CHAPTER 5 One-Dimensional Elements ‘The objectives of this chapter are to introduce the concepts of one-dimensional ele- ments and shape functions and their properties in more detail. The idea of local and natural coordinate systems is also presented here, In addition, one-dimensional ele- ments used by ANSYS are discussed, These are the main topics discussed in Chapter 5: Linear Elements Quadratic Elements Cubic Elements Global, Local, and Natural Coordinates Isoparametric Elements Numerical Integration: Gauss-Legendre Quadrature Examples of One-Dimensional Elements in ANSYS 5.1 LINEAR ELEMENTS ‘The heat transfer example in this section is employed to introduce the basic ideas of one-dimensional elements and shape functions. Fins are commonly used in a variety of, enginecring applications to enhance cooling. Common examples include a motorcycle engine head, a lawn mower engine head, extended surfaces (heat sinks) used in electronic equipment, and finned-tube heat exchangers. A straight fin of a uniform cross section is shown in Figure 5.1, along with a typical temperature distribution along the fin. Asa first approximation, let us divide the fin into three elements and four nodes. The actual temperature distribution may be approximated by a combination of linear functions, as shown in Figure 5.1. To better approximate the actual temperature gradient near the base of the fin in our finite element model, we have placed the nodes closer to each other in that region. It should be clear that we can improve the accuracy of our ap- proximation by increasing the number of elements as well. However, for now, let us be content with the three-element model and focus our attention on a typical element, asSection5.1 LinearElements 253 Actual temperature profile FIGURE 5.1. Temperature distribution for fin of uniform cross section. FIGURE 5.2 Linear approximation of temperature distribution for an element. shown in Figure 5.2. The temperature distribution along the element may be interpo- lated (or approximated) using a linear function, as depicted in Figure 5.2. The forthcoming derivation of the shape functions are similar to the one we showed in Chapter 4, Section 4.1. As a review and for the sake of continuity and convenience, the steps to derive the shape functions are presented here again. The linear temperature distribution for a typical element may be expressed as TO = +X (1) The element's end conditions are given by the nodal temperatures 7; and Tj, according to the conditions T=T a X=X (52) T =T; at X=;254 Chapter 5 One-Dimensional Elements Substitution of nodal values into Eq. (5.1) results in two equations and two unknowns: T =e + @X; (3) Tj =6 +X; Solving for the unknowns ¢; and ¢2, we get a= Bi AT (4) Xj- Xi oe (3) oS Xk The element's temperature distribution in terms of its nodal values is 1X,-TX, . G-T Grouping the 7; terms together and the 7; terms together, we obtain X,- X X - X; @ = (— aati T (2 = =n + (2 = xn (5.7) We now define the shape functions, §, and Sj, according to the equations Si 6.8) 5 69) where ¢ is the length of the element. Thus, the temperature distribution of an element in terms of the shape functions can be written as T® = 57 + ST; (5.10) Equation (5.10) can also be expressed in matrix form as To =[5, sth (5.1) Tj As you recall, for the structural example in Chapter 4, the deflection u for a typical column element is represented by @ = (5, sq4" 5.12) nls Sith (6.12) where u and u; represent the deflections of nodes and j of an arbitrary element (¢). It should be clear by now that we can represent the spatial variation of any unknown vari- able over a given element by using shape functions and the corresponding nodal values.Silex; = Sileex; = Section 5.1 Linear Elements 255 Thus, in general, we can write v vO =[s, 5] {v4 (6.13) A where W, and ; represent the nodal values of the unknown variable, such as temper- ature, deflection, or velocity. Properties of Shape Functions ‘The shape functions possess unique properties that are important for us to understand because they simplify the evaluation of certain integrals when we are deriving the con- ductance or stiffness matrices. One of the inherent properties of a shape function is that it has a value of unity at its corresponding node and has a value of zero at the adjacent node. Let us demonstrate this property by evaluating the shape functions at X = X; and X = Xj. Evaluating 5; at X = X,and X = Xj,.we get Xj-X x) - X, xXj-X = = 1andS\y.x,= CO lrex; ¢ i Evaluating S;at X = X;and X = X;, we obtain X-X, Xi- Xi aa: a = Oand Sly-x, = This property is also illustrated in Figure 5.3. Another important property associated with shape functions is that the shape fune- tions add up toa value of unity. That is, Xj-X X-% XX, Xj X, It can also be readily shown that for linear shape functions, the sum of the derivatives with respect to X is zero. That is, aA (MAX), a (X=% 1 aX\X;— X,] ~ aX\ Xj — X; X;-X; 5 5 i : A LA % 5 x, 5 FIGURE 5.3 Linear shape functions. S.+S)= (6.16) (5.17)256 Chapter5 One-Dimensional Elements EXAMPLE 5.1 We have used linear one-dimensional elements to approximate the temperature dis- tribution along a fin. The nodal temperatures and their corresponding positions are shown in Figure 5.4. What is the temperature of the fin at (a) X = 4cm and (b) X =8cm? In Chapter 6, we will discuss in detail the analysis of one-dimensional fin problems, including the computation of nodal temperatures. However, for now, using the given nodal temperatures, we can proceed to answer both parts of the question: a, The temperature of the fin at X = 4 cm is represented by element (2); X - X; TO = SOT, + SOT, Xr, 7; 4 4 é T == (Al) += 4) = 363°C b. The temperature of the fin at X = 8 emis represented by element (3); yn x X-X; T! = S07, + SPT = 4 en 10 - - 7 = P=8 Gy) + 85 (20) = 256°C 5 5 For this example, note the difference between S$? and S©). Lox 18°C ee} ‘ Th 50 I T\ _ Jail, 1m 2 @ 3 6) 4 Ta ~ )34 T, 20 |+2em->[+—3 m—+|+ sem ——+| FIGURE 5.4 The nodal temperatures and their corresponding positions along the fin in Example 5.1. 5.2 QUADRATIC ELEMENTS We can increase the accuracy of our finite element findings either by increasing the number of linear elements used in the analysis or by using higher order interpolation functions. For example, we can employ a quadratic function to represent the spatial variation of an unknown variable. Using a quadratic function instead of a linear functionSection5.2 Quadratic Elements 257 requires that we use three nodes to define an element. We need three nodes to define an element because in order to fit a quadratic function, we need three points. The third point can be created by placing a node, such as node k, in the middle of an element, as, shown in Figure 5.5, Referring to the previous example of a fin, using quadratic approx- imation, the temperature distribution for a typical element can be represented by TO =, +X +X? (5.18) and the nodal values are T=T, at X=X, (5.19) T=T at X=X% T=T a X=; Three equations and three unknowns are created upon substitution of the nodal values into Eq. (5.18): cr + OX; + 5X? (5.20) Ty = cy + 0X4 + 0X} Tj = cy + Xj + eX} Solving for c;, cz, and ¢; and rearranging terms leads to the element's temperature dis- tribution in terms of the nodal values and the shape function: TO = ST; + ST, + ST (821) In matrix form, the above expression is T TO= (5, S S147 (5.22) qh FIGURE 5.5 Quadratic approximation of the temperature distribution for an element.258 Chapter 5 One-Dimensional Elements where the shape functions are pa (X — Xj)(X — Xi) (6.23) Sx = XX = %) Se = (HX ~ XX ~ XK) In general, for a given element the variation of any parameter W in terms of its nodal values may be written as Ws WO=[S S, SIV; (5.24) Ve It is important to note here that the quadratic shape functions possess properties similar to those of the linear shape functions; that is, (1) a shape function has a value of unity at its corresponding node and a value of zero at the other adjacent node, and (2) if we sum up the shape functions, we will again come up with a value of unity. The main difference between linear shape functions and quadratic shape functions is in their derivatives. The derivatives of the quadratic shape functions with respect to X are not constant. 5.3. CUBIC ELEMENTS The quadratic interpolation functions offer good results in finite element formula- tions. However, if additional accuracy is needed, we can resort to even higher order interpolation functions, such as third-order polynomials. Thus, we can use cubic fune- tions to represent the spatial variation of a given variable. Using a cubic function in- stead of a quadratic function requires that we use four nodes to define an element. We need four nodes to define an element because in order to fit a third-order poly- nomial, we need four points. The element is divided into three equal lengths. The placement of the four nodes is depicted in Figure 5.6. Referring to the previous ex- ample of a fin, using cubic approximation, the temperature distribution for a typical element can be represented by TO =o +X +X? + 0X9 (5.25) and the nodal values are at X=X, (5.26) at X =X, at X=X, at X= Xj Four equations and four unknowns are created upon substitution of the nodal values into Eq. (5.25). Solving for ¢}, ¢, ¢;, and ¢, and rearranging terms leads to the element’sSection 5.3 Cubic Elements 259 x Xr Xe x kk 3—}-— 3-3 —| ‘ FIGURE 5.6 Cubicapproximation of the 4 iw? @ 3 @) 4 Ts 34 +2 em-—}-—3 m=} 5 em >|264 Chapter 5 One-Dimensional Elements a. Using local coordinates, we find that the temperature of the fin at X = 8 cm is represented by element (3) according to the equation T® = SOT, + SPT, = ( - =) +27 Note that element (3) has a length of 5 cm, and the location of a point 8 em from the base is represented by the local coordinate x = 3: T= (: = son +2(20) = 256°C . Using natural coordinates, we find that the temperature of the fin at X = 7.5m is represented by element (3) according to the equation 7 = SOF, + SPT, =F -OB+ FU +9 2 Because the point with the global coordinate X = 7.5 cm is located in the middle of element (3), the natural coordinate of this point is given by € TO = za = 0)(34) + 0 + 0)(20) = 27°C One-Dimensional Natural Quadratic and Cubic Shape Functions The natural one-dimensional quadratic and cubic shape functions can be obtained ina way similar to the method discussed in the previous section. The quadratic natural shape functions are 1 zal - 8) (537) 1 Sze +8) (538) S= +908) (539) The natural one-dimensional cubic shape functions are 1 5) = 7g (1 - GE + NGE- 1) (540) 1 76 + &)(3§ + 1)(3€ - 1) (5.41) 5 = + OE -NEE-H) (5.42) Suis 2 (1 + &)(1 — &)(3 + 1) (5.43)Section 5.6 Numerical Integration: Gauss-Legendre Quadrature 265 For the sake of convenience, the results of Sections 5.1 to 5.4 are summarized in Table 5.1, Make sure to distinguish the differences among presentations of the shape fune- tions using global, local, and natural coordinates. EXAMPLE 5.4 xy Evaluate the integral [ ‘S}dX using (a) global coordinates and (b) local coordinates. Xi a. Using global coordinates, we obtain a] Xi x — x, xj _ f S}dX = ( = Hs Ix, IX; x b, Using local coordinates, we obtain Xj fx\2 SEG SidX = [ =)ax=-) => f i hb \e) = 3@|,~ 3 This simple example demonstrates that local coordinates offer a simple way to evaluate integrals containing products of shape functions. 5.6 NUMERICAL INTEGRATION: GAUSS-LEGENDRE QUADRATURE As we discussed earlier, natural coordinates are basically local coordinates in a dimen- sionless form. Moreover, most finite element programs perform element numerical in- tegration by Gaussian quadratures, and as the limit of integration, they use an interval from —1 to 1. This approach is taken because when the function being integrated is known, the Gauss-Legendre formulae offer a more efficient way of evaluating an inte- gral as compared to other numerical integration methods such as the trapezoidal method. Whereas the trapezoidal method or Simpson's method can be used to evaluate integrals dealing with discrete data (see Problem 24), the Gauss-Legendre method is based on the evaluation of a known function at nonuniformally spaced points to compute the in- tegral. The two-point Gauss-Legendre formula is developed next in this section. The basic goal behind the Gauss-Legendre formulae is to represent an integral in terms of the sum of the product of certain weighting coefficients and the value of the function at some selected points. So, we begin with . a I [100m = Desi) (5.44) Next, we must ask (1) How do we determine the value of the weighting coefficients, represented by the w;s? (2) Where do we evaluate the function, or in other words, how do we select these points (x;)? We begin by changing the limits of integration from a — b to 1 to 1 with the introduction of the variable A such that X= Got Gd992 TABLES.1 One-dimensional shape functions Tnterms of global coordinate X Tn terms of oa cordiatex Tnterms of natural coordinate § Inertia incon i EN) oeree treet Lins set Said saki-o 5-5 set saho+o orate Sadur- ayer 9 5=(§-1)((3)-1) suo s-4ar-ner- x9 5-(30G)- ) shure 8 = Far ~ xr 3) a=d3)(i- ()) S=G-p0+9 cate 5 Baar myx = mca = 49 5-2er- mere x9 Fa =a ~ aur - agar 2 See Bex = mar-aper- x9 Si Fal OGE + G51) s-da+ocerneen 9 5-20 +oe-nerD san Za+ aa -oce+)Section 5.6 Numerical Integration: Gauss-Legendre Quadrature 267 Matching the limits, we get a= 0 + ¢(-1) b=) + (1) and solving for ¢ and ¢}, we have _ (b +a) q =o" and 4 aes Therefore, _(b+a) (b-a) eee Sse (5.43) and b- dx= Pada (5.46) Thus, using Eqs. (5.45) and (5.46), we find that any integral in the form of Eq. (5.44) can be expressed in terms of an integral with its limits at —1 and 1: F0) as = Sago) (647) The two-point Gauss-Legendre formulation requires the determination of two weighting factors 2, and tw; and two sampling points dy and A; to evaluate the function at these points. Because there are four unknowns, four equations are created using Legendre polynomials (1, 4, X2, 2°) as follows: \ waft) + wf) = fran =2 1 wf (dr) + af (ho) = pra =0 My : cof) + vfs) = [ran = 2 Mt 1 wif (My) + wef (2) = [ra 0 Mt The above equations lead to the equations aw,(1) + 203(1) = 2 20,(dy) + 20,9) = 0Chapter 5 One-Dimensional Elements TABLE 5.2 Weighting factors and sampling points for Gauss-Legendre formulae Points (n) Weighting factors (1) Sampling points (4)) 2 0577350269 0577350269 3 Ay = -0.774596669 a =0 0.774596669 4 0.3478548 0861136312 (0.6521452 (0.6521452 3 ty = 0.2369269 -0.906179846 0.4786287 8469310 0.568889 0.4786287 0.2369269 0538469310 0906179846 els tw (M + 1,04)? = wi)? + wy) = 0 Solving for 2,103.4. and Ay, We have w, = tw, = 1, = -0.577350269, and dy = 0.577350269. The weighting factors and the 2, 3, 4, and 5 sampling points for Gauss-Legendre formulae are given in Table 5.2. Note that as the number of sampling points increases, so does the accuracy of the calculations. As you will see in Chapter 7, swe can readily extend the Gauss-Legendre quadrature formulation to two- or three- dimensional problems. EXAMPLE 5.5 Evaluate the integral J = [ (2? + Sx + 3) dx using the Gauss-Legendre two-point sampling formula. This integral is simple and can be evaluated analytically, leading to the solution I = 161.333333333, The purpose of this example is to demonstrate the Gauss-Legendre procedure. We begin by changing the variable x to \ by using Eq. (5.45). So, we obtain pate =a), _ (642) (6-2) 7 7 7 + K=442d and = O59 9 = CD a - 200Section 5.6 Numerical Integration: Gauss-Legendre Quadrature 269 Thus, the integral / can be expressed in terms of d: fe) f0) 6 1 I [ (x + Sx + 3)dx = [ (2)[(4 + 2a)? + 5(4 + 2d) + 3)]dn 2 a Using the Gauss-Legendre two-point formula and Table 5.2, we compute the value of the integral / from = wf (hr) + wef(d2) From Table 5.2, we find that 0, = 20 = 1, and evaluating f(A) at Ay = —0.577350269 and hz = 0.577350269, we obtain Fu) = (2)[f4 + 2(-0.577350269)}? + 5(4 + 2(-0.577350269) + 3)] = 50.6444526769 F(X) = (2)[[4 + 2(0.577350269)} + 5(4 + 2(0.577350269) + 3)] = 110.688880653 I = (1)(50.6444526769) + (1)110.688880653 = 161.33333333 EXAMPLE 5.6 xj j Evaluate the integral [ S}dX in Example 5.4 using the Gauss-Legendre two-point Xi formula. Recall from Eq. (5.35) that S; = (1 + &) and by differentiating the relationship between the local coordinate x and the natural coordinate & (ie., & = ze -1= dg= 2 dx) we find dx = $f, Also note that for this problem, § = d. So, Ay 2 /e\2 cpt 2 r= f 'sjax = ax= [(2)-$/ [foro] Ix, | ty ln \E 24,12 Using the Gauss-Legendre two-point formula and Table 5.2, we compute the value of the integral / from —~ SS T= wyf(dy) + wef (2) From Table 5.2, we find that wo, = 2 = 1, and evaluating f(A) at dy = —0.577350269 and kz = 0.577350269, we obtain 2 2 F(&) = [3a + «| = ‘ [3c - osrs20) = 0,022329099389¢ ef P ef P f(&) =5] 50 + &) | =>] 5C + 0.577350269) | = 0.31100423389¢ 2|2 2(2 7 = (1)(0.022329099389¢) + (1)(0.31100423389¢) = 0.333333333¢ Note that the above result is identical to the results of Example 5.4.270 Chapter S One-Dimensional Elements 5.7 EXAMPLES OF ONE-DIMENSIONAL ELEMENTS IN ANSYS ANSYS offers uniaxial link elements that may be used to represent one-dimensional problems. These link elements include LINK31, LINK32, and LINK34, The LINK32 el- cement is a uniaxial heat conduction element. It allows for the transfer of heat between its two nodes via conduction mode. The nodal degree of freedom associated with this element is temperature. The element is defined by its two nodes, cross-sectional area, and material properties such as thermal conductivity. The LINK34 element is a uniax- ial convection link that allows for heat transfer between its nodes by convection. This element is defined by its two nodes, a convective surface area, and a convective heat transfer (film) coefficient. The LINK31 element can be used to model radiation heat transfer between two points in space. The element is defined by its two nodes, a radia- tion surface area, a geometric shape factor, emissivity, and the Stefan—Boltzman constant. In Chapter 6, we will use LINK32 and LINK34 to solve a one-dimensional heat- conduction problem. SUMMARY At this point you should 1. have a good understanding of the linear one-dimensional elements and shape functions, their properties, and their limitations. 2. have a good understanding of the quadratic and cubic one-dimensional elements and shape functions, their properties, and their advantages over linear elements. 3. know why it is important to use local and natural coordinate systems. 4, know what is meant by isoparametric element and formulation. 5. have a good understanding of Gauss-Legendre quadrature. 6. know examples of one-dimensional elements in ANSYS. REFERENCES ANSYS User's Manual: Elements, Vol. Ill, Swanson Analysis Systems, Inc. Chandrupatla, T., and Belegundu, A., Introduction to Finite Elements in Engineering, Prentice Hall, 1991. Incropera, F. P..and DeWitt, D. P., Fundamentals of Heat and Mass Transfer, 2nd ed., New York, John Wiley and Sons, 1985, Segrlind, L., Applied Finite Element Analysis, 2nd ed., New York John Wiley and Sons, 1984, PROBLEMS 1. We have used linear one-dimensional elements to approximate the temperature distribution along a fin, The nodal temperatures and their corresponding positions are shown in the ac- companying figure. (a) What isthe temperature of the fin at X = 7 em? (b) Evaluate the heat Joss from the fin using the relationship ar = kat where k = 180 Wim+K and A = 10 mn,Chapter 5 Problems 271 1@m? @ 3 @) 4 T; 64 7; 4l [+ 2em+f—3 em—+]-——5 em ——| 1, fi Evaluate the integral [ ‘S?dX fora linear shape function using (a) global coordinates and 1 (0) local coordinates. Starting with the equations T= bok +ext Ty = Cy + OXq + 3X} atakytox} solve for cy, 6p, and ¢s, and rearrange terms to vi fy the shape functions given by 5 = G(x ~ xa - XD 2x - xX =X) (X ~ XX — X)) For problem 3, use the Lagrange functions to derive the quadratic shape functions by the method discussed in Section Derive the expressions for quadratic shape functions in terms of the local coordinates and compare your results to the results given in Table 5.1. Verify the results given for one-dimensional quadratic natural shape functions in Table 5.1 by showing that (1) a shape funetion has a value of unity at its corresponding node and a value of zero at the other nodes, and (2) if we sum up the shape functions, we will come up with a value of unity. Verify the results given for the local cubic shape functions in ‘Table 5.1 by showing that (1) a shape function has a value of unity at its corresponding node and a value of zero at the other nodes and (2) if we sum up the shape functions, we will come up with a value of uni Verify the results given for the natural cubie shape functions in Table 5.1 by showing that (1) a shape function has a value of unity at its corresponding node and a value of zero at the other nodes and (2) if we sum up the shape functions, we will come up with a value of unity.272 Chapter S One-Dimensional Elements 9. Obtain expressions for the spatial derivatives of the quadratic and cubic shape functions. 10. As previously explained, we can increase the accuracy of our finite element findings either by increasing the number of clements used in the analysis to represent a problem or by using a higher order approximation. Derive the local cubic shape functions. Xj 11, Evaluate the integral US ‘$,4X for a quadratic shape function using (a) global coordinates, (b) natural coordinates and (¢) local coordinates. 12. Assume that the deflection of a cantilever beam was approximated with linear one- dimensional elements. The nodal deflections and their corresponding positions are shown in the accompanying figure. (a) What is the deflection of the beam at X = 2 ft? (b) Evaluate the slope at the endpoint. 100 1bitt 204 ort a= a El=2%10° I-82 y a 0 12 34 s |» 0.003275 om om sf =) 022075} in, | | | | Ys 0.034400 Jitter —$-laneelonl (os 0.046875 13. We have used linear one-dimensional elements to approximate the temperature distribution inside a metal plate. A heating element is embedded within a plate. The nodal temperatures and their corresponding positions are shown in the accompanying figure. What is the tem- perature of the plate at X = 25 mm? Assume that (a) linear clements were used in obtain- ing nodal temperatures and (b) quadratic elements were used. T, 120 tt z 9 B 116 qT, m1 & Tsp = 4 104 bc Ts 95 z 84 i Ts a ty 36 14. Quadratic elements are used to approximate the temperature distribution in a straight fin. The nodal temperatures and their corresponding positions are shown in the accompanying figure. What is the temperature of the fin at X = 7 em?ChapterS Problems 273 Th 100 qT, 74 > > Ss Be Tip =4 44 pec re 36 1 2 3 4 5 6 i) a Fi j2 em>f-2 em=|-2 em—[--2em-|--2, en |2 em qt 28 15. Develop the shape functions for a linear element, shown in the accompanying figure, using the local coordinate x whose origin lies at the one-fourth point of the element. 16. Using the natural coordinate system shown in the accompanying figure, develop the natural shape functions for a linear element. 17. In the accompanying figure, the deflection of nodes 2 and 3 are 0.02 mm and 0.025 mm, re~ spectively. What are the deflections at point A and point B, provided that linear elements were used in the analysis? = +10 em +} — 5 cm +} 7 cm ——>|274 Chapter5 One-Dimensional Elements 18. Consider the steel column depicted in the accompanying figure. Under the assumption of axial loading, and using linear elements, we determined that the downward vertical dis- placements of the column at various floor-column connection points are wy 0 0.03283 us » = § 0.05784 pin, Us 0.07504 Us 0.08442 Using local shape functions, determine the deflections of points A and B, located in the middle of clements (3) and (4) respectively. 30,000 Ib 1s ft om a0] @, Ise 2500010] @ sit dye J @ iste 1 | 19, Determine the deflection of points A and B on the column in problem 18 using natural coordinates. 20. A 20-f1-tall post is used to support advertisement signs at various locations along its height, as shown in the accompanying figure. The post is made of structural stec! with a modulus of elasticity of E = 29 X 10° Ib/in’, Not considering, wind loading on the signs and using linear elements, we determined that the downward deflections of the post at the points of load application are u 0 u| _) 6312 x 10“ us ) 871s x 10 [" Pn 11.470 x 104Chapter5 Problems 275 Determine the deflection of point A, located at the midpoint of the middle member, using (a) global shape functions, (b) local shape functions, and (c) natural shape functions. 1001 | 4) © Ap=075in? ate 1501 sit | 20016 | el 21. Evaluate the integral in problem 11 using Gauss-Legendre two-point formula, 22, Evaluate the given integral analytically and using Gauss-Legendre formula, fe + $x? + 10)dx t 23. Evaluate the given integral analytically and using Gauss-Legendre two- and three-point formulae. [ (3x4 + x? - 7x + 10)dx 24, In Section 5.6 we mentioned that when the function being integrated is known, the Gauss-Legndre formulae offer a more efficient way of evaluating an integral as compared to the trapezoidal method. The trapezoidal approximation of an integral deals with discrete data at uniformly spaced intervals and is computed from 1 1 [rede= iba tn to — + natmithn)276 Chapter 5 One-Dimensional Elements The above equation is known as the trapezoidal rule, and A represents the spacing among data points, yb. J. Yos-e+s and yy For problem 23, use ft = 1 and generate 11 data points: yy at x=-2y =1,..-+and yipait-x = 8. Use the generated data and the trapezoidal rule to approximate the integral. Compare this result to the results of problem 23. FQ) and cubic elements. Discuss the 25. Derive and plot the spatial derivatives for linear, quadral differences. femperature distribution along a section of a material is given by 7) = 80°C, 7, = 70°C, Ty = 02°C, Ty = 55°C with Xp = 0, X, = 2em, Xz = 4em, Xy = Gem. Use single tinear, quadratic, and cubic clements to approximate the given temperature distribution. Plot the ac- {ual data points and compare them to the linear, quadratic, and cubic approximations. Also, estimate the spatial derivative of the given data and compare it to the derivatives of the lin- car, quadratic, and cubie representations. 27. Use the isoparametric formulation to express the following information: 7, = 100°C, Ty = 58°C with the corresponding global coordinates X1 = 2.em, Xz = 5 em. Show the trans- formation equations between the global coordinate, the local coordinate, and the natural co- ordinate, 28. Use the isoparametric formulation to express the following deflection information: U, = 0.01 cm, U; = 0.025 em with the corresponding global coordinates X, = Sem, 12 em, Show the transformation equations between the global coordinate, the local 26.