20.500.12556/DKUM-81463 Sistem za podporo postopka tuširanja orodij za globoki vlek doktorska disertacija SUPPORT SYSTEM FOR SPOTTING OF DEEP DRAWING TOOLS Doktorska disertacija se ukvarja z digitalizacijo postopka tuširanja orodij za globoki vlek. Tuširanje orodja in analiza tuširne slike sta integralni del postopka izdelave orodij, ki služita za določanje prileganja orodja in preoblikovanca. Rezultat analize tuširne slike je pokazatelj tehnološke kakovosti orodja. Zaradi odsotnosti znanstvene obravnave postopka tuširanja doktorska disertacija opisuje postopek in ga umesti v proces izdelave orodij. V nadaljevanju je zaradi digitalizacije postopka tuširanja predstavljen sistem za analizo tuširne slike, ki je sestavljen iz 3D digitalno optičnega zajemanja barvnih in geometrijskih informacij tuširanih preoblikovancev, predhodne obdelave zajetih podatkov z decimacijo, segmentacijo po metodi k-voditeljev, določanja stopnje svetlosti, mreženja (2D in 3D) in prikaza področij z informacijo o kakovosti naleganja na CAD modelu. Ob tem je predstavljena tudi integracija analize tuširne slike z rezultati simulacije preoblikovanja za ocenjevanje naleganja orodij. Razviti so bili kazalniki, za prikaz stopnje pokritosti oblakov točk in stopnje homogenosti. Za delovanje sistema je treba s predobdelavo zajete 3D tuširne slike ločiti barvne točke spodnje in zgornje površine preoblikovanca. Učinkovitost segmentacije je empirično validirana na podlagi Davies-Bouldinovega indeksa. Rezultati sistema so tudi primerjani s strokovnjakovo sposobnostjo analize tuširne slike za določitev naleganja orodja. Presented PhD thesis tackles the digitalisation of the spotting process of deep-drawing tools. Tool spotting and spotting image analysis represent an integral part of the overall tool making process, and serve to assess the degree of fit between tool and sheet metal part. The result of the spotting image analysis is an indicator of the technological quality of the tool. Due to the lack of scientific description of spotting process, the PhD thesis also provides an extensive description of the spotting process and places it into the overall toolmaking process. In the following, due to the digitalization aim, a spotting image analysis system is presented, which consists of 3D digital optical capture of the spotted surface (colour and geometric information), pre-processing in terms of decimation, segmentation by k-means, brightness level determination, meshing (2D and 3D) and display of areas with information on the quality of the fit on the CAD model. The integration of the forming simulation results for assessing contact areas is also presented. Various indicators are given, such as the degree of coverage of each point cloud and the degree of homogeneity. For appropriate deployment of the presented system, the coloured point clouds need to be separated into the lower and upper surfaces by pre-processing. The segmentation performance is empirically validated based on Davies-Bouldin index. The presented system’s results are also compared with the expert’s ability to analyse the spotting image. izdelava preoblikovalnih orodij tuširanje 3D skeniranje tuširna slika analiza slik segmentacija forming tools manufacture spotting 3D scanning image analysis segmentation true true false [L. Berus] Slovenski jezik Angleški jezik Doktorsko delo/naloga 2022-03-30 10:48:36 2022-08-22 10:08:35 2022-08-27 03:08:33 0000-00-00 00:00:00 2022 0 Maribor 0 Maribor X, 92 str. 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 2025-07-11 621.98.04:004.9(043.3) 119314179 DOK_Berus_Lucijano_2022.pdf DOK_Berus_Lucijano_2022.pdf 1 FCA81082652B0DB5A67B4F2AFAE8370B 606332143f68e5b0a68d8c6810b9e1ef4918126cada1af3a555c5798e471ba8d aaa8d1c2-0bd4-11ed-befd-00155d000105 https://dk.um.si/Dokument.php?lang=slv&id=159542 Fakulteta za strojništvo 0 0 0