2 results listed
In this study, the welding images, which were taken in HURSAN Pres company, were classified as good welding /
bad welding. The aim is to minimize human errors by artificial intelligence methods during welding defact detection. Welding
images were taken on weldings, which were done by three different experts, with 13MP mobile phone camera. The images were
labelled as good welding / bad welding by an expert. The dataset consists of 27 good welding images and 70 bad welding
images.
International Iron & Steel Symposium
UDCS
Yavuz Selim Ceran
Musa Yılmaz
Mücahit Barstugan
Fused deposition modelling (FDM) is rapidly growing 3D-printing technology due to its ability of
building functional parts having diverse and complex geometries. However, printing materials are
limited to polylactic acid (PLA) or acrylonitrile butadiene styrene (ABS) in most FDM equipment.
Here, information about polycarbonate (PC) material, which is a high-performance printing material to
overcome these shortcomings, is provided. The mechanical properties of a built part depend on several
process parameters like layer thickness, air gap, build orientations, raster angle, raster width, fill
pattern and model building temperature. The aim of this study is to determine the effect of layer
thickness on the mechanical performance of PC samples manufactured with 3D printer. Samples with
three different layer thickness (178, 254 and 330 µm) were built using PC polymer material in the 3D-
printing system and their tensile and hardness tests were carried out. Results show that the tensile
strength (53.15 MPa) and hardness (57.9 HV) are highest in the samples with a layer thickness of 254
µm and 330 µm, respectively.
International Congress on 3D Printing (Additive Manufacturing) Technologies and Digital Industry
3D-PTC2019
Musa Yılmaz
Necip Fazıl Yılmaz
Ali Kılıç
Ömer Eyercioğlu