Hardware performance testing

The LC control software, either stand-alone or as part of an overall data-handling system, should be tested by means of a separate OQ protocol. This protocol only needs to address the communica-tions/control integrity of the hardware (e.g., setting up a run/sequence with the proper instrument parameters, the ability to start and stop the pump, etc.). It should cover all the required instrument control functions listed as part of the protocol s functional specifications. It does not need to include specific hardware performance testing, such as linearity or flow rate. The latter tests are performed separately, as part of the individual hardware validation described below. 

The most important hardware items appeared to be the detectors themselves. The gas detection system gave frequent spurious alarms, and on both platforms the ultraviolet (UV) fire detectors were also prone to spurious activation from distant hot work for example, and had a limited ability to detect real fires. The tmreliability of these systems had a general effect on response time and would, overall, lengthen the time to respond. The second aspect which was related to hardware was fimction and performance testing of the emergency blowdown systems. It is critical that the workers believe the systems will work when required, and this can only be achieved by occasional use or at least fimction testing. 

User requirement specifications (URS) for the computerized system are provided by the pharmaceutical firm to the computer systems vendor. The vendor generates functional and design specifications as a basis for designing and coding software for the computerized system. The system is then built, together with all the interfaces to the hardware, and tested by the vendor. After installation of the computerized system, IQ, OQ, and PQ are performed at the pharmaceutical facility to verify that the system is able to meet the URS and design and functional specifications. 

Hardware considerations for automated performance testing systems generally include the computer, software, monitor, and input devices. These features can be provided on multiple testing platforms, including personal computers, handheld personal digital assistants, Internet-based systems, or on platforms that are specially designed to simulate occupational contexts. 

Although not yet available, it is likely that Web-based performance testing systems will become available in the near future. Web-based systems will permit selected tasks to be presented on computers equipped with appropriate Web browser software. Hardware requirements include Internet access and appropriate memory and software to support Web-based applications. Depending on the design of the Web-based system, it is could be possible to tailor the specific tasks presented for the performance testing system from a menu of options. Alternatively, testing systems consisting of a standardized array of tasks can also be chosen. As such, the start-up costs of Web-based systems should be lower than with personal or handheld computer systems. Subject identification, date, and time can be recorded at the start of a test, and data from multiple subjects and test occasions can be stored in a central file for easy access to the data. 

The tests will be broken into Hardware Acceptance Tests (or Installation Qualification), System Acceptance Tests (or Operational Qualihcation), and Equipment Tests (or Performance Qualihca-tion). According to the size of the unit, all tests may be in one test document or there may be three separate Qualification documents. A test document will define the test philosophy and how the tests should be run. Each test will have a tide, a reason for the test, an oudine of any test equipment required, a description of the test, data to be recorded, and the test acceptance criteria. Calibrated test equipment must be supported with cahbration certificates. 

Validation of instrument hardware includes testing according to documented specifications. If HPLC instruments consist of several modules, individual modules (modular validation), as well as the entire system (holistic validation), should be validated. However, the latter is preferred, as individual module tests should be performed as part of the diagnosis if the system fails. 

This section is primarily intended for those who need to set-up experiments or those who have new hardware to install for which new calibrations are required. As with any analytical instrumentation, correct calibrations are required for optimal and reproducible instrument performance. All the experiments encountered in this book are critically dependent on the application of rf and gradient pulses of precise amplitude, shape and duration and the calibrations described below are therefore fundamental to the correct execution of these sequences. Periodic checking of these calibrations, along with the performance tests described in the following section, also provides an indication of the overall health of the spectrometer. 

QTS software validation specifications lack details such as the hardware platforms the software should run on (model number, CPU [central processing unit], RAM [random access memory], hard drive size, free hard drive space, etc.), the software operating system, software database specifications (such as number of records accommodated, number of fields, whether those fields are character or numeric or date or alphabetical fields), the number of concurrent users allowed, the development platform used to write the application, network information, report contents and other information needed to perform testing properly. 

IQ Testing. A record of the physical installation of the individual computer systems and their interconnections and interfaces is the result of IQ testing. The purpose is to uniquely identify the component parts of the individual systems, hardware, firmware and software and to verify that they were the items utilized in the supplier acceptance testing. It is often the case, however, that due to supplier acceptance failures, there may have been modifications to the software or replacement hardware performed under change control. 

For developing medical applications extensive research is performed in this field internationally. Before novel diagnostic methods can be developed in hardware and tested in vivo, research starts out with computer-models of the anatomy and physiology that are used to run computer-simulations of potential measurement configurations (testing in silica). 

After the assembly, the complete system was fully commissioned by performing the low level hardware tests followed by detailed functionality tests. The low level hardware tests were carried out using the standalone software and could be completed within less than one day while the detailed performance tests took about 3 h per sector. 

Testing schemes generally affect complete subsystems hence, consideration of each hardware element is unnecessary. Tests of redundant portions of a system are particularly important, and may be constrained by the technical specifications which must be reflected in the fault tree. Testing may require the reconfiguration of systems for the test, which may prevent the performance of their designed function. In this case, other members of the redundancy must be available, but may fail. Failure to restore a system after test significantly increases the risk. 

As the cost of testing vast quantities of equipment would be too great and take too long, qualification tests, particularly on hardware, are usually performed on a small sample. The test levels are varied to take account of design assumptions, variations in production processes and the operating environment. 

Many researchers choose to buy expensive GPC/SEC columns from one of the major producers because that producer s columns had been used in the past or because of a successful marketing campaign by one particular producer. It should be noted that repacked columns can be obtained for a fraction of the cost of new columns. American Polymer Standards repacked columns are guaranteed to perform just as well as new columns from any company. When a column is repacked the only parts reused are the stainless-steel tube and end caps. This hardware is then repacked using new frits and new ST-DVB gel. Each column is individually tested in a quality control laboratory and shipped in the customer s choice of solvent. American Polymer Standards offers a column repacking service because it is a practical, inexpensive way for customers to acquire state of the art GPC/SEC columns. 

A human error or reliability analysis (HRA) can be performed to identify points that may contribute to an accidental loss. Human errors may occur in all facets of a the hydrocarbon industry. They are generally related to the complexity of the equipment, human-equipment interfaces, hardware for emergency actions, and procedures for operations, testing and training. The probabilities of certain types of errors occurring are normally predicted as indicated in Table 29. Individual tasks can be analyzed to determine the probability of an error occurring. From these probabilities, consequences can be identified which detemline the risk of a particular error. 

Most blast door manufacturers opt to perform static load tests on prototype assemblies of low-range blast doors to demonstrate that the assembly will resist the blast overpressure specified. Static tests should be accepted only if the dynamic structural response and dynamic load factors have been considered and the door, frame, and restraining hardware are manufactured using the same materials, dimensions, and tolerances as those in the prototype static test. 

---