of drilling holes that does not produce stresses after screwing home. Fig. 8. Picture of the screw joint The arrangements of screws is presented in figure 9. The screws were arranged symmetrically and conditioned by dimensions of the clamping frame in the delamination laboratory stand. Fig. 10. The picture of laboratory stand with displacement sensors and loading systems The laboratory stand is built up of carrying frame, fixing frames and actuators and measurement elements. The stand is equipped in the force gauge (fig. 11) and the displacement sensor (fig. 12). In figure 11 the picture of the force gauge and the hydraulic motor operator is presented. Between the sample and actuator, the dynamometer is located. Fig. 9. Picture of the screw joint Tests were provided on laboratory stands (Silesian University of Technology in Gliwice, Institute of Automation of Technological Processes and Integrated Manufacturing Systems, Poland) presented in figures 10 and 11 and 12. Fig. 11. The picture of laboratory stand with the force gauge U2B 50kN At the other side of the tested sample there is a displacement sensor fixed (fig. 12). 4

5 The MGCplus laboratory amplifier (fig. 14) is a modular measurement device compatible with all HBM sensors. Fig. 12. The picture of laboratory stand with the displacement sensor WA L-20 In figure 12 the displacement sensor WA L-20 used for measurements in the laboratory stand is presented. Fig. 15. Windows of the Catman program For acquiring data the Catman software was used (fig. 15). III. Numerical simulation For numerical simulation the Unigraphics NX environment was used. This system is the advanced engineering environment enabling three dimensional modeling, generating technical and technological documentation, strength verification and other simulations. Modeling of composite demands creating of geometrical model of the analyzed system. Modelling of composite in Unigraphics (fig. 16) makes possible setting of parameters of each ply of laminate [4, 5]. Fig. 13. The picture of laboratory stand for fatigue tests The laboratory stand for fatigue testing is presented in figure 13. Testing is based on fixing the sample and loading in cycles using a pressure roll. Fig. 14. Picture of the MGCplus AB22A laboratory amplifier 5 Fig. 16. Setting parameters of the composite s plies in Unigraphics In the Unigraphics we can set parameters such as: ply material, thickness or angle of ply and nesting consecutive ply in every layer, making different structure from each layer. We can also define parameters such as:

6 matrix material - reinforcements layer, matrix volume fraction, warp fiber material, weft fiber angle, weft fiber material, and finished thickness. Fig. 19. The sample stress pattern of analyzed composite plate In figure 19 there is presented the sample juxtaposition of experimental results obtained from testing real models and numerical simulation results from theoretical models implemented in the Unigraphics. Force [kn] 0,1 0,3 0,5 1,0 1,5 2,0 2,5 Experimental test of composite with steel plate 0,7 mm Displacement [mm] 0,69 1,96 2,86 4,52 5,75 6,81 7,72 Fig. 17. The boundary conditions and results of numerical analysis of the composite plate In figure 17 the way of loading (boundary conditions) and the exemplary results of numerical simulation is shown. Displacement [mm] MES simulation in the Unigraphics NX 0,36 1,07 1,78 3,55 5,33 7,10 8,88 TABLE 6. The results obtained on experimental and theoretical ways In tables 6 and 7 the results obtained on experimental and theoretical ways are presented. Force [kn] 0 0,1 0,3 0,5 0,8 Experimental tests of laminate Displacement [mm] 0 1,84 3,99 5,46 7,10 MES simulation in the Unigraphics NX Fig. 18. The readout of displacement in nodes using the Identify Results tool In figure 18 there are presented sample results of numerical experiments. It is possible to obtain results for displacements (fig. 18) and for stress pattern (fig. 19). 6 Displacement [mm] 0 1,27 3,82 6,37 10,19 TABLE 7. The results obtained on experimental and theoretical ways

7 The presented in tables 6 and 7 results (displacements in function of loads) can be shown in charts. During the experiment, physical measurements and numerical simulation both in laminate and composite cases some discrepancy of displacement can be observed. The displacement curves transect in one point and there are adequate values, but other points are different. The reason for this can be inaccuracy of laboratory stands or the implemented ideal Hook s law in the Unigraphics environment. In figure 20 the juxtaposition of results from tables 6 and 7 is shown. Obtained results are similar in some ranges of acting force. IV. Analysis of experimental results Based on acquired data the comparative analysis is done. Assessment of impact on composite properties is done. Such properties are considered as: resins properties, reinforcement types, connection types. Fig. 22. The force-displacement chart juxtaposition of tested composite materials Fig. 20. The force-displacement chart juxtaposition of experimental and numerical simulation (theoretical) results for the fiberglass composite material There is presented exemplary comparison of experimental and numerical results of testing for the Kevlar composite material in figure 21. In figure 22 the juxtaposition of chosen tested composite materials is presented. The Epidian 6 and Polimal 1094 resins were used for manufacturing samples. Comparative analysis of samples S-1000-x-3-E-S and S-1000-x-3-P-S reinforced by fiberglass. In figure 20 the force-displacement chart is presented. The reason for discrepancy of displacement in this case can be different mass of samples. Laminate S x-3-E-S has mass larger than laminate S-1000-x-3- P-S. Fig. 21. The force-displacement chart juxtaposition of experimental and numerical simulation (theoretical) results for the Kevlar composite material 7 Fig. 23. The force-displacement chart juxtaposition of composite materials with Polimal (S-1000-x-3-E-S) and Epidian (S-1000-x-3-P-S) resin

8 Strengths both of resins, polyester and epoxide ones are similar. In case of high level of reinforcement participation this parameter can be negligible. In figure 23 a juxtaposition of composite materials with Polimal and Epidian resins is shown. In figure 25 a relation between different number of layers is analyzed. Results of analysis show direct proportion on displacements in function of number of layers. More layers means better stiffness. Fig. 24. The force-displacement chart juxtaposition of tested composite materials with carbon (W-600-x-5-E-S Laminate), Kevlar (S-320-x-10-P-S Laminate) and fiberglass (S-1000-x-3-E-S Laminate) reinforcements The comparative analysis of composite materials with different reinforcements is done. In figure 24 a juxtaposition of tested composite materials with carbon (W-600-x-5-E-S Laminate), Kevlar (S-320-x-10-P-S Laminate) and fiberglass (S-1000-x-3-E-S Laminate) reinforcements is presented. The body mass of samples is very similar (from 660 to 690 grams) so the comparison of samples makes sense. Discrepancy of displacements in figure 24 is connected with different characteristics of reinforcement fibers. The maximum resistance of sample has the carbon composite material and the minimum has fiberglass one. Fig. 26. The force-displacement chart juxtaposition of tested composite materials joint in adhesive (S-1000-x-3-P-K) and screws (S-1000-x-3-P-S) The adhesive joint composite material has more strength than screw joint one. Connections between a steel plate and the laminate using the collar shape joint and glue guarantee the concurrent displacements of steel plates and laminates. This type of connection ensures better strength properties but does not provide to possibility of disconnection. The realization of laminate collar is difficult and timeconsuming but the final product is more efficient than the others. In figure 26 a juxtaposition of tested composite materials joint in adhesive (S-1000-x-3-P-K) and screws (S-1000-x-3-P-S). Fig. 25. The force-displacement chart juxtaposition of tested composite materials with different number of layers: 4 (S-450-x-4-E-S Laminate), 6 (S-450-x-6-E-S Laminate) and 10 (S-450-x-10-E-S Laminate) 8 Fig. 27. The force-displacement chart juxtaposition of tested fiberglass composite materials with different number (4 and 8) of screws (M6) in the connection

9 In figure 27 two ways of connection are compared. The juxtaposition of tested fiberglass composite materials with four and eight screws is applied. was loaded by a pressure roll in these tests. The five thousands of cycles were provided and change of strength could be observed. The frequency of loading impulse was sixty cycles per minute (1 Hz). Fig. 28. The force-displacement chart juxtaposition of tested carbon composite materials with different number (4 and 8) of screws (M6) in the connection More efficient connection is the joint with eight screws. During testing of composites the solution with four screws had larger displacement than the one with eight screws. In figure 28 a juxtaposition of tested carbon composite materials with different number of applied screw is shown. Fig. 30. The force-displacement chart juxtaposition of tested composite materials before and after fatigue tests After fatigue testing samples were tested on strength again on a delamination laboratory stand. In figure 30 and 32 a juxtaposition of tested composite materials, carbon and fiberglass ones before and after the fatigue test are presented. Fig. 29. The force-displacement chart juxtaposition of tested Kevlar composite materials with different number (4 and 8) of screws (M6) in the connection In figure 29 a juxtaposition of tested Kevlar composite materials with different number of applied screw is presented. In this work the fatigue tests are also provided. Two samples S-1000-x-3-P-K and W-600-x-5-E-S are tested on the fatigue laboratory stand. Central point of samples 9 Fig. 31. Picture of the delaminated composite sample s collar After testing the samples were evaluated by visual method. The collar of sample is delaminated but the sample does not fail of a test piece. The photo of delaminated collar is presented in figure 31. In figure 32 a juxtaposition of tested carbon composite materials before and after fatigue testes. Increased difference in strength between a sample S-1000-x-3-P-K and a sample W-600-x-5-E-S arise from different properties of applied components, especially the reinforcement fibers type and used resins.

10 Fig. 32. The force-displacement chart juxtaposition of tested composite materials before and after fatigue tests (5000 cycles) Presented results are only initial research, the considered domain is very extensive and demands much more testing and considerations. V. Conclusions The tested samples are handmade and therefore there are some heterogeneities in laminates that can provide some result errors and provide distinctions between results obtained on theoretical and practical ways. Also the arrangement of reinforcement fibers is very important. Applied Polimal 1094 resin is dedicated for hand production. Polimal 1094 resin as distinguished from Epidian 6 resin has better processing features. The Polimal resin is very infiltrating one. It efficiently cut down time of infiltrating reinforcements fibers. The sample should be adequately prepared on the whole surface. Connections between a steel plate and the laminate using the shape collar joint and glue, guarantee the concurrent displacements of steel plate and laminate. This type of connection ensures better strength properties but does not provide to possibility of disconnection. The realization of shape laminate collar is difficult but the final product is more efficient than the others. The screw joint of laminate and steel plate is not adequate and provides some discrepancy of displacements. Therefore the strength properties are not so strong as they could be. Results confirm better properties of samples joint with the higher number of applied screws and the lower displacements of such a type realizations. Difference between the Young modulus of connected laminates and steel plates may occur destruction of elements. A major problem of modelling and simulating the composite materials in numerical environments is assuming physical and material parameters of analyzing elements. In literature there is some discrepancy between individual values, the producers of woven fabric do not provide any important values and parameters. Numerical simulations permit to testing of impact on strength 10 properties. In this type modelling most important are boundary conditions and properly assumed finished elements (FEM) providing appropriate reflection of warp and fibers. Handmade samples have different participation of warp and reinforcement fibers but results of testing on resin impact occurs no impact on this aspect. Based on obtained results the lowest displacement was for composite materials with carbon fibers. This type of reinforcement has high strength properties and low mass density. The analysis of force-displacement charts shows proportional increasing of strength together with increasing the number of layers. Carry out of fatigue tests shows declining strength properties of composite materials after five thousand cycles of loading. Acknowledgements: This work has been conducted as a part of research and development project N R supported by the The National Centre for Research and Development (NCBiR) in References [1] Hyla I., Śleziona J.. Composite materials. Basis of mechanics and designing. Publishers of Silesian University of Technology. Gliwice, 2004 (in Polish). [2] Boczkowska A., Kapuściński J., Lindemann Z., Witemberg-Perzyk D., Wojciechowski S.: Composites. Oficyna Wydawnicza Politechniki Warszawskiej, Warszawa 2003 (in Polish). [3] Web page [4] Reiner A., Peter B.: Simulationen mit Unigraphics NX 4 : Kinematik, FEM und CFD. Carl Hanser Verlag, Wien 2006 (in German). [5] Żółkiewski S.: Computer Aided Dynamic Analysis Of Composite Material Structural Components. Górnictwo odkrywkowe 2010 (in Polish).