Jämförande provning – ballast 2015: flisighetsindex, korndensitet och kulkvarn
2017 (Swedish)Report (Other academic)Alternative title
Profiency testing - aggregate 2015 (English)
Abstract [sv]
Detta notat redovisar resultat, viss statistisk analys och precisionsdata från jämförande provning (ringanalyser) utförda på tre ballastmaterial under 2015. De analyser som utfördes var flisighetsindex (SS-EN 933-3), korndensitet (SS-EN 1097-6) och kulkvarn (SS-EN 1097-9). Antalet deltagande laboratorier var för flisighetsindex 58 stycken., för korndensitet 60 stycken. och för kulkvarn 52 stycken. De flesta laboratorierna var svenska men sju stycken från övriga Norden (Island, Norge och Finland).
Spridningarna (i form av standardavvikelse) ökar med ökad nivå av resultat men inte lika mycket. Variationskoefficienten (standardavvikelse dividerat med medelvärdet per material) minskar med ökad nivå.
För precisionsdata matchar denna ringanalys i princip de precisionsdata som anges i standarderna för respektive analys. Data från denna ringanalys har ofta ett vidare spann på mätvärden än standarderna och man kan också se ett samband mellan nivå och reproducerbarhet. Repeterbarheten har inte riktigt kunna beräknas då laboratorierna generellt gjort enkelprov.
Analys av bakgrundsvariabler som visst handhavande, dimensioner eller utrustningsmodell visar att de har ingen eller endast liten betydelse.
Abstract [en]
This report shows results, some statistical analysis and precision data for profiency testing (round-robin) performed on three materials of aggregate during 2015. The properties that was determined was Flakiness Index (EN 933-3), particle density (EN 1097-6) and Nordic ball mill (EN 1097-9). The number of participating laboratories were for Flakiness index 58 pcs., for particle density 60 pcs. and for Nordic ball mill 52 pcs.
There are some scatter of the results which is not surprising. A few of the laboratories show clearly high or low results (for all three materials) for one of the analysis. Generally is the scatter in results varying. E.g. some of the laboratories have a clearly high result for one of the materials, a slightly low result for the second material and are quite close to the average for the third material.
The standard deviations increase with increase in level of results (average), but not as much. The ratio of standard deviation and average decrease with increase of level.
The precision data from this round-robin over all matches the levels from the test standards. The tests from this report has often a wider range of results. There is a correlation between reproducibility and level of average. The repeatability has not been calculated because tests actually have been performed as single tests.
Analysis of background data, like handling, dimensions and type of model, shows that they have none or very little significance.
Place, publisher, year, edition, pages
Linköping: Statens väg- och transportforskningsinstitut, 2017. , p. 45
Series
VTI notat ; 7-2017
Keywords [en]
Test method, Accuracy, Laboratory (not an organization), Group analysis (test), Aggregate, Properties
National Category
Infrastructure Engineering
Research subject
50 Road: Materials, 56 Road: Aggregate and stone materials
Identifiers
URN: urn:nbn:se:vti:diva-11726OAI: oai:DiVA.org:vti-11726DiVA, id: diva2:1085178
2017-03-282017-03-282022-10-21Bibliographically approved