Matavimų rezultatų patikimumas tiriant statybinių medžiagų degumą

Abstract

General characteristic of work Topicality of work Nowadays the harmonization within the fire field raise and we can feel increasing demands for accredited fire-testing laboratories to guarantee that the tests give repeatable results. In the new European classification system it is important to have high quality on fire testing within the whole of Europe. The quality assurance standard ISO/IEC 17025 demands that accredited laboratories have control over measurement uncertainty for the methods used within the laboratory. Nowadays there is quite complicated situation because of evaluation of measurement uncertainty in fire testing. For the fire testing method which have a formulas for calculating final results there are possibilities to calculate measurement uncertainty using members of formula as main sources of measurement uncertainty. But sometimes in this way calculated measurement uncertainties don’t show real situation about the precision of the measurement results because there are made a lot of “if” and assumptions. For fire test methods like EN/ISO 11925-2 “Ignitability of building products subjected to direct impingement of flame” where the result of the test is “the material ignites/does not ignite”, it is very difficult, if not impossible, to determine the total uncertainty for the test result. Reference materials would be one way of achieving this. Anyone who has experience in fire testing knows that most fire test methods don’t have a good precision and repeatability. Purpose and tasks. The purpose of the doctoral dissertation is to analyse two main fire test methods and to improve precision and reliability of test results. To realize the purpose the following scientific and technical tasks have been dealt: - to determine the most optimal method of measurement uncertainty evaluation for fire test methods; - to examine the reliability of two main fire test methods on the ground of Round Robin test results; - to create candidate reference materials for two main fire test methods; - to create main aspects and principles for development methodology of reference materials creation and measurement uncertainty evaluation for reaction to fire tests methods; Scientific originality. For the first time the reliability of SBI test method has been evaluated, i.e. measurement uncertainty, sources of measurement uncertainty, repeatability and reproducibility standard devations has been analysed in detail and sources of measurement uncertainty which contribute most to the uncertainty has been determined. Candidate reference materials for SBI and EN ISO 11925-2 test methods have been proposed and criteria for further development of reference materials for these tests methods have been determined. The most optimal method of measurement uncertainty evaluation has been determined and principles for development methodology of reference materials creation and measurement uncertainty evaluation for reaction to fire tests methods has been presented. Additional calibration requirements for SBI and EN ISO 11925-2 test equipments has been created and possible improvements of test method and test equipment have been proposed. Practice value. Evaluated measurement uncertainty of SBI test method will be useful information for European fire laboratories. Value of measurement uncertainty will give valuable additional information about SBI test method precision. Calculated standard uncertainties of separate sources will be useful for laboratory staffs which perform SBI equipment calibration, maintenance. Proposed candidate reference materials and criteria will be valuable information for further development of reference materials in the future. These proposed materials should be used in interlaboratory comparison tests and in fire laboratories on purpose maintain internal quality. In dissertation given principals and aspects of evaluation of measurement uncertainty and development of reference materials will be valuable background for preparation guidelines and standards of uncertainty evaluation and development of reference materials in reaction to fire testing field. Approval of work. 2 articles under review and 5 articles in different conference proceedings were published on research field of the thesis. The main results of the work were presented at the international and national conferences: - 3th Conference of Lithuanian junior scientists „Lithuania without Science –Lithuania without future“ – „Mechanical and material engineering, industry engineering and management“. Vilnius, Lithuania. October 12 – 13, 2000. - International Conference „Mechanika-2003“. Kaunas, Lithuania. April 3 – 4, 2003. - 4th International Conference „Measurement 2003“. Smolenice, Slovak Republic. June 15 – 19 d. 2003. - 14th Scientific Symposium with International Participation \"Metrology and metrology assurance 2004\". Sozopol, Bulgaria. September 14-18, 2004. - 8th International Symposium on Measurement and Quality Control in Production held on October 12-15, Erlangen, Germany, 2004. Size and structure of dissertation. The doctoral dissertation consists of introduction, 4 chapters and conclusion, list of 61 literature sources and 7 presentations according to the dissertation theme list. The size of work is 87 pages including 50 figures and 34 tables. Work contents The first chapter gives a review of the state of the art of measurement uncertainty evaluation in fire testing field. The new European fire classification system of construction products and building elements, two main test methods of this system are described. Detail analysis of existing methods of measurement uncertainty evaluation is presented. It was found out that in fire testing field quite complicated situation since real precision of most test methods is unknown. The analysis of literature shows that heat release rate obtained in SBI test is one of the most important parameter of new reaction to fire classification system and that there wasn’t real measurement uncertainty evaluation for this parameter. Survey of existing reference materials in fire test field shows that only several materials were certified as reference materials for use in certain fire test. It was found out that main part of them have been developed in USA. Also it was cleared up that only one European test method has certified reference material. The purpose and the tasks of work have been formulated after the critical and exhaustive literature analysis. The second chapter describes detail measurement uncertainty evaluation and detail analysis of HRR measurement uncertainty sources. Also detail analysis of SPR measurement uncertainty calculation is given. The evaluation results for HRR, SPR are given in 1 and 2 tables. The results in these tables clearly indicate which parameters in the SPR and HRR measurements that contribute most to the uncertainty. For the HRR, the oxygen concentration contributes most followed by the E-factor and the mass flow. The uncertainty in the oxygen measurement depends on the instrument used and the size of the fire. The E-factor is independent of the experimental apparatus but depends on the fuel used. If the fuel is known then the uncertainty decreases. The uncertainty in the velocity profile in the duct and the bi-directional probe constant are the most important for the mass flow.The uncertainty in the velocity profile can be decreased by designing the duct correctly and determining the velocity profile more precisely. For the SPR the most important factors are the calibration of the filters used for calibrating the equipment together with the temperature sensitivity of the photocell. The most interesting result is the very high uncertainty on the low SPR levels (Fig 1), a fact that should be considered when classifying products that produce little smoke.In the fourth part of the second chapter CFD models developed with two commercial codes Fluent and FDS are described. The models simulate the SBI test and they are able to describe the characteristics of the test and the influence of changed testing conditions (Fig 2 and 3). Calculations with Fluent and FDS models show significant changes in predicted temperature and velocity fields in the burning room and in the measurement section for changes in ventilation and different ambient air temperatures. All calculations were performed with non combustible specimens. Data from several laboratories for 210 to 270 s of the test (base line) has been selected from round robin tests. They show the connection of ventilation and temperatures in the measurement section as predicted. Changed temperatures in the burning room and on specimens may lead to changed ignition conditions and may influence the rate of heat release which is used for classification. Total heat release of specimen in first 600 s of exposure to the main burner flames, THR600s, has been chosen as characteristic benchmark to estimate impact of changed testing conditions on test results. Some selected round robin data seem to show the impact of ventilation on heat release of the specimen; other selected data doesn’t show the connexion that clear. The many parameters, ventilation, ambient air temperature and humidity make it difficult to see the impact of one parameter on the test results without further testing. Calculations with Fluent and FDS models show particularly impact on temperature and velocity fields in case of special combinations of the parameters ventilation and ambient air temperature. Two extreme cases, combination of high ventilation with low ambient air temperatures in contrast to low ventilation with high ambient air temperatures have been investigated. Results give reason to think about narrowing the ranges for ventilation and ambient air temperatures in the standard. In the fifth part of the second chapter performed experimental tests to investigate the influence of varying this extraction rate between 200 l/s and 1000l/s are present