ASME B89.4.10:2021 pdf free download

ASME B89.4.10:2021 pdf free download.Methods for Performance Evaluation of Coordinate Measuring System Software.
4.5.4 ComputationaL Effort. The effort or time required to compute the results is affected by the three previous factors. Fxcessive computing effort can adversely affect throughput of a CMS.
5 TEST METHODOLOGIES
This section establishes the general principles, procedures and practices for testing the performance of CMS software.
5.1 Test PrincipLes
For the purposes of this Standard, CMS software is evaluated strictly in terms outs intended function. No assumptions are made regarding the internal structure or operation of the software. The software is subjected to variations of inputs while its outputs are evaluated with respect to a specified objective.
CMS software Is tested by the Input ofsets of test data that reflect the expected range and variability of actual data. Such testing cannot guarantee that the software is completely error free because exhaustive functional testing is impossible.
5.2 Apparatus
The apparatus shall be a testing system interacting with the software under test through exchange of data sets and fit results. Software under test shall be executed in the computing environment in which it will be used. Modifications of the software under test. ifany, are limited to those necessary to input the supplied data sets and to extract the fit results. Such modifications shall not change the fit results from what would be produced by the software under test when presented the same data in a production environment.
The data formats, data resolution, and related characteristics of the test must be defined prior to its execution. The person operating the software under test shall be trained in the operation of that software to the extent necessary to input the data, run the software, and gather the output (fit results) in the required format. Figure 5.2-1 illustrates the following major components of a functional CMS software testing system:
(a) a Reference Pair Generator (RPG) capable of producing reference pairs of data sets and fit results, for specified feature types with controlled range and variability
(b) a means to transfer data sets to, and receive fit results from, the software under test
fr) a comparatordesigned to compare the results of the software under test to the reference results with respect to the objective function and generate an appropriate report
5.2.1 Reference Pair Generation and Validation
5.2.1.1 Reference Pair Generation. The RPG must be capable of creating a data set and fit. This may be done in two ways.
(aS) A predefined fit result is pocessed to produce a data set that meets the fit criteria.
(b) A data set Is generated that approximates the feature, and a reference fit is generated from the data by a reference algorithm.