- Six Sigma and Robust Design
- Identify Project and Organize Team
- Develop VOC Models
- Formulate Critical-to-Quality Characteristics
- Control Energy Transformation for Each CTQ Characteristic
- Determine Control and Noise Factors
- Assign Control Factors to the Inner Array
- Summary and Road Map
- Bibliography
This chapter is from the book
This chapter is from the book
This chapter is from the book
1.5 Control Energy Transformation for Each CTQ Characteristic
To fulfill the intent of the system, the customer does something that initiates a transfer of energy, which produces a CTQ that might be categorized as either the intended result or the unintended result (error state). Because energy transfer creates CTQs, the system must be studied in terms of this transfer. Such a study will help the team identify a response that quantifies the system's production of intended results. It is important to consider the following:
- Energy can neither be created nor destroyed.
- Energy can be transformed into various states.
- Only one energy state is intended, or ideal.
Maximizing the amount of energy used to produce an intended result will minimize the amount available to produce unintended results, or error states (Figure 1-4). Robust design shifts from examining the error states and searching for remedies, to studying the functional intent of the system and exploring ways of optimizing it.
)
Figure 1-4 Control energy transformation effectively.
Engineering robust products with Six Sigma requires a shift from measuring the symptoms of poor quality to measuring the transformation of energy. This philosophy requires a shift in thinking by the engineer. Robust design enhances quality through a focus on optimizing the system's intended functions—the efficient transfer of all energy.
To depict the intended function in terms of engineering metrics, study the underlying physics of a system, which should yield an engineering metric that quantifies the amount of energy used to produce a result. Use this metric as the Critical-to-Quality characteristic. Maximizing such a CTQ will optimize system functionality.
Clearly, CTQ characteristics depend on the system chosen for study. Many systems are composed of several subsystems and related processes, each with its own intended function. Therefore, what to study must be determined before the team can identify the transfer function and the related CTQs.
The following are the three types of metrics commonly used in industry:
- Customer metrics— usually subjective and expressed in nontechnical terms
- Management metrics— typically related to productivity or economics
- Engineering metrics— quantitative, objective, and physics-based
All of these metrics have their place in the development of quality products and processes. In experimentation though, engineering metrics will provide more useful and reproducible information than either management or customer metrics.
Levels are the different settings a factor can have. For example, if you want to determine how the response (speed of data transmittal) is affected by the factor (connection type), you need to set the factor at different levels (e.g., modem and local area network).
Example 1.1
Consider a robust design experiment with the objective to reduce the production of defective donuts. Suppose the management metric yield is the CTQ and the factors (time and temperature) are both tested at two levels.
Yield is the percentage of a product that is free of defects (i.e., the percentage of defect-free products over the total number of products produced). At the low temperature, B1, increasing time (from A1 to A2) increases yield; whereas, at the high temperature, B2, increasing time (from A1 to A2) decreases yield. So, the effect of cook time (on yield) depends on the temperature level. This implies that cook time and temperature interact, as is indicated by the nonparallel nature of the lines on the interaction plot of level combinations (see Figure 1-5). An interaction occurs when the response achieved by one factor depends on the level of the other factor. On the interaction plot, when lines are not parallel, there's an interaction.
)
Figure 1-5 Time and temperature interact with respect to yield.
Suppose, instead, that the CTQ is defined as "color." With color as the CTQ, increasing cook time increases color at either temperature, and increasing temperature increases color at either level of cook time. An interaction occurs when the response achieved by one factor depends on the level of the other factor. On the interaction plot, when lines are not parallel, there's an interaction. As shown in Figure 1-6, the factors interact little, if at all, with respect to the color engineering metric. Similar results would follow if the CTQ were another engineering metric such as moisture or density. In this example, with an engineering metric as Critical-to-Quality rather than a customer metric as Voice-of-Customer, the size of interactions can be reduced.
)
Figure 1-6 The interaction between color and temperature; the effect of cook time is to increase color at either temperature.
In sum, the Voice-of-Customer is what the customer wants. Systems transform the intent into the perceived result for VOCs. The perceived result is what the customer gets. A Critical-to-Quality characteristic is an engineering metric that quantifies the output of the system. A VOC is expressed in nontechnical terms and is frequently subjective. CTQ characteristics are expressed in technical terms and should (1) be related to the perceived results for VOCs, (2) quantify energy transfer, and (3) be an engineering metric.