Close

Forum

Questions about mechanical properties determination of composite materials

  1. Lina Shang

    Hi all,

          I want to use VABS method to analyze the twisted composite tailored rotor blades developed and tested by Bao and Chopra. The cross section of these composite rotor blades consists of an IM7/8552 graphite/epoxy D-spar laid up on a foam core mandrel with embedded leading-edge weights, an aft cell foam core and IM7/8552 graphite/epoxy weave skin. Now I know the size and material of each cross section constitution part but have a problem to get some material mechanical properties.

          1) The D spar is made of IM7/8552 graphite/epoxy unidirectional prepreg. Bao and Chopra measured the mechanical properties of IM7/8552 graphite/epoxy unidirectional prepreg: E1=169.6Gpa, E2=10.3Gpa, G12=8.3Gpa, v12=0.34 and thickness=0.0069inch. From these data, I can get E3, v13 and G13 by using E3=E2, v13=v12, G13=G12, but I don't know how to determine v23 and G23?

          2) The skin is made of a ply of  IM7/8552 graphite/epoxy weave prepreg with ±45 degree. I don't know how to calculate or get the mechanical properties of the weave composite prepreg.

          3) In order to minimize the size of the leading edge weights, tungsten alloy (class IV, with a density of 18500 kg/m3) was used to build the weights. The tungsten alloy Designation is 97W2Ni1Fe(97W,2.1Ni,0.9 Fe), can anybody hlep me find this alloy's Poisson's ratio and Young's modulus?

    Reply Report abuse

    Replying to Lina Shang

    1. Wenbin Yu

      1) if you cannot find the properties from the literature, you can assume v23=0.3 and assume the material is transversely isotropic, then G23=E2/(1+2 v23)

      2)To be rigorous, a micromechanics approach is needed for you to figure out the property for a weave prepreg. As a simple approximation, you can use smeared properties. Just transfer the properties into a single coordinate system, then average it out. 

      3) If google cannot help you, then a simple dog-bone test in the lab will.

      Reply Report abuse

      Replying to Wenbin Yu

    2. Lina Shang

      Dr. Yu, I still have some questions about your answer:

      1)  Should the equation for G23 be G23=E2/2(1+v23) rather than G23=E2/(1+2 v23)?

      2) For a simple approximation, my understanding is as follows: First, the properties of the +45 degree weave and the -45 degree weave are transformed into the common ply coordinate system seperately; Then, in the common ply coordinate system, by assuming the strains for +45 degree weave and -45 degree weave are the same and the stresses for both weave obey the superposition principle, the smeared young's modulus equals to the sum of young's modulus of both weave, the smeared shear modulus equals to the sum of shear modulus of both weave and the smeared poisson ratio equals to the average of possion ratio of both weave. Is my understanding correct?(I think there are some problems with my assumption, by assuming the strains are the same for both weave in the common ply coordinate system, the poisson ratio for both weave should be the same and this is not the case for the unsymmetric weave prepreg)

      Reply Report abuse

      Replying to Lina Shang

    3. Wenbin Yu

      1) You are right. It should be the well known formula from mechanics of materials E=2(1+v)G.

      2) You need a good micromechanics code to compute the effective properties if you are not satisfiied with your assumptions. My code SwiftComp can serve this purpose. 

      Reply Report abuse

      Replying to Wenbin Yu

    4. Lina Shang

      Dr Yu,

         I read some of your papers and your students’ dissertations about VAMUCH. To predict effective material properties for composite materials by VAMUCH, Young’s modulus and poisson’s ratio of fiber and matrix and fiber volume fraction should be known. I cannot find these properties for IM7 carbon fiber and 8552 epoxy matrix, so I think I’d better use a simple approximation first.

        In my aforementioned simple approximation, I think the transformation matrix should act on the stiffness matrix D (sigma=D*epsilon) rather than engineering constants. Then by considering the +45 weave and -45 weave separately and assuming the strains for both weave are the same and the stresses for both weave obey the superposition principle in the common ply coordinate system, the smeared stiffness matrix equals to the sum of the transformed +45 weave stiffness matrix and the transformed -45 weave stiffness matrix .

      Reply Report abuse

      Replying to Lina Shang

    5. Wenbin Yu

      You are right that transformation should be applied to as a fourth order tensor. Even if you don't have fiber matrix properties, you can also directly input weave properties into SwiftComp, it is still better than your smeared properties. The beauty of SwiftComp is that it allows you to capture details as needed or possible   

       

       

      Reply Report abuse

      Replying to Wenbin Yu