Medical Device Link.

MPMN: March 1997

Development of Blood-Gas Analyzers Benefits from DFMA Software

Hospitals will always want equipment featuring the latest technology, but with shrinking healthcare dollars, they are being forced to limit their equipment budgets. Medical device manufacturers are therefore in the difficult position of having to meet conflicting demands: providing market-leading technology cheaply. Ciba Corning Diagnostics Corp. (Medfield, MA) sought to meet both demands when it began work on its new line of blood-gas analyzers, the 800 series. Instrumental in achieving these tough goals was their use of Design for Manufacture and Assembly (DFMA) software from Boothroyd Dewhurst Inc. (Wakefield, RI).

David Yeo, the manufacturing project manager at Ciba Corning, championed the use of DFMA in developing the analyzers. The design team established high goals: besides being durable and easy to use, the analyzers had to be easy to manufacture, assemble, and upgrade.

In order for the three analyzers (Models 840, 850, 860) to be easy to upgrade, they were all based on the same platform. Thus, if a lab using the 840 model finds that, in addition to analyzing blood gas and pH, it also needs to analyze electrolytes, it can simply add the needed analyte to the machine. Likewise, if it needs to analyze hemoglobin, it can upgrade its machine with a CO-oximeter module.

"If you look at the benchmarks for the 800 project," says Yeo, "you can see that we met and surpassed some very aggressive assembly and cost targets. The measurement module on the 800 series, for example, is currently being produced with fewer than half the parts of the 200-series [the previous model] measurement module. The result is an 81% reduction in assembly operations and a 20% reduction in assembly time and labor costs for this subassembly. In addition, it holds three more analytes than its predecessor."

Yeo began by arranging a two-day training course that focused on a team approach. Design engineers at Ciba Corning were paired with manufacturing engineers during the course and for the rest of the project. A computer station was set up so that everyone involved with the project could work together.

Yeo says that the groups embraced the DFMA concept right from the start. "It was a very effective brainstorming tool because it created debate and consensus resolution. As we were doing the analysis, we worked through a range of direct and secondary issues and made key manufacturability decisions."

Critical design changes were made by using DFMA time and cost databases for assembly, and by employing basic Design for Assembly (DFA) guidelines. Those guidelines ask three questions about a design: Does the part move relative to all the other parts already assembled, must it be made of a different material, and must it be separate from all other parts?

"The designers really stuck to these rules. They made it one of their goals to design with manufacturing and assembly in mind."

Unlike the previous 200-series analyzer, which had separate panels for the top, bottom, and all four sides, the new analyzer consists of just two pieces hinged together like a clamshell. Yeo says that the new analyzer's biggest improvement is in its highly functional structural-foam molded parts that make up its outer shell. The upper and lower caseworks that make up the two pieces of this clamshell construction are designed with features that allow a top-down assembly approach for the other components.

The hydraulic system that is located in the upper caseworks benefited from some of the most important changes made as a result of using DFMA. The designers sought to reduce the number of hydraulic connections because they saw each one as a potential leak path.

In the product's measurement module, the number of assembly operations for the fluidic system was reduced by 80%--even with the addition of three electrodes that increased system capability by 20%. Furthermore, for this subassembly, overall parts and assembly time were reduced by half.

"We ran all the subassemblies through the analysis and got estimates for assembly cost. But, most importantly, DFMA analysis promoted discussions about whether a material needed to be different, or whether parts could be incorporated into each other. Having trade-off numbers early on was extremely beneficial and helped us develop a fluidic circuit that eliminated the need for most of the tubing."

The fluidic system of the 800 series builds on the technology of the 200 series. The fluidic system's centerpiece is the reagent manifold, a solid acrylic block with pathways running through its center. This design eliminates the need for most of the tubing and several other parts that were used on the 200 series. Looking at the assembly, it is hard to imagine how curved pathways could be drilled through a clear, solid block.

After using DFA to brainstorm parts-consolidation strategies, Ciba Corning worked closely with its supplier, Eastern Plastics Inc. (Plainville, CT), to develop and produce the manifold using lamination. Eastern Plastics begins with a 1Ž2-in.-thick block of acrylic, and machines and polishes it flat. A CNC machine then cuts a U-shaped cross section into the block before another layer of acrylic is placed over it and heat-fused together.

The entire manifold self-aligns easily into the analyzer's upper caseworks. The caseworks has several hooks to ensure that the parts
fit snugly. Only three screws are needed to hold it in place, and they are easily serviceable since they are all accessible from the front.

The upper caseworks is a very complex molded composite that was developed using 3-D solids modeling. DFMA helped the design team define the trade-offs involved in designing the upper caseworks as a single, complex component. From solid-model geometry, stereolithography was used to develop the first model. Rubber molds and cast urethane functional prototypes were then created before the actual molded structural-foam components were made.

The overall improvements to the hydraulic system were dramatic. DFA analysis helped reduce the number of parts by 48% and the number of fluidic connections by 60%. According to Yeo, the 22% reduction of the fluidic system's standard cost is a direct result of applying DFA.

DFMA was also instrumental in developing the lower caseworks. The engineers used DFMA to compare alternatives. They discussed the trade-offs of molding in snap-fits versus molding in holes for screws or tinnerman fasteners. In several cases they decided on captive fasteners because they wanted to avoid having loose hardware fall into the electronic components during repair. This type of fastener also provided assembly advantages over traditional screws.

When the 200 and 800 series are laid out together on a bench, their differences are dramatic. The new model has fewer components, and most of them are self-aligning. "Answering the DFA questions helped bring out important features," says Yeo. Throughout the analysis, the team paid special attention to self-locating features because they didn't want to reorient any part more than necessary.

The success of the 800 series has prompted Yeo to look at DFMA as a strategy for future projects. The company may use DFMA software in relation to machining issues. "A machining module would have helped us even further in deciding if it was economical to produce separate parts. For this reason, we will look closely at future DFMA shape-forming modules."

Return to the MPMN page