2. Design of experiments

When intending to make a decision about an entity based on measurement, it is often not only a question of performing just one measurement of a single test object.

Minimising errors. Planning for reliability

Often there are several tests in a planned trial, for example a factorial trial, in which case the measurements on different test objects and possible repeated measurements of a particular object must be performed in such a way that the effects of measurement error and variations in the studied object can be minimised as well as estimated reliably.

A particular challenge is to separate reliably the sought, intrinsic entity variation from apparent variations arising from limited measurement quality.

Lecture*

– Specifications of Process and Measurement Capabilities

Balansing measurement efforts against available resources and demands

Design of experiment also includes the formulation of:

  • a realistic specification of the scope of measurements in relation to available resources and demands on process and measurement capabilities
  • Optimised uncertainty

 

Processes of production &  measurement
Specification of measurement problem
Design of experiments
Specification of demands on a measurement system
Choice/development of a measurement method
Evaluation (validation) of a measurement method
Verification
Performing the measurements
Presentation of measurement results
Use of measurement results. Conformity assessment

_________________________________________________________________

*Contact leslie.pendrill@sp.se for password
Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: