prikaz prve stranice dokumenta Correlation between MIG Brazing Parameters of a Hybrid Arc and Brazing Quality of Hot-Dip Galvanized Steels
No public access
doctoral thesis
Correlation between MIG Brazing Parameters of a Hybrid Arc and Brazing Quality of Hot-Dip Galvanized Steels
2024. urn:nbn:hr:262:718152
Strutz, Josip Florian
University of Slavonski Brod

Cite this document

Strutz, J. F. (2024). Correlation between MIG Brazing Parameters of a Hybrid Arc and Brazing Quality of Hot-Dip Galvanized Steels (Doctoral thesis). Slavonski Brod: University of Slavonski Brod. Retrieved from https://urn.nsk.hr/urn:nbn:hr:262:718152

Strutz, Josip Florian. "Correlation between MIG Brazing Parameters of a Hybrid Arc and Brazing Quality of Hot-Dip Galvanized Steels." Doctoral thesis, University of Slavonski Brod, 2024. https://urn.nsk.hr/urn:nbn:hr:262:718152

Strutz, Josip Florian. "Correlation between MIG Brazing Parameters of a Hybrid Arc and Brazing Quality of Hot-Dip Galvanized Steels." Doctoral thesis, University of Slavonski Brod, 2024. https://urn.nsk.hr/urn:nbn:hr:262:718152

Strutz, J. F. (2024). 'Correlation between MIG Brazing Parameters of a Hybrid Arc and Brazing Quality of Hot-Dip Galvanized Steels', Doctoral thesis, University of Slavonski Brod, accessed 11 December 2024, https://urn.nsk.hr/urn:nbn:hr:262:718152

Strutz JF. Correlation between MIG Brazing Parameters of a Hybrid Arc and Brazing Quality of Hot-Dip Galvanized Steels [Doctoral thesis]. Slavonski Brod: University of Slavonski Brod; 2024 [cited 2024 December 11] Available at: https://urn.nsk.hr/urn:nbn:hr:262:718152

J. F. Strutz, "Correlation between MIG Brazing Parameters of a Hybrid Arc and Brazing Quality of Hot-Dip Galvanized Steels", Doctoral thesis, University of Slavonski Brod, Slavonski Brod, 2024. Available at: https://urn.nsk.hr/urn:nbn:hr:262:718152

Please login to the repository to save this object to your list.
Metadata
TitleCorrelation between MIG Brazing Parameters of a Hybrid Arc and Brazing Quality of Hot-Dip Galvanized Steels
Title (croatian)Korelacija između parametara MIG tvrdoga lemljenja pomoću hibridnoga luka i kvalitete lemljenja vruće pocinčanih čelika
AuthorJosip Florian Strutz
MentorKatica Šimunović (mentor)
MBZ: 211943
MentorIvan Samardžić (komentor)
MBZ: 165270
Committee memberAntun Stoić (predsjednik povjerenstva)
MBZ: 190286
Committee memberLeon Maglić (član povjerenstva)
MBZ: 190275
GranterUniversity of Slavonski Brod
Slavonski Brod
Defense date and country2024-11-14, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Mechanical Engineering
Universal decimal classification
(UDC)		621 - Mechanical engineering. Nuclear technology. Machinery
Abstract
Od postupaka MIG tvrdoga lemljenja vruće pocinčanih čeličnih limova srednje debljine, uglavnom se primjenjuju postupci s kratko spojenim i pulsnim električnim lukom te CMT (od Cold Metal Transfer) postupak. U ovome radu, noviji postupak, s hibridnim lukom, koji predstavlja kombinaciju kratko spojenoga i pulsnoga luka, uspoređen je s prethodno navedenima. Nakon toga je pomoću algoritama strojnoga učenja dobiven metamodel za predviđanje kvalitete zalemljenoga spoja. Ova četiri postupka su uspoređena na temelju mjerenja signala struje i napona lemljenja, ispitivanja mehaničkih svojstava (mjerenje tvrdoće i statički vlačni test) i korozijske otpornosti te makroskopskog ispitivanja zalemljenoga spoja. Pri tome je, za učinkovitije istraživanje, razvijen poluautomatski algoritam toka istraživanja, u kojem je, između ostaloga, programirana metoda obrade slike za makroskopsko ispitivanje. Rezultati istraživanja i višekriterijske usporedbe postupaka su ukazali na poboljšanje svojstava zalemljenoga spoja dobivenog s hibridnim lukom. Makroskopsko ispitivanje i ispitivanje korozijske otpornosti također su pokazali bolje rezultate u usporedbi s ostalim postupcima. Zatim je razvijen metamodel za a-dimenziju i vlačnu silu zalemljenoga spoja, primjenom različitih algoritama strojnoga učenja, uključujući umjetne neuronske mreže (ANN). Za optimalno parametriziranje metamodela primijenjen je zatim genetički algoritam. Za daljnju optimizaciju, razvijena je analiza osjetljivosti koja je uzela u obzir eksperimentalne vrijednosti. Konačno, razvijeni su ansambli neuronskih mreža s različitim težinama pojedinačnih mreža, pomoću genetičkoga algoritma i analize osjetljivosti, čime je dobiveno odstupanje od 0,1 mm za a-dimenziju i 1,35 kN za vlačnu silu.
Abstract (english)
In the field of MIG brazing of hot-dip galvanized, medium-thickness sheet metals, the short-circuiting metal arc (SCMA), pulsed arc and cold metal transfer (CMT) brazing processes are mainly used to join components. In this work, a hybrid arc consisting of short-circuiting metal arc and pulsed arc is compared with these processes. Afterwards, a metamodel for predicting setting parameters and quality features was created using machine learning algorithms. The current and voltage responses, the mechanical characteristics (hardness testing and quasi static tensile test), the macroexamination properties and also the corrosion resistance were evaluated and compared. A semi-automated workflow was developed for the evaluation, in which, among other things, an image processing method was programmed for the macroexamination. The results of the evaluation and multicriteria comparison proved a complete improvement of the properties with the hybrid arc. The macroexamination and corrosion tests were also convincing. Then the metamodel for the a-dimension and tensile force was derived using various machine learning algorithms - including artificial neural networks (ANN). A genetic algorithm was used to optimally parameterize the metamodel. For further optimization of ANN, a sensitivity analysis was developed that takes the experimental values into account. Finally, an ANN ensemble with different weighting of the individual neural networks was developed using the genetic algorithm and optimized with the sensitivity analysis. This achieves a deviation of 0.1 mm for the a-dimension and 1.35 kN for the tensile force.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:262:718152
Promotion2024
Study programmeTitle: Obtaining a doctorate of science outside of doctoral studies
Study programme type: university
Study level: postgraduate
Academic / professional title: doktor/doktorica znanosti (doktor/doktorica znanosti)
Type of resourceText
File originBorn digital
Access conditionsClosed access
Terms of useLicence In copyright icon
Created on2024-11-28 08:45:55