Home Michell truss type theories as a Γ-limit of optimal design in linear elasticity
Article
Licensed
Unlicensed Requires Authentication

Michell truss type theories as a Γ-limit of optimal design in linear elasticity

  • Heiner Olbermann EMAIL logo
Published/Copyright: April 17, 2020

Abstract

We show how to derive (variants of) Michell truss theory in two and three dimensions rigorously as the vanishing weight limit of optimal design problems in linear elasticity in the sense of Γ-convergence. We improve our results from [H. Olbermann, Michell trusses in two dimensions as a Γ-limit of optimal design problems in linear elasticity, Calc. Var. Partial Differential Equations 56 2017, 6, Article ID 166] in that our treatment here includes the three-dimensional case and that we allow for more general boundary conditions and applied forces.

MSC 2010: 49J20; 74B10

Communicated by Jan Kristensen


A Derivation of the variational form of the compliance minimization problem

Here we repeat basically our presentation from [23, Section 2.4]. We include this part in order to keep the present article self-contained.

Let g W - 1 , 2 ( Ω ¯ ; n ) . The aim of this appendix is to give a derivation of the compliance minimization problem in its variational form

(A.1) inf 𝒢 λ ; g ( σ ) ,

where the infimum is taken over the set

S g ( Ω ) = { σ L 2 ( Ω ; sym n × n ) : - div σ = g } .

Here the equation - div σ = g is to be understood as an equation in the distributional sense in a neighborhood of Ω ¯ , with σ extended by 0 on the complement of Ω. In this way we incorporate boundary conditions in the equation; see our discussion in Section 1.1. We want to derive this variational problem starting from the standard formulation of a linear elasticity problem. More details can be found in [2].

Consider Ω n as an elastic body, characterized by its elasticity tensor A 0 Lin ( sym n × n ; sym n × n ) , where for simplicity we assume here that A 0 = Id sym n × n is the identity. We remove a subset H Ω from the elastic body and the new boundaries from that process shall be traction-free. The resulting linear elasticity problem is to find u : Ω H n such that

σ = A 0 e ( u ) ,
- div σ = g in  Ω ¯ H ¯ ,
σ n = 0 on  H ,

where e ( u ) = 1 2 ( u + u ) T . The compliance (work done by the load) is given by

c ( H ) = Ω g u d 1 = Ω H ( A 0 e ( u ) ) : e ( u ) d x ,

where u : Ω H 2 is the unique solution to the linear elasticity system above. We want to minimize the compliance under a constraint on the “weight” 2 ( Ω H ) . We do so by the introduction of a Lagrange multiplier λ, and are interested in the minimization problem

min H ( c ( H ) + λ 2 ( Ω H ) ) .

The “equivalence” between the mass constrained problem and the problem including a Lagrange multiplier only holds on a heuristic level; see [17, 18, 19] for a discussion of this point. Accepting this step, taking the limit of vanishing weight corresponds to the limit λ . We now rewrite the problem by considering space-dependent elasticity tensors of the form A ( x ) = χ ( x ) A 0 , where χ L ( Ω ; { 0 , 1 } ) . The equations from above turn into the system

(A.2) { σ = A ( x ) e ( u ) , - div σ = g in  Ω ¯ .

The compliance turns into a functional on the set of permissible elasticity tensors, and is given by

c ( A ) = Ω ( A ( x ) e ( u ) ) : e ( u ) d x ,

where u is the solution of (A.2). By the principle of minimum complementary energy, the compliance can be written as

c ( A ) = Ω G ( A ( x ) , σ ( x ) ) d x ,

where

G ( A ¯ , ξ ) = { + if  ξ 0  and  A ¯ = 0 , 0 if  ξ = 0  and  A ¯ = 0 , ( A ¯ - 1 ξ ) : ξ else, 

and σ L ( Ω ; sym n × n ) is a solution of the PDE

- div σ = g in  Ω ¯ ,

i.e., σ S g ( Ω ) . We see that the compliance minimization problem can be understood as the variational problem of finding the infimum

inf { Ω ( G ( χ ( x ) A 0 , σ ( x ) ) + λ χ ( x ) ) d x : χ L ( Ω ; { 0 , 1 } ) , σ S g ( Ω ) } .

Of course, the compliance of a pair ( χ , σ ) is infinite if there exists a set of positive measure U such that χ = 0 and σ 0 on U. Hence the above variational problem is equivalent with

(A.3) inf { Ω F λ A 0 ( σ ) d x : σ S g ( Ω ) } ,

where

F λ A 0 ( ξ ) = { 0 if  ξ = 0 , ( A 0 - 1 ξ ) : ξ + λ else.

Up to a factor λ - 1 / 2 , this is just the integrand (3.5), and hence (A.3) is just the variational problem (A.1), with A 0 = Id sym n × n . As is well known, this problem does not possess a solution in general and requires relaxation.

B A very brief presentation of Michell trusses

In this section, we want to sketch very briefly how the limit integral functional 𝒢 , g is linked to Michell truss theory for the case n = 2 . What we say here is mainly taken from [9].

A truss is a finite union of bars (line segments that can resist compression or tension parallel to them) between points x i 2 , i = 1 , , M . We write ( x 1 , , x m ) = x 2 × M , and let w sym M × M . To every bar [ x i , x j ] = { t x i + ( 1 - t ) x j : t [ 0 , 1 ] } , we associate w i j , where | w i j | is the strength of the bar, and the sign of w i j is chosen according to whether the bar has to withstand compression or tension.

The force, provided the bar [ x i , x j ] , is given by

f i j ( x , w ) = ( δ x i - δ x j ) w i j x i - x j | x i - x j | .

The set of all bars shall counterbalance a given force

g = i = 1 N g i δ y i ,

where y i , g i 2 , i = 1 , , N , are given. The truss ( { x i } i = 1 , , M , { w i j } i , j = 1 , , M ) withstands g if

g + i , j = 1 , , M f i j ( x , w ) = 0 .

The weight of the truss ( x , w ) is given by

𝒲 ( x , w ) = i , j = 1 M | w i j | | x i - x j | .

The task is now the minimization of the weight, given the external forces, as a function of x , w . To express how this variational problem relates to 𝒢 , g , we note that the force supplied by the bars can be written as the divergence of a stress, f i j ( x , w ) = div σ i j ( x , w ) with

σ i j ( x , w ) = w i j x i - x j | x i - x j | x i - x j | x i - x j | 1 | [ x i , x j ]
σ ( x , w ) = i , j = 1 M σ i j ( x , w ) ( 2 ; sym 2 × 2 ) .

With this notation, the balance of forces becomes the equation

- div σ = g ,

and the weight of the truss is given by 𝒲 ( x , w ) = | σ ( x , w ) | (the total variation of the measure σ).

Summarizing, we are dealing with the variational problem

inf { | σ | : σ = σ ( x , w )  for some truss  ( x , w ) , - div σ = g } .

To guarantee the existence of a minimizer, this variational problem requires relaxation, as has already been remarked by Michell in 1904 [22]. We will not discuss the derivation of the relaxation here and refer the interested reader to [9]. We only state the result: Namely, that it becomes the variational problem defined by the Γ-limit in the main text:

inf { ρ ( 2 ) ( σ ) ( 2 ) : σ ( 2 ; sym n × n ) , - div σ = g } .

Requiring additionally supp σ Ω ¯ leads to

inf { ρ ( 2 ) ( σ ) ( Ω ¯ ) : σ ( Ω ¯ ; sym n × n ) , - div σ = g } .

References

[1] J. J. Alibert and G. Bouchitté, Non-uniform integrability and generalized Young measures, J. Convex Anal. 4 (1997), no. 1, 129–147. Search in Google Scholar

[2] G. Allaire, Shape Optimization by the Homogenization Method, Appl. Math. Sci. 146, Springer, New York, 2002. 10.1007/978-1-4684-9286-6Search in Google Scholar

[3] G. Allaire, E. Bonnetier, G. Francfort and F. Jouve, Shape optimization by the homogenization method, Numer. Math. 76 (1997), no. 1, 27–68. 10.1007/978-1-4684-9286-6Search in Google Scholar

[4] G. Allaire and R. V. Kohn, Optimal design for minimum weight and compliance in plane stress using extremal microstructures, European J. Mech. A Solids 12 (1993), no. 6, 839–878. Search in Google Scholar

[5] L. Ambrosio, N. Fusco and D. Pallara, Functions of Bounded Variation and Free Discontinuity Problems, Oxford Math. Monogr., Clarendon Press, Oxford, 2000. 10.1093/oso/9780198502456.001.0001Search in Google Scholar

[6] A. Arroyo-Rabasa, G. De Philippis and F. Rindler, Lower semicontinuity and relaxation of linear-growth integral functionals under PDE constraints, Adv. Calc. Var. (2018), 10.1515/acv-2017-0003. 10.1515/acv-2017-0003Search in Google Scholar

[7] J.-F. Babadjian, F. Iurlano and F. Rindler, Concentration versus oscillation effects in brittle damage, preprint (2019), https://arxiv.org/abs/1906.02019. 10.1002/cpa.21953Search in Google Scholar

[8] G. Bouchitté, Optimization of light structures: The vanishing mass conjecture, Homogenization 2001, GAKUTO Internat. Ser. Math. Sci. Appl. 18, Gakkotosho, Tokyo (2003), 131–145. Search in Google Scholar

[9] G. Bouchitté, W. Gangbo and P. Seppecher, Michell trusses and lines of principal action, Math. Models Methods Appl. Sci. 18 (2008), no. 9, 1571–1603. 10.1142/S0218202508003133Search in Google Scholar

[10] G. De Philippis and F. Rindler, On the structure of 𝒜 -free measures and applications, Ann. of Math. (2) 184 (2016), no. 3, 1017–1039. 10.4007/annals.2016.184.3.10Search in Google Scholar

[11] R. J. DiPerna and A. J. Majda, Oscillations and concentrations in weak solutions of the incompressible fluid equations, Comm. Math. Phys. 108 (1987), no. 4, 667–689. 10.1007/BF01214424Search in Google Scholar

[12] I. Fonseca, G. Leoni and S. Müller, 𝒜 -quasiconvexity: Weak-star convergence and the gap, Ann. Inst. H. Poincaré Anal. Non Linéaire 21 (2004), no. 2, 209–236. 10.1016/j.anihpc.2003.01.003Search in Google Scholar

[13] I. Fonseca and S. Müller, Quasi-convex integrands and lower semicontinuity in L 1 , SIAM J. Math. Anal. 23 (1992), no. 5, 1081–1098. 10.1137/0523060Search in Google Scholar

[14] I. Fonseca and S. Müller, Relaxation of quasiconvex functionals in BV ( Ω , 𝐑 p ) for integrands f ( x , u , u ) , Arch. Ration. Mech. Anal. 123 (1993), no. 1, 1–49. 10.1007/BF00386367Search in Google Scholar

[15] I. Fonseca and S. Müller, 𝒜 -quasiconvexity, lower semicontinuity, and Young measures, SIAM J. Math. Anal. 30 (1999), no. 6, 1355–1390. 10.1137/S0036141098339885Search in Google Scholar

[16] W. S. Hemp, Optimum Structures, Clarendon Press, Oxford, 1973. Search in Google Scholar

[17] R. V. Kohn and G. Strang, Optimal design and relaxation of variational problems. I, Comm. Pure Appl. Math. 39 (1986), no. 3, 113–137. 10.1002/cpa.3160390107Search in Google Scholar

[18] R. V. Kohn and G. Strang, Optimal design and relaxation of variational problems. II, Comm. Pure Appl. Math. 39 (1986), no. 3, 139–182. 10.1002/cpa.3160390202Search in Google Scholar

[19] R. V. Kohn and G. Strang, Optimal design and relaxation of variational problems. III, Comm. Pure Appl. Math. 39 (1986), no. 3, 353–377. 10.1002/cpa.3160390305Search in Google Scholar

[20] J. Kristensen and F. Rindler, Characterization of generalized gradient Young measures generated by sequences in W 1 , 1 and BV, Arch. Ration. Mech. Anal. 197 (2010), no. 2, 539–598. 10.1007/s00205-011-0477-0Search in Google Scholar

[21] T. Lewiński, T. Sokół and C. Graczykowski, Michell Structures, Springer, Cham, 2018. 10.1007/978-3-319-95180-5Search in Google Scholar

[22] A. G. Michell, The limits of economy of material in frame-structures, Phil. Mag. S. 8 (1904), 589–597. 10.1080/14786440409463229Search in Google Scholar

[23] H. Olbermann, Michell trusses in two dimensions as a Γ-limit of optimal design problems in linear elasticity, Calc. Var. Partial Differential Equations 56 (2017), no. 6, Article ID 166. 10.1007/s00526-017-1266-xSearch in Google Scholar

[24] D. Preiss, Geometry of measures in 𝐑 n : Distribution, rectifiability, and densities, Ann. of Math. (2) 125 (1987), no. 3, 537–643. 10.2307/1971410Search in Google Scholar

[25] G. I. N. Rozvany, Structural Design via Optimality Criteria. The Prager Approach to Structural Optimization, Mech. Elastic Inelastic Solids 8, Kluwer Academic, Dordrecht, 1989. 10.1007/978-94-009-1161-1Search in Google Scholar

[26] W. P. Ziemer, Weakly Differentiable Functions. Sobolev Spaces and Functions of Bounded Variation, Grad. Texts in Math. 120, Springer, New York, 1989. 10.1007/978-1-4612-1015-3Search in Google Scholar

Received: 2019-09-12
Revised: 2020-01-15
Accepted: 2020-03-24
Published Online: 2020-04-17
Published in Print: 2022-07-01

© 2020 Walter de Gruyter GmbH, Berlin/Boston

Downloaded on 23.9.2025 from https://www.degruyterbrill.com/document/doi/10.1515/acv-2019-0074/html?lang=en
Scroll to top button