Let's keep those manufacturing defects to a minimumMarch 2006Designing for Six Sigma. Take steps to improve your position in the marketplace and avoid costly waste by ensuring that your product is not only manufactured as near perfectly as it can be, but also performs faultlessly once it leaves your factory gate Six Sigma was pioneered in the 1980s by Motorola. It is a rigorous and systematic methodology that applies statistical analysis techniques to data drawn from the manufacturing process. The aim is to measure, and subsequently improve upon, a company's performance by identifying and then taking steps to prevent defects - either in manufactured products or business services. To achieve the ultimate goal of Six Sigma, a product manufacturing company, for example, must not produce more than 3.4 defective parts per million. The key Six Sigma operations are to measure and analyse. If you are trying to improve an existing process you use this information to take better control over your manufacturing procedures; if you are setting up a new production line, then you utilise it in the design of your production processes. Any company can implement the technique, no matter what their size, and benefit in terms of reputation in the marketplace, customer satisfaction and waste reduction.
But can the designer have any influence on the attainment of Six Sigma, or is the methodology purely the province of the production engineer? Well, the designer can actually have a very big impact on a successful Six Sigma implementation. Design for Six Sigma (or 'Robust Design') designs quality into a product. By assessing the variations that a product experiences as it passes through the manufacturing process - or during use - it is possible to create a product that performs faultlessly regardless of these variations.
Ansys' (www.ansys.com) DesignXplorer VT ('Variational Technology') and Design Xplorer software modules, which work within the Ansys Workbench simulation environment, allow users to create these robust designs by means of a process that defines both design and uncertainty variables. The software then optimises a set of reliable goals for parameters such as fatigue life, stress or deflection. DesignXplorer VT lets the user perform quick and accurate 'what-if' scenarios to test their design ideas, while DesignXplorer allows the user to perform optimisation and Design for Six Sigma with any application via a third party plug-in.
NWA Quality Analyst, which is supported in the UK by Adept Scientific (nwa.adeptscience.co.uk) is a statistical process control (SPC) charting and analysis software product - now in version 6 - that meets the needs of users engaged in process improvement programmes such as Six Sigma. The latest version has improved drill-down and chart annotation in the Graphics Viewer, the inclusion of text reports in grouped output and the addition of 'assignable cause/corrective action' capability, which directly supports quality compliance programmes such as Six Sigma. And as it is now more pressing to integrate SPC with other quality-related applications, Version 6 offers improved database connectivity. In particular, it is now much easier to read specifications from external databases, and there is automatic linking of SPC data with appropriate records in these external databases.
Excel is a powerful modelling tool that can help you define a process and frame a decision, but Excel alone won't help you understand how variable or uncertain inputs will affect your forecasts. Crystal Ball (www,decisioneering.com/sixsigma) is a suite of Excel add-ins that will turn your existing engineering and process spreadsheet models into dynamic forecasting tools, while retaining all of Excel's functionality.
At its basic level, Crystal Ball lets you define uncertain inputs in your spreadsheet as probability distributions (rather than average values); these more realistically represent the possible values of the inputs. The software uses the defined distributions and Monte Carlo simulation to create thousands of possible alternative outcomes for your model; for example, if you are simulating how tolerance will impact a new product design, you can use Crystal Ball to generate, say, 5,000 scenarios that represent 5,000 products. By analysing the statistics of your simulation you can quantify the likelihood that your new product will meet system requirements.
Crystal Ball includes a powerful set of statistical and graphing tools that let you create and present your risk models. The software adds its own toolbar and menus, and the output charts, graphs and reports are all created in native Excel formats.
Dr Know's recommended download is Quality Analysts 6.1 Demo. Download yours today:
http://www.adeptscience.co.uk/download/dlddsp/0/All/NWA=Quality+Analyst+6.1+Demo.html
Most Viewed Articles...
<< Back to Dr Know Archive index
|