NASA Reference Publication 1351 Basic Mechanics of Laminated Composite Plates AT. Nettles October 1994NASA Reference Publication 1351 Basic Mechanics of Laminated Composite Plates ALT. Nettles Marshall Space Flight Center * MSFC, Alabama National Aeronautics and Space Administration Marshall Space Flight Center + MSFC, Alabama 35812 a October 1994UL. ML. WV. Vi. VIL TABLE OF CONTENTS. INTRODUCTION ‘A. Intent and Scope B. Terminology and Notation C. Summary of Sections .. GENERALIZED HOOKE’S LAW FOR NONISOTROPIC MATERIALS ... ‘A. Normal Stress and Strain, Uniaxially Applied Force.. B. Stress and Strain, Plane Stress for Specially Orthotropic Plate ‘in, Plane Stress for Generally Orthotropic Plates......... D. Invariant Stiffnesses.. 5 MECHANICS OF LAMINATED COMPOSITES Assumptions... Definitions of Strains and Displacements Definitions of Stress and Moment Resultants Constitutive Equations for a Laminate.... : Physical Meanings of the [A], [B], and [D] Matrices. MOODD> NOMENCLATURE FOR DEFINING STACKING SEQUENCES... A. Coordinate System B. Nomenclature .......... IN-PLANE ENGINEERING CONSTANTS FOR THE LAMINATE... ‘A. Orientation of the Laminate . B. Symmetric Laminates C. Nonsymmetric Laminates..... D. Summary ENVIRONMENTAL EFFECTS .. A. Importance . B. Coefficients of Thermal Expan C. Moisture Effects... STRESSES AND STRAINS WITHIN LAMINAE OF A SYMMETRIC LAMINATE. oscsrce sree even A. Strains Within the Laminae. B. Stresses Within the Laminae. iii 28 28 28 33 4B 43 43 45 46 50TABLE OF CONTENTS VII. STRESSES AND STRAINS WITHIN LAMINAE OF AN UNSYMMETRIC LAMINATE. A. _ Difference From Symmetric Laminate: B. Example (0/+45)z Laminates . C. Determination of the Neutral Plane IX, REFERENCE OF IMPORTANT EQUATIONS REFERENCES. APPENDIX iv Page 63 63 63 R 89 95Figure 1 2 9. 10. LIST OF ILLUSTRATIONS Title Page Difference between an isotropic and an orthotropic plate 3 Definition of shearing strains 5 Generally orthotropic lamina 1 Displacements of a plate 12 Total displacements in a plate.. 13 Definitions of plate curvatures 14 Definition of stress resultant 15 Direction of stress and moment resultants 7 Cross section of a laminate 19 Displacements in an unsymmetrical plate... : = 2B Coordinate system for a typical laminate 26 Some laminate stacking sequences and their notation .. 2REFERENCE PUBLICATION BASIC MECHANICS OF LAMINATED COMPOSITE PLATES I. INTRODUCTION A. Intent and Scope This report is intended only to be used as a quick reference guide on the mechanics of continuous fiber-reinforced laminates. By continuous fiber-reinforced laminates, the following is assumed: (1) The material to be examined is made up of one or more plies (layers), each ply consisting of fibers that are all uniformly parallel and continuous across the material. The plies do not have to be of the same thickness or the same material. (2) The material to be examined is in a state of plane stress, ie., the stresses and strains in the through-the-thickness direction are ignored, (3) The thickness dimension s much smaller than the length and width dimensions. An attempt is made in this report to develop a practical guide that can be easily referenced by the engineer who is not familiar with composite materials, or to aid those who have seen this subject matter before. The scope of the report will be limited to the elastic response of the above-mentioned class of material. Strength-of-laminated composites will not be covered. General composite material mechanics and strength are developed in more detail in texts such as Jones! and Halpin? It is assumed that the reader has a general knowledge of elastic stress-strain behavior. B. Terminology and Notation Some terminology important to composite materials follow: Isotropic—Possessing the same mechanical properties in all directions, Composite laminates are never isotropic. Laminate—A material consisting of layers (laminae) bonded together. Transversely Isotropic—Possessing one plane that has the same mechanical properties at any direction in that plane, i, the laminate will have the same stress-strain behavior at any direction in the plane of the material (sometimes called quasi-isotropic). Orthotropic—A material that has different mechanical properties in three mutually perpendicular planes. Note that the properties of the material are direction specific in this case, All unidirectional laminae are individually orthotropic. Most laminated composites fall into this category.Homogeneous—Material properties do not change from point to point within the material. Since filamentary composites consist of at least two distinct phases (fiber and matrix), laminated composites are never truly homogeneous, although on a macroscopic scale when discussing the linear elastic response (no damage) of laminated composites the material may be generalized as homogeneous. This assumption is termed “smearing” of fiber and matrix. In reality all composites are heterogeneous. Principal Material Directions—Directions parallel and perpendicular to the fibers in a lamina. Note that these directions are not necessarily the directions of principal stress as defined by continuum mechanics. Balanced Laminate—For each +0 ply in the laminate there is an equally thick -6 ply in the laminate. This does not apply to 0° and 90° plies. Symmetric Laminate—The plies of the laminate are a mirror image about the geometrical midplane, Angle Ply Laminate—Containing plies oriented at angle(s) other than 0° or 90°. ‘The notation used throughout this report denotes the directions parallel, perpendicular, and through-the-thickness to the aligned fibers of a ply of material as 1, 2, and 3, respectively. The coordinates of the laminate are denoted by x,y (in-plane) and through-the-thickness z. Some texts and laminate analysis computer software have reverse notation, which is exactly opposite to that presented here. Some users prefer to use a primed and unprimed notation. Thus, it is important to note which notation is being used. Since a state-of-plane stress is assumed, the term “plate” will be interchangeable with material and composite laminate. Please note that in actual practice, the laminate does not need to be a plate, but may be a shell or other shape such that the material is considered “thin.” C. Summary of Sections In section II, the behavior of an individual ply or lamina is considered. This behavior is the “building block” upon which laminated plate theory is based. The important concept of coupling, ‘unique to anisotropic materials, will be introduced in this section. The main body of this report is considered in section III. It develops the relationship between loads and deflections for a composite laminate. Most information on the elastic response of laminated composites can be calculated from the equations given in this section. The constitutive equations that govern the laminate load/deflection behavior are given in this section. For those who have not dealt with laminates, section IV shows how stacking sequences of the plies that make up the laminate are denoted. This section simply presents a standard “code” by which a laminate can be described. Many will already be familiar with this material. Section V contains derivations of the most sought after numbers when dealing with the elastic response of these materials, the engineering constants. Many texts on laminated composites omit this part, or only present it for special types of laminates that make the calculation relatively simple. The engineering constants for any laminate can be found from the equations in this section.Section VI introduces the effects of temperature and moisture on the strain of a composite laminate. These effects are often neglected, but are very important in determining the stresses and strains within each ply of the laminate. Determining these ply stresses and strains for symmetric laminates is presented in section VII and in section VIII for unsymmetric laminates. These sections provide information necessary to study the strength of composite laminates. Section IX is a summary of the most important equations presented in this paper. I. GENERALIZED HOOKE’S LAW FOR NONISOTROPIC MATERIALS A. Normal Stress and Strain, Uniaxially Applied Force Normal stress is defined as the force per unit area acting perpendicular to the surface of the area. The corresponding strain is defined as the elongation (or stretch) per unit length of material in the direction of the applied force. For isotropic materials, the relationship between stress and strain is independent of the direction of force, thus only one elastic constant (Young's modulus) is required to describe the stress-strain relationship for a uniaxially applied force. For a nonisotropic material, at least two elastic constants are needed to describe the stress-strain behavior of the material. Figure 1 is a schematic of an isotropic and a unidirectional fiber-reinforced material. The stiffness of the isotropic plate can be described by one value, the modulus, E, of the material, regardless of direction of load. The stiffness of the orthotropic plate must be described by two values, one along the longitudinal direction of the fibers, commonly referred to as Ey, and one transverse to the direction of fibers, usually denoted by Er. Subscripts 1 and 2 will be used such that Ey = E; and Er = Ep. Thus, indices must be added to the stress, strain, and modulus values to describe the direction of the applied force. For example, for an isotropic material, the stress/strain relationship is written: sEe. qa) Isotropic Plate Orthotropic Plate 1 1 Reinforcing fibers aligned in I-drection 2 2 stiffness in 1-direti iffness in 2-direction stiffness in L-direction >> stiffness in = stiffness in any direction 2-direction # stiffness in other directions Figure 1. Difference between an isotropic and an orthotropic plate.For the orthotropic system, the direction must be specified. For example: O,=E,e, oF 0,=E 2 - (2) If the applied load acts either parallel or perpendicular to the fibers, then the plate is considered specially orthotropic. B, Stress and Strain, Plane Stress for Specially Orthotropic Plates The previous section dealt with an extremely simple type of stress state, uniaxial. In general, plates will experience stresses in more than one direction within the plane. This is referred to as plane stress. In addition, Poisson's ratio now becomes important. Poisson's ratio is the ratio of the strain perpendicular to a given loading direction, to the strain parallel to this given loading direction: & as @) Poisson’s ratio = v, or vy=—=— - io a & a & & For loading along For loading the fibers perpendicular to the fibers The strain components are now stretch due to an applied force, minus the contraction of Poisson’s effect due to another force perpendicular to this applied force. Thus: (4a) Using equation (2): >: (4b) Shear forces can also be present. Shear stress and shear strain are related by a constant, like the normal stresses and strains. This constant is called the shear modulus and is usually denoted by G. Thus: t= %2Gi2 - (5) Where 72 is the shear stress (the 1 and the 2 indices indicating shear in the 1-2 plane), and 2 is the shear strain. Figure 2 gives a definition of shear strain. Since it is known that a relationship exists between Poisson's ratios and the modulii in each of the two axes directions, namely: VE i= VE 2 » -)thange in right angle of element Defined as "Engineering Strain" @ Defined as "Tensorial Strain” () Figure 2. Definition of shearing strains.Equations (4b) and (5) can be written in matrix form as: 4} [SuSn 0 |] % &|=]Si252 0 || a) Ya] [9 9 S66}] te where, So9= E> (8) Seo= UG Note that at the 3,3 position in this 3x3 matrix (called the compliance matrix), the subscripts are 6,6. This evolves from a detailed treatment of arriving at a constitutive equation for an orthotropic material from an anisotropic one. By inverting the compliance matrix, one can get stress as a function of strain. This turns out to be: %) |QuQn 9 4% %|=|Q20n 0 |le} (9) t| | 9 9 Qe) y> where: Ey On=—— : 1-VQVo1 (10) Q65=Gir - ‘The Q’s are referred to as the reduced stiffnesses and the matrix is abbreviated as [Q). C. Stress and Strain, Plane Stress for Generally Orthotropic Plates Now suppose that the unidirectional lamina in figure 1 is loaded at some angle other than 0° or 90°. The lamina is now referred to as generally orthotropic, since, in general, the loading direction does not coincide with the principal material directions. The stresses and strains must now be transformed into coordinates that do coincide with the principal material directions. This can be accomplished using the free-body diagram in figure 3. From free body diagram (a), and summing forces in the 1-direction:Figure 3. Generally orthotropic lamina. EF, =0= 6, dA-o,(dA cos cos @ — 0,{dA sin 6) sin 0 ay ~ ta{a4 cos 6) sin 9 — 1, (dA sin 6} cos @ From the free body diagram of figure 3(b) and summing forces in the 2-direction: BF) =0= 0, dA-o,(aA sin 9) sin @-0,(dA cos 6)c0s 8 (12) +1, {dA cos asin @ + 2,{ dA sin 6)e0s @ From free body diagram of figure 3(b) and summing forces in the 1-direction: EF, =0= 1,glA+odA sin 6) cos @-o,(dA cos 6) sin (13) ~ty{dA cos 6) cos 6 +2, aA sin d)siig equations (11), (12), and (13); 0,00" 6+0, sin? 942r, sin 8 cos 6 , @, Sin? 949,008" 8-2,,sin Bos 8 , (4) 44379, Sin 8 cos B+0, sin Boos O+7,, (cos? @-sin” 6) . Equation (14) can be written in matrix form as; 2 in? ; o, cos? 6 sin?@ —-2sin @cos@ || a, o,|=| sin’ cos’@ 2 sin cos @ |} o,| (1s) ti] | -sin@cos@ sin @cos@ {cos @~sin? ) ty ‘The 3x3 matrix in equation (15) is called the transformation matrix and is denoted by [7]. The same matrix is used to transform strains. Note that the tensorial shear strain must be used, not the engineering shear strain, when transforming strains. This arises from the geometrical considerations that the amount of shear must be equivalent with respect to both the x- and y-axes, since these axes will be transformed into new ones (fig. 2(b)). If it is desired to transform from the 1-2 coordinate system to the x-y coordinate system, the inverse of [7] must be found. It is given by: cos’ @ sin’@ 2 sin 8cos 8 (7y'- sin? 9 cos” 6 2sincos@ | . (16) sin @cos@ —-sin 9cos 8 (cos? -sin® 9) Thus; % o, oO a %,|=(7]|o,} and | 9,)=[77"| %] - a7) TR Ty by TaSimilarly for strain; & & & a -1 s| and | 6 [=o] e| - as) by by fn Putting equation (9) into the second part of equation in (17): ic cr 100} % 6, \=[7)" [el] & |=[7}[a]}o10]] | - a9) fy te o02}le Now putting the first equation of equation (18) into equation (19): = S S Q 100 ¢,|=[7}'[ rts : (20) Defining a new matrix called the lamina stiffness matrix (sometimes called “Q-Bar”) as: 100 (2]=[7F'fal}o1 off] « ay 002 and letting: m=cos 6 , the components are: Dir =Qrsm'+2(02+20g5)m*n® 40m" , Bi2=(On+0rz-40 eg) m'n®+Oy{m'en') , Bro=O1sn*42(O12+20 ee) mn+Onm* , (22) Brg=(O11-Q12-2065) mn + (Q12-a2+20} mn? ,Bas=(Q11-O 12-20) 2m +(Q12~Qra¥20 5) nm, (22) (cont.) Bes= (Qirt 20-20 12-20) m'n®+Den*n4) - Note that if @ is any angle other than zero, there will be nonzero Jy, and J, terms. Putting this into equation (20): S (23) %| | Qu Q12Qr6}| & 9,|=| O12 O22 26] &, | » Fy} [BisDe6Be|] x, it can be seen that a shear strain will produce normal stresses, and normal strains will contribute to a shear stress. This is referred to as extension-shear coupling and will take place in a lamina that is loaded at an angle to the fibers (other than 0° and 90°). That is, there will be coupling if the Tg and/or Og terms in the lamina stiffness matrix are nonzero. D. Invariant Stiffnesses The “Q-Bar” terms can be written as B11 = Uy+U; cos (26)+U5 cos (46) , Dy. = Uj-U; cos (26)+U; cos (46) , G,2= Us-Uy cos (46) , Des = Us-Us cos (46) , D=> Uy sin (26)+U; sin (46) , (24a) One $e: sin (26)-U; sin (48) , 10where, =2(01tOn)+t Ont Qos» = 3 (11-22) » 1 1 *ealemcels te aee- (24d) Note that only U and Us are coefficients to the sine or cosine terms in equation (24a). This implies that when calculating the Q-Bar values, Uj, Us, and Us are independent or invariant to the ply orientation @, This concept of “invariant” quantities can make some calculations easier. This paper will not go into detail on this subject since only the basics are being presented. III. MECHANICS OF LAMINATED COMPOSITES A. Assumptions The following assumptions are made for the remainder of this paper: (1) The laminate thickness is very small compared to its other dimensions. (2) The lamina (layers) of the laminate are perfectly bonded. (3) Lines perpendicular to the surface of the laminate remain straight and perpendicular to the surface after deformation (4) The laminae and laminate are linear elastic. (5) The through-the-thickness stresses and strains are negligible, These assumptions are good ones as long as the laminate is not damaged and undergoes small deflections, B. Definitions of Strains and Displacements A displacement of the plate in the x-direction is designated as u. For the y-direction, it is designated as v and for the z-direction w. Figure 4 shows these displacements. The strains are now defined as: MWy ™ ! | Normal displacement Bending displacement Figure 4, Displacements of a plate. av au av ees: roe(5 3] : (25a) The slope of the plate if itis bending is given as: 2 stong the x-direction , & (25b) ee Cy The total in-plane displacement at any point in the plate is the sum of the normal displacements plus the displacements introduced by bending. Denoting the displacements of the midplane of the plate for the x and y directions as u, and V, respectively, with the help of figure 5 the total displacements are: along the y-direction . ow ow wauce sf vers. 26) a Note that for figure 5, x can be replaced by y when u is replaced by v (ie., the view could be from any side of the plate). It is assumed that there is no strain in the thickness direction, only a displace- ment. 127 Undeformed edge of plate Top surface of plate Bottom surface of plate @=Slope= 2% For small 0 Displacement along z-direction due to bending is sind times z Since 0 is small, sind = 8. ‘Therefore the displacement is = 20 where negative is compression and positives tension | \ ce Figure 5. Total displacements in a plate. From equations (25) and (26): Qn 7 ou Ny ye Defining: Quy Woo ° SSS ashe (28) to be the midplane strains and defining: aw aw . aw ~ GEEK ei md ts 5 29) to be the plate curvatures will make notation easier. 13Equation (27) can now be written in matrix form as: & e K, g,|=| 6 |+2| x, 0) ty] |e) [Ke From figure 6, it can be seen that the plate curvature K;, or Ky is the rate of change of slope of, the bending plate in either the x- or y-direction, respectively. The plate curvature term Kyy is the amount of bending in the x-direction along the y-axis (ie., twisting). sope = 3% axe xt x Stopee SH atx Rate of change of slope across x-direction = Rate of change of slope across y-dreation = & Ky Figure 6. Definitions of plate curvatures. 14From equation (23), the stresses in each ply of the laminate can be determined with equation (30): ee 2 Se %! |QunBrO6||*| |G Gi2 A164] Ks 6, |=] D122 ae] €) | +2| O12 Brn Dre]| Ky (30a) 14] [Bis@2e] P| [Brs262ec|| hy C. Definitions of Stress and Moment Resultants Since the stress in cach ply varies through the thickness of the laminate, it will be convenient to define stresses in terms of equivalent forces acting at the middle surface. Referring to figure 7, it can be seen that the stresses acting on an edge can be broken into increments and summed. The resulting integral is defined as the stress resultant and is denoted by Nj, where the i subscript denotes direction. This stress resultant has units of force per length and acts in the same direction Total force in x-direction = 2 ox(dzy) 2 As d2—90, E enphop] 6, dz hi ‘h/2 N= 6x dz h/2 Figure 7. Definition of stress resultant. 15as the stress state it represents, Figure 7 could also be drawn for the y-direction stress and shear stress. The three stress resultants are therefore: We Naf od. GN) We Ny | to! As can be seen from figure 7, the stress acting on an edge produces a moment about the midplane. The force is o, (dz)(y) as denoted in figure 7. The moment arm is at a distance z from the midplane. Following the same procedure as for the stress resultants, the moment resultants can be defined as: rs me f ode . Li % we M, f nt (32) My fe tyede These moment resultants have units of torque per unit length. The directions for all of the stress and moment resultants are shown in figure 8 for clarity. The double-headed arrow indicates torque in a direction determined by the right-hand-rule (.e., point the thumb of your right hand in the direction of the double-headed arrows and the direction of rotation of the torque is in the direction that your four fingers are pointing). Note that M, and M, will cause the plate to bend and My will cause the plate to twist. As an example of the relationship between stress and stress resultants, if a tensile test specimen is 2.54-cm (1-in) wide and 2-mm (0.08-in) thick, and is pulled on with a force of 4,500 N (1,000 Ib), then the average stress on the cross section is: 4,500 N ©, = > = 88.6 MPa = 12,500 (33) (0.0254 m) (0.002 m) i ‘The stress resultant would be the average stress multiplied by the specimen thickness: = (88.6 MPa) (0.002 m : (34) 16Ny Ne My My Figure 8, Direction of stress and moment resultants. Since 1,000 Ib was applied over 1 inch of specimen length, the expected result of 1,000 Ib/in is obtained. Each individual ply of the specimen may have a stress other than 88.6 MPa (12,500 Ib/in?), but the average stress will be 88.6 MPa (12,500 Ib/in?). D. Constitutive Equations for a Laminate Putting equation (31) in matrix form: (35) fyand putting equation (32) in matrix form also: My, o% W M, fr 9, ede (36) My by The integrals in equations (35) and (36) must be performed over each ply and then summed, since discontinuities in stresses can occur at ply interfaces. Using the schematic of a laminate in figure 9, equations (35) and (36) must be written as: Ny rs “he x o,| a, (G7) Ny ria | Vey a and, M, x hk M, flo] - (38) hea) * My ty Now equation (30) can be substituted into equation (23), which can then be substituted into equations (37) and (38) to give: o ,, [O82 Br6l | & | meee || by | 212222 Doo} | & | ae "0 16O26Bos}, |? Ik | Yay 21212016] | Kx , (39) "hy +f. DiDnDre) | x, [eae Rlegece DieDrs Des}, | Ky and, 18Ply Plym ht hha | Piy2 Ply 1 [Note that ply k and ply K+ are the same lamana (layer), but are separated into two plie by dhe geometric midplane Figure 9. Cross section of a laminate. _ ate y, [211 F12Dr6) | & 4 eet a, [212922 Da} | ede B15 B25 Oe, 4? Dir DirDr6] | Ke ao ny [22nd : +f" |Qi2D22026} | K, fe ae nore DisDrsDes|,| Ky 19Since the middle surface strains and curvatures (the €"s and K’s) are not a function of z (because these values are always at the middle surface z = 0), they need not be included in the integration. Also, the laminate stiffness matrix is constant for a given ply, 80 it too will be a constant over the integration of a lamina thickness. Pulling these constants to the front of the integral in equations (39) and (40) gives: _ ote Ne] on |{ Gi Gi2 Gre] | & zaanagsz of} phe N%|=¥,(|OnDra Dal |] ff" ae My] |[OreDaeBes} | 2 aS , 41 BBB] [Xe], “© +| 12D Dr6| | Ky he D6Br6 ss}, [Ky and az a =m a My) on |] Qu Q12Qi6] | “= eras ol} ste My l= % (0128200) |e], ade My| ** || O16 O26 es], |? a SS 42} Bubs) [X] 4. , ” +|2n2:0o| |x, | f™ fee BisDrs Ove], | Koy Performing the simple integrations gives: aznazlle Br Bi2Br6] | & i222 D6} | 69 | (ha-he) O16 OO, 9 DisDrsBe|,| (43) BuBuBul [Ke], +] B22 Bas | Ky 15 (He-mea) BBs) Ke and 205 M, B80 ||, , eu eats ls M, 212005) | & 2 (mH) My ZieBrBes),| BuO20r6] [Ke Gd 91020 Df +1822 25} | K, | (ri-¥e) BisBas Deel [Ky Since the middle surface strains and curvatures are not a part of the summations, the laminate matrix and the hy terms can be combined to form new matrices. From equations (43) and can be defined as: (45) (46) (47) N 1 ef x] {Ani | Bu Bi Bisl| 5 Ny | |ApAnA26 | Biz Br Br6|| Fy Nig | fAtsAasAes | Bie Bas Beal] 2, --"]=|-- -- -- | -- -- --||-] (48) M, By B2Bie | PuPrPi6 Ky My | |Bi2 B22 B25 | DizD2Da6|| Ky My| | B16 B26 Bos | Dis Drs Do|| Ky Written in contracted form, equation (48) becomes: AIB oo |e (49) BID|LK 2This can be partially inverted to give: [| . (50) where, [e‘}=-LA ta] . (51) (c']={2)4l° [']-[01-Bl4re] The fully inverted form is given by: a) [AndinAiel Bi Hie] 5 | [Arma | Bio Bas Bas|| 9 | [Auton Aes | Bic Ba Be eae yon lM, | > (52) Ke] [On Ci2Cis | Pu Dia Pioll kK, Cian Cae | Pra DD My C16 Crs Cos | Dig Dag Des where, oa ]=[]-e Tic). apo 2'][0'] (52a) [PT 'lc’). The fully inverted form is the most often used form of the laminate constitutive equations. 22For symmetric laminates (laminates that are configured such that the geometric midplane is a mirror image of the ply configurations above and below the midplane), the geometric midplane is also the neutral plane of the plate, and the [B] matrix will have all elements equal to zero (as will be shown later). However, if the laminate is unsymmetric, i.e., if the plies near the bottom of the plate are much stiffer in the x-direction, then the geometric midplane will not be the neutral plane of the plate; and the neutral plane will be closer to the bottom of the plate for x-direction bending as shown in figure 10. This is accounted for in the constitutive equations, since the {B] matrix will have some nonzero elements (as will be shown later), implying that a bending strain (plate curvature) will cause a midplane strain as depicted in figure 10. Likewise, a midplane strain will cause a bending moment. A method to find the neutral axis of the plate will be discussed in a later section about stresses within the plies of a laminate. . Undeformed edge of plate “op nude tp ‘Vey aif pes oto sce of pie Deformed edge of pate Pre bending cos a cmpesson of rome! ipa Figure 10. Displacements in an unsymmetrical plate. E. Physical Meanings of the [A], [B], and [D] Matrices Recalling the definitions of the (A], (B], and [D] matrices, Elk (orbs) « Ly obbi-n,)« ail(ot-s2,) « (45) (46) (47) 23and referring to figure 9, it can be seen that the last term in equation (45) is the kth lamina thickness which will be denoted by fy. Thus, equation (45) can be written as: y= [Bilets - (53) This matrix is called the extensional stiffness matrix. From the constitutive equation (48), it ‘can be seen that these terms relate the normal stresses and strains (much like the modulii of elasticity), except for the Ay and Ag terms which relate shear strains to normal stresses and normal strains to shear stresses. Thus, when Ayg and Ag are nonzero, and the laminate has a shear strain applied to it, normal stresses will result and vice-versa. These terms are analogous to the Qig and Qag terms mentioned in the final part of section Il Equation (46) can be written as: 1 1S lr moa) bvehes)=¥ aye ee B, ii where fy is the thickness of the kth ply, and (/4,+hy_1)/2 is the distance from the geometric midplane to the center of the kth ply. This matrix is called the coupling stiffness matrix. From the constitutive equation (48), it can be seen that these terms relate bending strains (plate curvatures) with normal stresses and vice-versa. The Bj and Bag terms relate twisting strains to normal stresses and shear strains to bending stresses. If the laminate is symmetric, then the By terms will be the same for each mirrored ply above and below the midplane, with the exception of the sign of the (h+H,.1)/2 term being negative if itis below the midplane (-2) and positive if itis above the midplane (42). Thus, when summed, the result will be zero for all By. Now define: (a ith gy 2 Part of equation (47) can be writen as: (ihm) = [ees (eache = [lm res) +30, 3h eh] = [lec tea) +3(ehas)(rath LU = A(nj-n,)=@+128, ee el (55)Therefore, equation (47) can be written as: Dy=> [8s] aia), (56) It can be scen that the last term is the second moment of the kth ply with respect to the geometric midplane. Dj is called the bending stiffness matrix and relates the amount of plate curvatures with the bending moments. IV. NOMENCLATURE FOR DEFINING STACKING SEQUENCES A. Coordinate System The choice of coordinate system used for the laminate will determine its stacking sequence. For example, a unidirectional composite can be said to be made up of all 0° plies in the x-direction, and all 90° plies in the y-direction. Alternatively, the same composite can be said to be made up of all 0° plies in the y-direction, and all 90° plies in the x-direction. The composite can also be referenced by any other x-y coordinate system in the plane of the plate. The choice of coordinate system is totally arbitrary, but some general procedures are usually followed to make calculations or ‘communication about the laminate easier for others to understand, A coordinate system is almost always chosen such that one of the axes runs in the direction of fibers of one of the plies of the laminate. This will make analysis much easier. The x-axis is usually chosen as the “longitudinal” axis, with the corresponding y-axis being the “transverse” direction. This is similar to what was defined earlier for laminae. The main load bearing fibers (if these are known) are usually called the 0° fibers, longitudinal fibers, or x-direction fibers. The other ply orientations will then be defined with this coordinate system. An example of a typical coordinate system is given in figure 11. B, Nomenclature There is more than one way to denote the stacking sequence of laminates. However, once one method is leamed, any other is easy to interpret, even though it may not be in the form that the user is accustomed to. ‘Once the 0° fiber direction has been defined (and thus the x-axis), the plies that are not at 0° must be assigned an angle. To do this, start from the x-axis and rotate to the fiber direction of the ply being defined. Clockwise rotations are positive angles, and counterclockwise rotations are negative angles, although the reverse can also be used since only plane-stress is being examined for plates and the material is the same whether viewed from one surface or the other surface. Now that all plies have an angle associated with them, a method of presenting the stacking sequence follows. If the laminate is symmetric, then start with the angle of the outermost ply and write the ply angles, separated by a comma, until the midplane is reached. Enclose this string of angles in brackets or parentheses and subscript the brackets or parentheses with an “S” to denote “symmetric.” If the laminate is not symmetric, then proceed as above until the bottom ply is 25Direction of primary loading oes ee Fiber orientation of the 8 plies as seen from the primary loading direction edge to bottom is now 0°, o 90°, 90°, -45°, 445°, y ‘The stacking sequence from top -45°, Figure 11. Coordinate system for a typical laminate, 26reached. Subscript the brackets or parentheses with a “7” to denote “total” laminate. Referring to figure 10, this laminate is denoted as [0, +45, -45, 90]s . This is more convenient than writing [0, +45, -45, 90, 90, -45, +45, O}r. Further simplifications can be made when two or more plies of the same orientation are grouped together. The angle of these plies need only be written once with a subscripted number denoting the number of plies in the group. For example, 0, 90, 90, 90, 90, O}r can be written as [0, 904, Olr. Since this laminate is symmetric, further simplifications can result, and this laminate could be described by [0, 902]s. If a symmetric laminate consists of an odd number of plies, then the geometrical midplane of the laminate will lie at the midplane of the center ply. In this case, a bar is placed over the angle of this ply to denote that half of it resides in the top half of the laminate and the other half resides in the bottom half of the laminate. For example, a laminate with stacking sequence {0, 90, 90, 90, O}r can be written as, {0, 90, 90}. Any repeating units within the laminate can be placed in parentheses with a subscripted number representing the number of repeats. For example, a {0, 90, 0, 90, 0, 90, 0, 90}s laminate can be written as [(0, 90)4]s. If adjacent plies are of the same angle, but with different signs, then a plus-minus sign is usually placed in front of the angle of the plies. For example, a [0, +45, ~45, 90, +30, ~30]7 laminate can be written as (0, +45, 90, + 30]7. Some examples of stacking sequences and how they can be denoted are given in figure 12. a5 a 5 90 0 45 a (445, 90, -45}5 Sods S| 10,285)5 a 0 “45 0 ° a5 oo ° 0 5 ° vs 48 90 a5 90 ° 0 ° 0 5; 90 | (445,90, 0, 50} 0 | 10.9020 ~45_ | (G45, 0} 0 0 ° 0 ° ry 5 0 4s 5 0 a Figure 12. Some laminate stacking sequences and their notation. 21V. IN-PLANE ENGINEERING CONSTANTS FOR THE LAMINATE A. Orientation of the Laminate For a given stacking sequence of laminae whose engineering properties are known, it is possible to determine the in-plane engineering constants of the laminate from the Ajj matrix for symmetric laminates, and the Ay, By, and Dy matrices for unsymmetric laminates. The choice of coordinates will determine the directions of the laminate engineering constants being evaluated. ‘These directions are arbitrary, but are usually chosen as described in section IV. A. B. Symmetric Laminates Recall that for symmetric laminates, the By matrix consists of all elements being zero. This greatly simplifies finding the in-plane engineering constants of the laminate. To find the x-direction modulus, the value of the x-direction stress to the x-direction strain must be calculated. In equation form: o, NJh = . (57) & & where h is the thickness of the laminate. Since the By’s are zero, the constitutive equations are: o Ny Andris i Ny |=|A 124224261] &, (58) Ny] [AisA26Aco|| ? hy Since a relationship between Nx and & are being sought, when a load is applied in the x-direction, from equation (58): Oe eeORr nO N= AE, +A 26,44 16% + (59) OR Ane +A 6 tA 6h « (60) O=A see tAne6,*Agoley « (61) From equations (60) and (61): = (62) 0 _ ,0[ A26A16-A 2.66 6 =&(—— >| + = 2 AnAge-Are 28and 2 Arg Ar6A12A6e-A 26 A 16 1 (63) Ae Age ~Ar6 Ao Equations (62) and (63) can be substituted into equation (59) with the result: 2 ‘N, AAA pA -A A2¢A 1A g5-A 6 A & AnAge-Ar 66 ArAgs -Ar6 Ace Thus, E, can be calculated by dividing equation (64) by the thickness of the laminate, A which will give equation (57). The same procedure is followed to obtain Ey. The con: tive equations are: OR Ane tA 2644 chy + (65) Og ug a? Ny=A 2+ nb +A rey + (66) on AaiArdenen co) From equations (65) and (67): (68) and (69) Equations (68) and (69) can be substituted into equation (66) with the result: 2 ~Ar Ais 1rAc6-A 16 A 26 rapt | <2 See ae Mas (70) 66 AAs Ai6 Ace 7 70 A191 2 & Aj Aes-A 16 Dividing equation (70) by the laminate thickness will give E,. 29Gay is found in the same manner, The constitutive equations are: Oy eeOn ceo O=Ane Ae +A hy + DRA nee tAneytAreny + ea ea a? =A 166,44 266,44 66% ey From equations (71) and (72): 0_ 0] A12A26~A 16422 &= 2 AyAnA iz and 2 0» Are | AwArrAn-A 12. Ar6 * An AyAn’-Ay'An Equations (74) and (75) can be used with equation(73) to obtain: 2 — Neg Ae | 2AraA agar ia'Aae Ae Ans 9 =A66~ 2, 2, Nyy An AiAn -AiAn Dividing equation (76) by the laminate thickness will give Gy. To find Poisson’s ratio of the laminate, use equations (60) and (61) to obtain: Rearranging to get: 30 (71) (72) (73) (74) (75) (76) (77) (78)Using equations (65) and (67): O=Ayer+A bond (79) Rearranging to get: AAr6 0 -2 | Aes Vyp=— = (80) & Ais Example 1: Suppose a 4-ply laminate of AS4/3501-6 is laid up in a [0,+45]s stacking sequence. The ply properties are given below for this material: Ey = 20,010,000 Ib/in? , Eq = 1,301,000 Ibfin? , Gia = 1,001,000 Ibfin? , v2 =03 , va = 0.02 , Ply thickness = 0.005 in . This is all of the information needed to calculate the in-plane engineering constants. Calculating Ex will demonstrate the general procedure applied to any of the engineering constants. From equation (10): Ey __ 20,010,000 ibfin in? 1. = 20,130,785 Ib T=¥y¥q,__1-10.3)(0.02) m Qu 301,000 Ibvin® = 1,308,853 Ib/in” , 1-(0.3)(0.02) (03(4,301,000 Ibn”) EL) = 392,656 Ib/in” , 1-(0.3)(0.02) Qp5= 1,001,000 Ibfin? . From equation (22), for the 0° plies: 31= 392,656 Ibvin® , 3p = 1,308,853 Ibiin? , Di=0. (E1.2) 23,=0. Dio 1,001,000 Ibvin? . For the +45° pli it = (20,130,785 wwin?x0-107 9 52,656 Ibin?+2(1,001,000 ini?) }o.r079%0.700? + (1,308,853 Ibvin?\(0.707)* = 6,557,237 Ibfin? , oy = (cossozs Ibfin?+1,308,853 Ibvis 4.00100} 070% 07078 +392,656 Ibvin® (0-707 i*+00-707') = 4,555,238 Ib/in” , 54° = (20,130,785 tia?x0-107) 92686 n2¢,001,000 wins hoon 07008 +(1,308,853 Ibvin?)(0.707)* = 6,557,237 Ibvin? 2 1,308,853 Ibvin?+2(1,001,000 lbvin”) }(0.707)(0.707)° = 4,705,483 Ibvin? , oe = (coaanzss Ibvin?~392,656 Ib/in—2(1,001,000 Ibvin )JoronXa + (sv2156 Ibi } Ons = (2onanzes Ibfin?~392,656 Ib/in?-2(1,001,000 Ibvin?) (0.707)°(0.707) + (205 To/in?~ 1,308,853 Ibfin?+2( 10010019? f.rr90700) = 4,705,483 Ibfin? Ore = (zusozs Ibfin?+1,308,853 lb/in?~2(392,656 Ib/in?+ 1,001,000 Ibvin } (0.707(0.707? +1,001,000 bn? ((0-707/8«0.207) 163,582 Ibvin® . (E13) 32Now from equation (53), the Ay are obtained: Ib Aq, = [20,130,785+6,557,237 +20,130,785+6,557,237]lb/in” (0.005 in) = 266,880 me i 1 Ap = (392,656+4,555,238-+392,656+4,555,238] Ib/in” (0.005 in) = 49,479 a : ‘Ag = [1,308,853+6,557,237 + 1,308,853+6,557,2371Ibfin? (0,005 in) = 78,6612 , mn 1.4) Ib A 1g = [0+4,705,483+0+4,705,483] bin” (0,005 in) = 47,055 ae i A= (0+4,705,483+0+4,705,483] b/in® (0.005 in) = 47,055 = , in Ib i A eg= {1,001,00045,163,582+1,001,000+5,163,582}IbVin? (0,005 in) = 61,6462. in From equations (64) and (57): 266,880 Eat (4979) '55}(47,055}-(49,479)(61,64 (78,661](61,646)-(47,055) Ib ~ (47,085) _| (47,055)(49,479}(61,646)-(47,085) (47,085) || in (orcas) * (78,661)(61,646) -(47,055) (61,646) | |0.02in + (47,055) (E15) E,= 11,333,000 Ibvin” C. Nonsymmetric Laminates Since for a nonsymmetric laminate there are nonzero B,; terms, the calculation of in-plane engineering constants become more involved; however, the same basic procedure is followed as for symmetric laminates. Since there are now six equations instead of three, it will be much easier to use matrix notation. The constitutive equations are given by equation (48): 33222 r= y ay = To find E,, only the x-direction in-p 1 AnAnAr | Bu Br Bis Ap An Ar | Ba Bm A164 26 66 | Bis Bas Bos Du Dy2Dig Diy Dy Do Dig Dag Dos 1 a pp By By Bye | By Bo Brg | Bie Bas Bos | sought. The constitutive equation now becomes: Using Cramer’s rule to solve for e : An AA | Bu Br Bis AA Ar | B12 Bx Bag Ae Are Mee | Big Ba Bes Diy DyDig Dy, Dy Dr Dig D5 Dos a By Bi Bie | By By By | Big Brg Bos ! MN An Aig 0 An Ar 0 Ar Ace By By Bi6 Bi Br Br Bis Brs Bes 0 By Bis 0 By Brg 0 Br Bos Du Di2Dr6 Dy2D2Drg Dip Dag Dog 34 An AA Bu BBs ArrAnAr Bi BnBr Ai6A26A6 Big B25 Bos By By2Bys Dy Di2Di6 By2B2By5 Dy2D22Dag By Brs Bos Dig Dag Do5 (48) Jane load is applied and a relationship between Ny and e® is & 0 7, (81) (82)Determinants for two 6x6 matrices must be found. At this point, it should be clear why a computer (or calculator) program makes calculations much easier. To write out the solution for equation (82) would require 6! = 720 terms for the numerator and the denominator. Cofactor expansion can be used in the numerator for some simplification: AnAr% BiB Br ArwAgs — Big Bas Bos Bi2B yg Bis Dy Dia Die B22 B25 B25 Di2 D2 Dr Br B65 Bos Dig Dr5 Dow AnA2Aie Bu Bi2Bie ApAnA25 Br Bx Brs A Ar As B15 B25 Bos By ByBys Du Di Dig By2 Bx By5 Di2D Dre By Bap Bos Dip Drs Dos From equation (57), Ey can be found by evaluating: Ay Ans Bu BB. A222 A 26 By2 Bop Bao AwAr6A 65 Bis Bas Boo By By2Byp Dy Dy Dig By2 B22 B26 Dig Dx Dag By6 B25 Bes Dig D6 Doo Ax By By Acs Bis Bas By Dy Dy By Diy Dy Bos Dig Dag (83) (84) Appendix A shows a Fortran program to calculate the determinant of a 6x6 and 5x5 matrix. 35E, can be found in a similar manner. The denominator will be different since equation (81) is being solved for £8 An Ai2A 6 Bu Bi2Bis Ai.AmA26 Biz Bx Brg 4164264 66 B16 Bae Boo By By2Bis Du DrDis By2 Bp Bag Dy Don Dog By6By6 Bes Dis Drs Des (85) i sie Ais Bu Bi Guy will be given by: AnApAr Bu Bi2Bis AnAnAr6 BirBr Br AwAr Ag BisBr6 Bos By ByBys Di DiDie By, B22 Bag D12D22 Dag Bi6Br6 Bos Dig Drs Dos (86) >is 42 Bu Ba Bis An Bn Bn By 36Poisson’s ratio will be determined as it was for symmetric laminates. For Vs, the contraction in the y-direction upon an applied s Thus, vsy will be given by: Vx is given by: ess in the x-direction must be obtained. This is given by: An Aw Bo Bx Bos Ais Ave Bis Brg Boo. -NIBy By Du Pa Dis By Bx Dn Dn Drs Lr An AAs Bu BBs AnAn Ar Br. Br2 Bas AywArAos Bie Bas Bos By BipBis Dir DD By, Br By Di.Dz2 Dis Bis B26 Bos Dis P26 Dos, An Aw Bo Bx Br Ais Aw Bis Bas Bos By By Pu Pp Dw By By Dp Dn Dy Bis Boo Die Das Dow An Ax Bo Ba Bas Ar Acs Bis Bre Bes By By Dy Dy Dis Bn By Di Dy Dy By Beg Dig Dr Dos (87) (88) 37-€, a (89) Example 2: Given a 2-ply laminate of stacking sequence [0,+45]7, E will be calculated. From example 1, the Aj’s are known, The By’s and Dy’s must be calculated. To obtain the By, use equation (54) with the already determined values in equations (E1.2) and (E1.3). By, = 20,130,755 Ibvin? (0.005 in)(0.0025 in) + 6,557,237 Tbfin? (0.005 in)(-0.0025 in) = 170 Ib , B yy = 392,656 Ibvin? (0,005 in)(0.0025 in) + 4,555,238 bin? (0.005 in)(0.0025 in) B yp = 1,308,853 Ib/in? (0.005 in)(0.0025 in) + 6,557,237 Ibvin® (0.005 in)(-0.0025 in) = -66 Ib B yg = 044,705,483 Ib/in” (0.005 in)(~0.0025 in) = -59 Ib , By = 0+4,705,483 Ibvin” (0,005 in)(-~0.0025 in) = -59 Ib , Bog = 1,001,000 Ibvin? (0,005 in)(0,0025 in) + 5,163,582 Ib/in” (0,005 in)(-~0.0025 in) = -52 Ib -521b , (E2.1) Use equation (56) to find the Dj's: 383 2{ (0.005) 2.3 Dy = 20,130,785 lbfin® |“ + (0.005 \(0.0025)" hin + 6,551,237 Ibfin” + (0,005(-0.0025)" fi =112indb , (0.003)? 12 0.005)" Dy = 392,656 Ib/in” leans) + (0.005)(0.0025)° fin® 2{ (0.005)" 21.9 02061 + 4,555,238 Ibvin” | “+ (0,005)(-0.0025)" fin’ = 0.206 in-lb 1 2| ( 2 Dy = 1,308,853 Ib/in’ | ~~ + (0.005)(0.0025) 0. (0.005)? + 6,557,237 Ibfin? + (0.005)(-0,0025)* fin’ = 0.328 inb , 2 3 005) 12 0: 12 Dyg= 0+ 4,708,483 Ibfin” + (0.005(-0.0025)" fin? = 0.196 in-b , (0.008)? 12 (0.005)° Dag = 0+ 4,705,483 Ibfin? yt 005)(-0.0025)" fin’ = 0.196 in-tb , (0.005) ? (coos) + (0.005(0.0025)" Jin® 3 + (0.005)(~0.0025)" fin? = 0.257 in-b . | (0.005 + 5,163,582 Ibfin® |“ OE) Now use equation (84) and the Fortran programs in appendix A to obtain: NJh 843x10' DS 1 naaaaaeninae = 5,839,000 Ibvin” . (£23) a 14,440,750 Ib°-in 0.01 in 39D. Summary The equations for the in-plane engineering constants of a symmetric laminate are: Ajp[ AzA | 2 HA 2A 65 A264 124 66~Ar64 16 A . 2 a2 Ard 66-Arg Ar 66-Az6A 66 2 An AeA 124 66-A 6A 26 Air] Ars i6-A 2A 66 2.2 AitAGe~A AsFor unsymmetric laminates, the engineering constants are given by: AuAndw BuBrBis AA Ar Bi2Bx Bre AeA 26 Acs Big B25 Boo By By2Bys Dy DpDi6 B12 By Ba5 Dy2Dx Dog B6Br5 Bos Dig P26 Doo An Aw Bua Bn Arm Ass Bis Bas An Apr BuBnBis Ai2A2Ar5 Bir Br Brg AwAr Acs Bis B25 Bos By By2Bys Du Di Dig By2B22Br5 Di2Dn Dr By Br5 Bos Dis Pr Dos Bu By Bis Bye Dy Pr Dy Dy Dig Dag le 442 AyArArw Bu Bi2 Bis Ar2AnA26 Bi. Bx Bas AAA 66 Bi¢ B26 Bos =loVI. ENVIRONMENTAL EFFECTS A. Importance Like all engineering materials, laminates will contain residual stresses from processing, or will experience strains due to thermal effects or moisture absorption. However, because of the anisotropic nature of composite laminates, these effects become much more important. It is well known that most laminates are processed or “cured” at elevated temperatures. It is at this elevated temperature that the molecular structure of the material is set. Upon cooling, the laminate will experience many internal stresses as each ply contracts a different amount in different directions. For unsymmetric laminates, this can be seen by warpage of a flat plate as it is removed from its molding platens. In some cases, this warpage is desirable to achieve a natural “twist” in the material (such as for helicopter rotors). In general, the more anisotropic the laminate, the more important the residual thermal and moisture absorption stresses become. The thermal effects will be quite important in the next section on ply stresses. B. Coefficients of Thermal Expansion Just like any other material, fiber/resin systems will experience a change in strain with a change in temperature. These strains are defined by the coefficients of thermal expansion. These values are material constants in each principal material direction. Thus, two constants will describe the thermal expansion coefficients for any lamina. These are defined as: x, — coefficient of thermal expansion in the fiber direction and 0 ~ coefficient of thermal expansion in the direction perpendicular to the fibers.These values have dimensions of inch/inch/°F or inch/inch/°C, depending on the temperature scale that is being used. Just as with the mechanical strains, the thermal strains must be transformed into the laminate coordinate system: cos” 6 sin’@ —-2sin 8cos @ a, a, a, @ cos” 8 2sin acos@ |f a} - (90) oy sin @cos@ —_-sin 8cos (cos? @-sin” 4 : Note that a coefficient of thermal expansion for shear is formulated if the lamina being examined does not have its material axes as principal material axes (ie., @ # 0° or 90°). The amount of thermal strain induced in each lamina is given by: e=a,aT, (91) es - eles tyAT . where the superscript T denotes “thermal,” and AT denotes the change in temperature from cure to operating temperature. These thermal strains are now treated just like the mechanical strains considered earlier. Therefore, from equation (43), it can be seen that when these thermal strains are combined for each layer of the laminate, thermal stress resultants are present: r Ms] on [OnBnBre r A N, B12 On Bas (92) My 216 O25 Qos A From equation (44), it can be seen that thermal moment resultants also develop: MY 6 [Onder] [4] , M, =z G29 026) | &, £ (i al . (93) al] "Biers Be|These thermal stress and moment resultants can be added to the mechanical stress and moment resultants to arrive at the total stress and moment resultants: ro] | yr ta) [ yr Ne | Ns} [n, Me) 1M) [m, qot|_| ra|_| 7 wh lelntle| ny] and [ac |=| ad |e] a} (04) ra! || Ny ra} | rl | a, Ne] [Ny me) [Me] [Me where the superscript Tot denotes “total”. Typical values of the thermal expansion coefficients for carbon/epoxy is xl = -0.072x10-6 invin/°P and o2 = 32.4x10° in/in/*F. Note that in the fiber direction, the material actually contracts upon heating (or expands upon cooling), Also, note the large differences in the two coefficients. For this reason, composite materials have been used in many applications where a certain thermal expansion is desired, since plies can be combined to give a wide range of values (including near zero which is very convenient for optical benches). C. Moisture Effects Swelling of a composite material due to moisture absorption is handled in the exact same manner as expansion due to temperature differences. The moisture swelling coefficients in each principal material direction must be known. They are designated as 1, along the fiber direction, and B2, perpendicular to the fiber direction. The strain due to moisture absorption is given by: el =BAm, at Bam, (95) M 2= Byam, where the f;’s are the transformed moisture expansion coefficients, Am is the moisture concentration in weight moisture/weight material, and the superscript M denotes “moisture.” Stress and moment resultants can be determined due to the effects of moisture absorption. Typical values of moisture expansion coefficients for carbon/epoxy are, 1 = 0.01 inv/in/g/g and 0.35 in/in/g/g. The moisture concentration is usually a very low number under normal operating ‘umstances ~0.0005 g/g, However, in humid environments this number may be much higher. 45VII. STRESSES AND STRAINS WITHIN LAMINAE OF A SYMMETRIC LAMINATE A. Strains Within the Laminae So far, in this paper, methods have been shown to describe the stress-strain behavior of a laminate as a whole. In many instances, however, the stresses and/or strains in each ply may need to be known. To do this, the constitutive equation (52) is used to find the total midplane strains and curvatures. From these values, the total strain in each ply can be calculated from equation (30). These strains can then be transformed into strains in the principal material directions by the first equation in (18). As in finding the engineering constants of a laminate, the most involved step in performing this analysis is the manipulation of matrices. The A’, B', and D' matrices must be found from the A, B, and D matrices. Example 3: Using the symmetric laminate in example 1, {0,+45]s, find the strains in each ply in each ply’s principal material direction given a tensile load of 1,000 Ib on a 2-in wide specimen in the x-direction. ‘The x-direction is parallel to the 0° fibers. Assume no environmental effects. The specimen is 2-in wide, therefore, the stress resultant is 1,0001b _ 4 Ib 2in in” (E3.1) From example 1, the Ay are known, Since the laminate is symmetric, the By are zero. The Dy are given by: 3 Dyy = 20,130,785 Ibvin? (coos), (0.005)(0.0075)* in 3 +6,557,237 Ib/in® (o0es) (0.005)(0.0025)* Jin (loos? 12 + 6,557,237 Ibvin (0.005)(-0.0025)" 3 +20,130,785 Ibvin? tonto in? = 12.3 ind ,3 D2 = 392,656 Ibvin” f os) + asta in? 3 4+4,555,238 Ibvin (coos) +(0. asf in? | [oss ons} + 4,555,238 Ibvin +(0.005}(-0. os in 3 sons 9.005) +(0.005)(-0.0075) ihe 0.609 inlb , Dyy = 1,308,853 Ibfin” € us) + omsoan)} in? + 6,557,237 Ibvin [lcs os}! +(0.005)(0.0025) he + 6,557,237 Ibvin” (a +(0.005)(-0.0025)"}in* (0.005)? 12 +(0.005)(-0.0075)" Jin? = 1.31 indb , S_-B + 1,308,853 wl 3 Dyg=0 Ibvin® [ow + oasfun| in? (0.005)° at (0.005}(0.0025)" J in’ + 4,705,483 w 3 +4,705,483 Ib/in? | (oms) + (0.005}(-0. on in? 473 +0 Ibvin? (ess + ows aun in?=0.392intb , Dag= 0 Ib/in [lost os)" soasuan} 3 $4,705,483 Ibfin? [ost + (sf in? (0.005)? 12 +4,705,483, ws +(0. ast ews in? +0 Ibvin [oss ons) + (0.005}(-0.0075) rhe =0,392inIb , Dgg= 1,001,000 Ibvin (ost sos) +(0. ast + 5,163,582 Ib/in [oss (0. asta 3 45,163,582 Ibn? [ost + os-ms} in? 3 + 1,001,000 Ibvin? [ost + oso in = 1.014 intb Using equations (51) and (53), the A' and D' matrices can be found. The B' and C’ matrices will be zero since the B matrix is zero. 266,880 49,479 47,055 441-14 -23 | [a]=[4]"=|| 49.479 78,661 47.055 -14 23.8 -17.1 pes . (E3.2) 47,055. 47,055 61,646 23 “in 3.0 4812.3 0,609 0.392 — 0.609 1.31 0.392 }intb} =| -0.033 0.876 0.392 0392 1.014 0.0195 -0.326 Therefore, the constitutive equation is: ol) foun & | [Andina 0 0 0 of fey Ny S| [Andndg 090 1G % | | AisAretes 99% JLo K,| | 000 DyPyPef}° K bone 0 y 00 0 DyDy»DPr |} 0 Ky 0.0836 -0.033 -0.0195 0.326 000 DDyDeg 1.12 tb)" (B3.3) (E3.4) These are the midplane strains and curvatures. Using equation (30), it can be seen that since the midplane curvatures are zero, then the strains in each ply will be equal to the midplane strains. In other words, the strains are constant through the thickness in any given direction since there is no bending and therefore z does not enter into the calculation. To transform these strains into the principal material directions for the +45° ply: 49e cos” 45° sin? 45° 2sin 45° cos 45° | 0.00221 &| =] sin’4se cos 45° -2sin 45° cos. 45° |} -0.0007 te — sin 45° cos 45° sin 45° cos 45° (cos?4se -sin’4s°) au 2 +45, . (E35) ef =o.00018; «*°=0,00133; 73° =-0.00291 Note: Recall that the engineering shear strain must be put into tensorial shear strain, y = 1/2 Yoy before transformation. B. Stresses Within the Laminae ‘The stresses in a lamina can be determined once the strains are known by applying equation (23) to each ply. If thermal or moisture strains are present, these strains must be subtracted from the total strain since they are not caused by an external force. Example 3a: Find the stresses in each ply of the laminate in example 3. | |20,130,785 392,656 0 0.00221 || %} [44,214 7,| =| 392.656 1,308,853 0 Ib/in?} -0.0007 o,| =| 48 Ibn? | (B3a.1) + 0 © 1,001,000 ~0.00115].}- | | -1151 9 Io aoe lo %, 6,557,237 4,555,238 4,705,483 0.00221 | | 5,892 4g, | =|4.555,238 6,557,237 4,705,483 | tbvin”| 0.0007 || o,} =| 66 |ib/in? . (E3a.2) + 4,105,483 4,705,483 5,163,582 -0.00115 | 44) « 1,167 das OL Oy J 145 Transforming the +45° ply into its principal material directions: 3 cos? 45° sin? 45° 2sin 45° cos 45° |[ sgon. o,| =| — sin?4se cos” 45° ~2sin 45° cos 45° |} 66 | Ibfin? Tia] 4g [sin 45° cos 45° sin 45° cos 45° cos? 45° — sin? as’) ae (B3a.3) oF =4,146 Ibfin?; of" =1,812Ib/in?; —rfg° =-2,913 tbfin? 50Example 4: The importance of thermal effects can be found by working problem 3, taking into account that the laminate was cured at 176.7 °C and tested at 21.1 °C. This implies a AT of -155.6 °C. The coefficients of thermal expansion for AS4/3501-6 are given below: ~0.04x10° 5 a, = 18x10 (E4.1) in°C in Using equation (90), for the 0° plies the coefficients are: -6_in -6 = 0.0410 a, = 18x10 4,2 o nee § YEN nee and for the +45° plies: 4s -¢_in_ as sin as -6_in = 8.98x10 ; =8.98x10 =-9,02x1 —=.- 4.3) ox nee * % x ine * An nec (SA) From equation (91), the thermal strains in the 0° and the 45° plies can be found: e°7=0,0000062" ; e°7=-0.0028 ; 727 (B4.4) 2 in?“ oy 0.0014 ; 74°" =0.0028 (4.5) r Ny 33.57 wi |=] -60.42 - (E46) at] [1283 7 ‘The thermal moment resultants will be zero for the same reason the By are zero, the laminate is symmetric. Thus: 51AyAndy 000 {Ns ot AiAndrg 000. |/% N Toy | | AygAneAgs 9.00 & 000 DyDyDPel] 0 Ky Ky 000 DyDyDyy ; K 7 ae oe Dy Dag Deg =4anaoe = (ss 2). 1.4x10°% (ae! 2) in in ry ea-l aao*ite 466.4322) 423.810 "| 60,42 © | 17. 1x10 , 1b in b in in Pa -17.1076 | 60,42 | 431.05.10 in Ib (sscua® a= K,=0 3 Kj=0 3 Ky=0 (E4.7) Transforming these into principal material directions for the 45° plies gives: & cos? 45° sin? 45° 2sin 45° cos 45° || 0.00211 | =| sin?4se cos’ 45° —2sin 45° cos 45° || -0.00231 Na ~sin 45° cos 45° sin 45° cos 45° {cos"45° — sin”45° oe 2 a - (48) + 0.0001 ; off =~0.00042 The lamina stresses are: 52| | 20,130,785 392,656 0 0.0021 1-0.000006 @,| =| 392,656 1,308,853 0 —_|Ibvin”| -0.00231-(-0.0028) + 0 0 1,001,000 0.00036-0 Io E49) Ox] | 42,543 49) o,| =| 1.467 | b/in® 360 ty lo For the 0° ply, the x-direction, stress is altered slightly, but the y-direction and shear stresses are quite different due to the thermal effects. o, 6,557,237 4,555,238 4,705,483 0,00211-(-0.0014) 7 =| 4,555,238 6,557,237 4,705,483 Ibvin? ~0.00231-(-0.0014) + 4,705,483 4,705,483 5,163,582 0.00036-0.0028 |. teas o £4.10) ox 7,389 : : =| -1,460 | tbvin® + 366 ).45, Transforming the +45° ply stresses into principal material direction stresses: o, cos” 45° sin? 45° 2sin 45° cos 45° |[ 7,389 o| =| sin?4se cos?45° 2 sin 45° cos 45° || -1,460 | thi Tra] jg | sin 45° cos 45° sin 45° cos 45° cos? 45° sin®4s?) 365 . (E411) of = 2,600 Ib/in? ; of" =3,3301Ibin® ; 74° =—4,425 bin” For the + 45° plies, the stresses vary substantially in all directions due to thermal effects. Example 5: Perform the same exercise as above taking into account moisture uptake. The amount of moisture absorbed by the composite is 0.007 g/g, and the moisture swelling coefficients are By = 0.01 in/in/g/g and By = 0.35 in/in/g/g 53‘The moisture expansion coefficients need to be transformed for the +45° plies; cos? 45° sin? 45° -2 sin 45° cos 45° B, 0.01 B,| = sin? 45° cos? 45° 2'sin 45° cos 45° |] 0.35 2 By] 4, | sin 45° cos 45° ~ sin 45° cos 45° cos? 45°— sin’ 45° +45 45 an 45, ae (5.1) Br =0.18 infinlele ; "0.18 inlinlgle ; B%° =-0.17 infinfole The strains due to moisture uptake are given by equation (95): 2" = 0.00007 ; "= 0.00245 = * in" *Y in (E5.2) +45, Hs "0.001262 ; & 7 : mo) SM in as in = 0.00126 — =~0.00238 — in * in ‘The stress and moment resultants due to moisture swelling can now be calculated using equations (92) and (93): =] [siz]. in w= es me (E5.2) Ny ‘The moisture moment resultants will be zero since the laminate is symmetric. The new total stress resultant is the sum of the mechanical, thermal, and moisture stress resultants: Tot Ns"! [500 nt] =}°0 (5.4) wre] Lo sy Using the constitutive equation, the midplane strains can be calculated as they were in equations (3.4) and (4.7). 54Ted araeae ema iloe| |v 0 Tot & AiaAnaAag 000 a ° ot N Ares 000 y K, 000 DD, Die|| 0 K, ee alo || 000 DpPPss|} 5 oy we: oe Dip Pas Doo, 0.00227 Ib in? 518.12} -14x10%* [0+ in 1b 7 lb oe I 22 tb oe -1ax10* = [518.1 >} 423.810 = (0 >]-17..0 = [8.5] =-0.00087 Ib in Ib in Ib in aso ™ (sig) rrct0 Ib in Transforming these strains into principal material direction strains for the +45° plies; > > 0.00227 cos” 45° sin? 45° 2 sin 45° cos 45° sin? 45° cos” 45° ~2 sin 45° cos 45° || -0,00087 Ne —sin 45° cos 45° sin 45° cos 45° (cos?ase 5 sin? 45°) 0.00093 2 2 (E5.6) ef = 0.00024 ; ef =0.00117 ; f° =-0.00314 The stresses in the 0° plies are given by: 55%,| | 20,130,785 392,656 0 0.00227-0.000006-0.00007 o,| =| 392,656 1,308,853 0 Ib/in” | -0.00087-(-0.0028)-0.00245 a 0 0 1,001,000 -0.00093-0-0 “| lo 43,963, _ 181 |tbvin? -931 The stresses in the +45° plies are given by: o 6,557,237 4,555,238 4,705,483 0,00227-{-0.0014)-0.00126 4g, | =| 4,555,238 6,557,237 4,705,483 | Ibvin”| -0.00087-(-0.0014)-0.00126 4,705,483 4,705,483 5,163,582 ~0,00093-0,0028-{-0.00238) S445, +45, 6,125 . (E5.8) = a Ibfin® Transforming these stresses into principal material directions for the +45° plies: a, cos” 45° sin? 45° 2 sin 45° cos 45° || 6125 161 | yin? sin? 45° cos’ 45° =2 sin 45° cos 45° |] | Ibvin’ oO " Tra] gy |—Sin 5? cos 45° sin 45° cos 45° (cosas — sins) 934 (B5.9) 3,143 Ibvin? 53.916 Ibjin? ; ofS=2,048 vin? ; The swelling strain of moisture uptake tends to diminish the contraction due to processing thermal differences. Example 6: Find the stresses and strains in each lamina of the laminate used in the previous examples if the laminate is subjected to a bending moment in the 0° fiber direction of 5 in-Ib/in.The constitutive equation is: 2) [hate 000 s ApAy Ay, 000 ; | Alig 990 Ilo K,| | 000 PyuDiPie|| Me K, 000 Dy DyDrg i = K, 0418 = = in-lb Wy Didh= 009558 ss) =-0.165 = (B6.1) iF ( a ) = 0.0995 = 0 9 y= 5 Ny Ky =D M,=-0.0195 Using equation (30), it can be seen that the strains in each ply will vary across the thickness of the ply. Thus, the distance the ply is away from the geometric midplane must be taken into account, as well as the direction of the fibers as in the previous examples. For clarity, assign the four plies of the laminate numbers as shown in the following: Z .005in. | Ply#1 o° .00Sin. | Ply#2 4 45° oosin, | Ply#3 +45° .005in, | Ply #4 0° Since this laminate is symmetric, the geometric midplane is also the neutral plane of the plate. Thus, the strains will be zero at the bottom of ply No. 2 and at the top of ply No. 3. The following are calculations for other planes within the plate.