Optimal Subassembly Partitioning of Space Frame Structures for In-Process Dimensional Adjustability and Stiffness

[+] Author and Article Information
Naesung Lyu

Department of Mechanical Engineering, University of Michigan, Ann Arbor, MI 48109-2125nlyu@umich.edu

Byungwoo Lee

 General Electric Global Research Center, Niskayuna, NY 12309leeb@research.ge.com

Kazuhiro Saitou1

Department of Mechanical Engineering, University of Michigan, Ann Arbor, MI 48109-2125kazu@umich.edu


Corresponding author.

J. Mech. Des 128(3), 527-535 (Aug 11, 2005) (9 pages) doi:10.1115/1.2181599 History: Received August 04, 2004; Revised August 11, 2005

A method for optimally synthesizing multicomponent structural assemblies of an aluminum space frame (ASF) vehicle body is presented, which simultaneously considers structural stiffness, manufacturing and assembly costs and dimensional integrity under a unified framework based on joint libraries. The optimization problem is posed as a simultaneous determination of the location and feasible types of joints in a structure selected from the predefined joint libraries, combined with the size optimization for the cross sections of the joined structural frames. The structural stiffness is evaluated by finite element analyses of a beam-spring model modeling the joints and joined frames. Manufacturing and assembly costs are estimated based on the geometries of the components and joints. Dissimilar to the enumerative approach in our previous work, dimensional integrity of a candidate assembly is evaluated as the adjustability of the given critical dimensions, using an internal optimization routine that finds the optimal subassembly partitioning of an assembly for in-process adjustability. The optimization problem is solved by a multiobjective genetic algorithm. An example on an ASF of the midsize passenger vehicle is presented, where the representative designs in the Pareto set are examined with respect to the three design objectives.

Copyright © 2006 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Figure 1

Two assembly sequences for automobile floor pan design (modified from (2)), where the total length is the critical dimension (KC). (a) Poor design (cannot adjust total length in the final assembly) and (b) better design (can adjust total length in the final assembly).

Grahic Jump Location
Figure 2

(a) A sample space frame structure with two KCs (KC1, KC2), and (b) a possible components set with 4 components, shown in different shades

Grahic Jump Location
Figure 3

Potential joint locations (grey boxes) and possible joint types at each location (joint library). Arrow(s) near each joint type indicates the adjustable direction during assembly.

Grahic Jump Location
Figure 4

(a) Seven basic members and (b) structural topology graph with seven nodes and ten edges

Grahic Jump Location
Figure 5

Selected joint types and topology graph with the corresponding edges removed

Grahic Jump Location
Figure 6

(a) four subgraphs and (b) corresponding four components

Grahic Jump Location
Figure 7

A sample structure with four components and its beam-spring FE model

Grahic Jump Location
Figure 8

(a) Basic members and (b) corresponding configuration graph C=(M,T,A). (c) A sample components set design and (d) corresponding liaison graph L0.

Grahic Jump Location
Figure 9

Partitioning with best adjustability

Grahic Jump Location
Figure 10

Detailed procedures for finding the optimal cut in the first partitioning of Fig. 9

Grahic Jump Location
Figure 11

Binary tree representation of the subassembly partitioning illustrated in Fig. 1

Grahic Jump Location
Figure 12

(a)ASF for a passenger car and (b) simplified frame model used in the case study with 5 KCs(KC1∼KC5)

Grahic Jump Location
Figure 13

(a) potential joint locations and (b) possible joint types (joint library) for each location type. Arrows in (b) indicate the adjustable directions.

Grahic Jump Location
Figure 14

(a) Beam cross-sectional design variable (beam thickness yi0 and yi1) and (b) joint design variable (weld thickness, zi0 and zi1)

Grahic Jump Location
Figure 15

Convergence history of GA run for the case study

Grahic Jump Location
Figure 16

Designs at the terminal condition (generation=50). Pareto solutions are colored as darker dots.

Grahic Jump Location
Figure 17

Individual designs from Pareto set. (a) R1 with 12 components and (b) R2 with 22 components.

Grahic Jump Location
Figure 18

Assembly partitioning of R2: (a) six subassemblies and (b) binary tree representation




Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In