Concept for IP Design Born of DFMA

Cost represents the major factor weighing on automotive designers' minds these days. A survey sponsored by DuPont Automotive at the recent SAE '98 proved that fact. But these aren't just raw material costs we're talking about--the concern spans the entire system cost of automotive assemblies, from concept to manufacture and even to disassembly. But how does a designer, pressured by the day-to-day task of meeting performance requirements and market demands, get his or her hands around such a complicated set of issues?

Fortunately, researchers have made the task a bit easier by going through the design for manufacture and assembly process in conceptual form. One such study in particular, conducted by Peter Dewhurst of the University of Rhode Island and sponsored by GE Plastics, offers to pave the way for better, more cost-efficient instrument panels.

IMM caught up with Dewhurst at the SAE show, where he presented the results of his study. "Our team considered the possibilities for redesigning a current-model, full-size truck instrument panel (IP) with two goals in mind--to reduce manufacturing costs and to increase the chance for end-of-life material recovery," he explains. "Keep in mind that this is a concept only, intended to suggest how IPs can be designed for moldability, assembly, and disassembly. Our hope is that skilled automotive designers will be able to incorporate portions of this concept into an optimized future design."

Before going into greater detail on the design itself, shown in Figure 1, check out the results Dewhurst and his team compiled with the new IP concept:

Concept Consolidation
Using software developed by Boothroyd Dewhurst, the research team subjected the original design to DFA (design for assembly) and DFE (design for environment) analysis. The current IP studied by URI's Design for Manufacture Center is a hybrid of injection molded plastic and metal subassemblies. The latter structures--welded steel tubes with a variety of components--were the first to be eliminated in the proposed redesign. One of the obvious benefits is parts consolidation, but this change also eliminated several assembly steps. The original truck IP contained seven main moldings and four steel subassemblies, which the proposed redesign replaces with four main moldings.

How did Dewhurst replace the structural steel members without sacrificing rigidity in the new IP design? The proposed design is based around two structural injection moldings, a back and front unit as shown in Figure 2. "The front molding consists of a horizontal rib structure that provides the required lateral bending stiffness and strength," Dewhurst says. "Cantilever snap elements to hold other components onto the front molding are only added to vertical ribs so that the horizontal rib structure is not compromised."

These two main moldings, along with a trim molding and steering column cover, were designed with ABS/PC. With wall thicknesses of .14 inch for the structural members and .08 inch for the trim, the combined weight would reach 17 lb. At an estimated cost of $1.25 per/lb, the raw material cost for these two main moldings exceeds the cost of individual manufactured parts in the original design. However, when assembly cost savings are considered, the higher-priced redesign actually costs less to produce (see Table I).

What about mold costs? Dewhurst considered them in developing part cost estimates. "We calculated mold cost estimates for both the current and proposed designs using DFMA software. Required investment in injection molds for the new design is roughly $500,000 more than the current one, because longer cycle times for the main structural moldings require more molds to meet production volumes." When tooling costs are amortized, however, the per part cost of the redesign is only 37 cents higher.

Comparing Assembly and Disassembly
Consolidating parts in the redesign may not have lowered the raw material or mold costs, but it did work wonders in the areas of assembly and disassembly. The current design requires operators to assemble 28 subassemblies onto the main assembly and consists of a total of 314 assembled parts. Also, the current design requires 137 separate assembly operations, for example, attaching electrical connectors, applying adhesive, etc. Applying DFA software, Dewhurst's study calculates the 482 assembly tasks take 3780 seconds at a burdened labor rate of $42/hour for a grand total of $44 assembly cost per IP.

On the other hand, the proposed IP redesign would require 121 assembly steps that take 730 seconds for a total of $8.52 per IP. According to Dewhurst, an assembly "bonus" occurs when the number of items and tasks is drastically reduced. "That's because it's possible to give workers adequate access to carry out assembly tasks without difficulty, so the assembly time reduces by even more than the proportionate decrease in the number of tasks."

As for disassembly, the results are equally stunning. Using the DFA analysis results as input, researchers performed DFE analysis to establish a disassembly procedure. In addition, the research team used the software to rate both designs on an MET scale (material use, energy use, toxicity effects). For the current IP, profitable disassembly occurred at 894 seconds with a maximum profit of $3.74. For the redesign, time to disassemble all recyclable materials dropped to 195 seconds, with a potential net profit of $6.39.

Dewhurst agrees with a researcher working on a structural plastic IP for the Dodge Dakota who told him that recovering a substantial amount of PC/ ABS is a much more attractive recovery target than the diverse material mix found in previous designs.