Risk Management through Design for Manufacture
As one of the nine knowledge areas of project management, risk management is certainly one of the most important. The Design for Manufacture (DFM) Phase of product development is essentially the management of project risks, including the functional, technical, operational, schedule, and financial risks.
The DFM engineer should gather the potential risks into a risk register and incorporate the appropriate reduction of those risks into the final product design. This final design serves as a key component of the Risk Management Plan.
A medical client recruited Natech to produce a drug delivery device for a combustible substance in a glass ampoule. The user had to be able to crush the glass ampoule and express the substance in a controlled, safe manner. The client’s major concern was that after use either the combustible substance or the broken glass shards could escape the device through one of the three join locations.
Given the existing constraints from the supplied components, the Natech engineers performed an initial review of the design. After analyzing the risks, the initial efforts focused on the joins because these were identified as the highest risk areas.
Of the three joins, the join closest to the point the user would pinch posed the highest risk. That was the join which would experience the most additional stress while the user crushed the ampoule. The tip and ampoule were fixed constraints which meant the application required a widening of the diameter moving from the tip toward the back where the ampoule was held.
As with any DFM process, the goal was to reduce the number of components. If the two components forming the high-risk join were made as a single component, a very long core would be required to form the distal end. In many instances experience can be a risk mitigator, and Natech has experience molding components with a very long draw and extremely low draft angles, so the parts were combined into one.
Contingency Planning by Design
The two remaining joins required hermetic seals to properly hold the material. The options included ultrasonic welding, laser welding, spin welding, adhesive, and compression. The team evaluated these for level of hermeticity, flash risk, cost, cycle time, and ampoule breakage risk.
While laser welding provided a high certainty of success, it also came with a high price tag. Spin welding offered a fast cycle time coupled with the greater possibility of creating flash. Compression was the preferred solution but was far from guaranteed to work. The more critical of the two joins was the join closer to the device tip because this would be where the substance would be expelled.
As part of the risk mitigation, the contingency plan was built into the design of this joint. First, it would be designed as a compression joint leveraging the existing “bump” geometry in the mating component. If the testing proved the compression fit to be inadequate, a spin shear joint would be available for a spin weld. If neither the compression fit nor the spin weld proved adequate, the feature would also allow for an adhesive joint. Rather than lose time and budget redesigning and adjusting the mold after test results, the same parts could be used for all three join methods to determine the best solution.
Moving to the mold design phase means identifying a new set of risks. Modeling the mold design and molding process is another method of risk management.
The team performed a mold flow analysis on the main body to anticipate the potential molding risks. Four variations of materials, setup parameters, and gating were modeled.
The analysis identified potential weld line and air trap risks toward the end of the main housing. Both temperature and pressure address these risks. A higher pressure and the optimal temperature range at the weld line will improve the strength of the weld line.
The Design of Experiments identified the processing window for the injection pressure, temperature at the flow front, and fill time. The model also predicted an optimal gating solution to maximize the processing window.
Once mold build was complete, parts were molded in three different custom material blends. Leak tests were set up in horizontal and vertical positions in both open and closed systems over a four-hour period.
In addition, a custom fixture was designed to collect and analyze the push force that each join could hold at a 95% confidence interval.
The test data confirmed that the compression fits would be adequate for both joins. Eliminating the most at-risk join reduced the number of components in the housing from four to three. This brought a cost savings by eliminating both a purchased item as well as an assembly step.
By looking at the DFM phase as a risk management effort, engineers can take a broader look at the project to more fully manage their project risks.