Utjecaj kemijskog sastava i parametara deformacije na mehanička svojstva cijevi

master's thesis

• Title (english): Influence of composition and parameters of deformation on mechanical propertis of pipes
• Author: Darko Ević
• Mentor: Stoja Rešković (mentor)
• Committee member: Ladislav Lazić (predsjednik povjerenstva)
• Committee member: Stoja Rešković (član povjerenstva)
• Committee member: Irena Žmak (član povjerenstva); MBZ: 228900
• Granter: University of Zagreb; Faculty of Metallurgy; Sisak
• Defense date and country: 2017-01-19, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Metallurgy; Mechanical Metallurgy

Abstract

U radu su prikazani rezultati primjeneumjetnih neuronskih mreža u određivanju utjecaja kemijskog sastava čelika na mehanička svojstva toplovaljanetrake i toplovaljanih šavnih cijevi. Proveden je trening neuronske mreže na rezultatima ispitivanja kemijskih i mehaničkih svojstava od 159 uzoraka. Za zadani je problem određena najpovoljnija građa neuronske mreže. Dokazano je da je korištenjem većeg broja eksperimentalnih mjerenja, moguće pouzdano istrenirati mrežu i naučiti je da uspješno procijeni mehanička svojstva trake i cijevi. Utvrđen je utjecaj ugljika i mangana na mehanička svojstva toplovaljane trake. Posebno je istražen utjecaj kemijskog sastava čelika, mehaničkih svojstava toplovaljane trake, početne temperaturae i stupnja deformacije toplog valjanja šavnih cijevi na njihova mehanička svojstva. Utvrđeno je da ugljik ima veći utjecaj na mehanička svojstava od mangana. Provedena ispitivanja su pokazala da početna temperatura valjanja cijevi ima značajan utjecaj na njihova mehanička svojstva.

Abstract (english)

The paper presents the results of the application of artificial neural networks to determine the influence of the chemical composition of steel on the mechanical properties of hot rolled strip and hot rolled welded pipes. The training of the neural network was conducted on the test results of chemical and mechanical properties of 159 samples. Best structure of neural network for given problem was defined. It has been proven that with use of large number of experimental measurments it is possible to train and learn network to predict mechanical properties of hot-rolled strip and hot rolled welded pipes. Influence of carbon and manganese on mechanical properties has been determined. Especially influence of the chemical composition of steel, hot-rolled strip mechanical properties, initial temperature detected and the degree of deformation of the hot rolling of welded pipes to their mechanical properties was inspected. It was found that the carbon has a greater influence on the mechanical properties than manganese. The conducted tests have shown that the initial temperature of the rolling of pipe has a significant impact on their mechanical properties.

• Keywords: neuronske mreže; čelik; kemijski sastav; traka; toplovaljane šavne cijevi; mehanička svojstva
• Keywords (english): neural networks; steel; chemical composition; strip; hot rolled welded pipes; mechanical properties
• Language: croatian
• URN:NBN: urn:nbn:hr:115:865440
• Study programme: Title: Graduate study program Metallurgy; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka metalurgije (magistar/magistra inženjer/inženjerka metalurgije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2017-01-24 11:39:19