Likenesses and differences in the fragmentation of full-scale and model-scale blasts
2006 (English)In: Proceedings of the 8th International Conference on Rock Fragmentation by Blasting, Santiago: Editec , 2006, 207-220 p.Conference paper (Refereed)
Fragmentation results for model- and full-scale blasts in three different rock materials are presented. The model-scale blasts were conducted in Ø100-300 mm cylindrical specimens charged with PETN. The full-scale Bårarp and Hengl bench rounds consisted of a single row of 5-10 holes, the Vändle rounds of multiple rows. The Bårarp rock mass consists of massive, damage free granitic gneiss whereas the other rock masses were both jointed and blast damaged. In all cases production explosives were used.There are three major likenesses in the fragmentation. Firstly all the sieving curves can be accurately described by the Swebrec function. Secondly, in the fines range the curves shift upward in parallel in a log-log diagram when the specific charge increases, in accordance with one of the NBC precepts. Thirdly, the sets of sieving curves for one rock all seem to have an inflection point at the same x-value in the log-log diagram. This point varies; from about 0.5 mm for the Hengl amphibolite, to 1.0 mm for the Bårarp gneiss, and 1.5 mm for the Vändle granite. Taken together these likenesses strongly suggest that the breakage mechanisms in model- and full-scale have strong similarities, despite the very different conditions.The model- and full-scale fragmentation also differs in important aspects. Firstly the definition of specific charge q is based on different volumes related to the actual breakage and the presence of free faces, which are shown to influence the fragmentation. Further the exponent for how the average fragment size x50 decays with increasing q is larger in model-scale, about 1.1 (x50∝1/q1.1) as opposed to 0.8 in full-scale. Further, the influence of size, given by the charge size Q1/6 in the Kuz-Ram model doesn't fully explain the observed x50 differences when q is constant. The rock factor A also seems to have a size dependence built in, which leads to different results for the Vändle and Bårarp tests. There is further a difference in the slope values s50 at x50 between model- and full-scale blasting. The values follow a previously established experimental relationship quite well. Thus the undulation exponent in the Swebrec function may be expressed as a function of the other two parameters, x50 and xmax. In order to revise the size dependence expressed by the Kuz-Ram equation, which in the case of x50 apparently involves not only Q, but also A and the exponent of the specific charge, it will probably be necessary to broaden the specific charge concept.
Place, publisher, year, edition, pages
Santiago: Editec , 2006. 207-220 p.
Civil engineering and architecture - Geoengineering and mining engineering
Samhällsbyggnadsteknik och arkitektur - Geoteknik och gruvteknik
Research subject Mining and Rock Engineering
IdentifiersURN: urn:nbn:se:ltu:diva-27750Local ID: 149cbc00-8aa3-11db-8975-000ea68e967bOAI: oai:DiVA.org:ltu-27750DiVA: diva2:1000939
International Symposium on Rock Fragmentation by Blasting : 07/05/2006 - 11/05/2006
Godkänd; 2006; 20061208 (ysko)2016-09-302016-09-30Bibliographically approved