| QG12: Base-line Parameter Diagrams |
|
After a Design Concept is generated, its specific, controllable parameters (Xs) are documented with preliminary set points that are available for exploration, characterization and optimization. It is extremely important to define measurable, physics-based, continuous response variables for the function so that the model(s) representing the design are fundamental to the physical principles upon which the design is based. Later, quality measures and attributes can be defined for the function, but at this point in development it is essential to pay very close attention to the governing physics between the Xs and the measured Ys.
.jpg) | | Base-line P Diagram |
Once a Base-line set of control parameters are identified and fully illustrated and the NUD parameters highlighted, the Team can proceed with Modeling and Simulation as well as Designed Experimentation (DOE) on physical proto-types. Prior to building DOEs, we recommend applying the methods of Design for Additivity to the Control Parameters to assure the DOE is efficient as well as economical. Design for Additivity provides guidelines to develop "engineered or grouped" parameters that govern the physics of the function. This technique increases the likelihood of the DOEs producing strong results that are easily verified in verification tests.
It is common for analytical iteration of set point development until it becomes obvious that physical proto-types must be built and evaluated to confirm or adjust the models and fill in knowledge gaps. These analytical and empirical iterations of base-line design development are done under nominal lab conditions prior to exposing the design to stressful sources of variation contained in the Noise Diagrams.
The Base-line Parameter diagrams serve two purposes: 1) They are essential to constructing a coherent Design Guide so people understand how the design works under nominal lab conditions, and 2) They set the stage for conducting Modeling and Simulation and Designed Experimentation (DOEs) for improving the math models describing the function and its governing parametric coefficients, linear and quadratic (or higher order) terms, interactive terms and set points. This information underwrites the design's stability, adjustability, independence, interactivity and statistical significance of controlling Xs, as well as any hyper-sensitivity in the design.
|
| QG13: Robust Parameter Diagrams |
|
After Base-line parameter characterization, modeling and simulation and DOEs for screening and modeling are completed under nominal conditions, it is time to apply stressful noises to the design while certain controllable parameters - Xs - are changed. We need to identify the Xs that have a strong potential to affect the standard deviation of the measured Y during active changes in the noise parameters. Here we specifically study control factor interactions with noise parameters. This is an essential form of interaction analysis often overlooked by development teams.
Control parameters that have a very strong effect on the mean of the measured Y variables under nominal conditions will have been identified and validated from the aforementioned base-line characterization work (see QG12, above). The one or two strong mean-shifters will not be candidates for use on minimizing the standard deviation during active noise changes. We now begin to discriminate between types of Xs: Mean shifting Xs are called out as Xmean while Xs that are good at controlling the standard deviation are called out as Xsigma. Xsigma variables are controllable parameters found to have useful interactions with noise factors that can be illustrated as follows:
.jpg) | | Engineering Control & Noise Factors |
.jpg) | | Robust P Diagram |
Once a Robust Parameter Diagram is fully illustrated, the Team can construct specific Designed Experiments to explore the interactions between the controllable parameters and the noise parameters. This is done to identify regions of low sensitivity to noise, where the Ys standard deviation is at a minimum while the stressful noises are actively turned on during the experiments.
The Robust Parameter diagrams serve two purposes: 1) They are essential to constructing a coherent Design Guide so people understand how the design is made robust, and 2) They set the stage for conducting Modeling and Simulation and Designed Experimentation (DOEs) for robustness optimization.
|
| Links to Prior CPD&M Quick Guide Newsletter Issues |
|
There are 24 Critical Parameter Development & Management (CPD&M) Quick Guides being published in installments in this newsletter. Below are links to each of the prior newsletters with CPD&M Quick Guides:
The CPD&M Quick Guide TOC (Nov 2013)
CPD&M QG1&2: Intro & Process (Jan 2014)
QG3&4: Prioritize Req'ts & Design Guide (Feb 2014)
QG5&6: Functional Diagramming & Functions, Complexity & Risk (Mar 2014)
QG7&8: Fn's, Design Controls, DG O'view & I-O-C Diagrams (Apr 2014)
*Note: there was no May 2014 issue
QG9: Design Failure Modes & Effects Analysis (DFMEA) (June 2014)
QG10&11: Fishbone & Noise Diagrams (July 2014)
|
| |
Is there a topic you'd like us to write about? Have a question? We appreciate your feedback and suggestions! Simply "reply-to" this email. Thank you!
Sincerely,
|
Carol Biesemeyer
Business Manager and Newsletter Editor Product Development Systems & Solutions Inc. |
