Fraunhofer Chalmers Research Centre for Industrial Mathematics
Chalmers Publication LibraryNot a member yet
70439 research outputs found
Sort by
Algorithms for Pure Categorical Optimization
No full textOptimeringsproblem med kategoriska variabler är vanligt förekommande exempelvis inom bilindustrin och andra industrier där mekaniska komponenter ska väljas ut och kombineras på gynnsamma sätt. Avsaknaden av naturlig ordning på beslutsvariablerna gör att kategoriska optimeringsproblem oftast är svårare att lösa än diskreta eller kontinuerliga problem. Det är därför viktigt att ta fram metoder som löser kategoriska optimeringsproblem. Den här rapporten presenterar tre olika algoritmer som kan användas för att lösa kategoriska optimeringsproblem: en lokalsökningsalgoritm, en globalsökningsalgoritm, och en genetisk algoritm. Dessutom presenteras två olika omgivningsdefintioner att använda ihop med lokalsökningsalgoritmen, en diskret, och en kategorisk. Algoritmerna implementerades i Matlab och testades på två olika kategoriska optimeringsproblem: ett artificiellt problem, och ett balkproblem. De framtagna algoritmerna applicerades på ett stort antal instanser av testproblemen och deras prestanda utvärderades med hjälp av prestandaprofiler och dataprofiler. Lokalsökningsalgoritmen utrustad med den kategoriska omgivningen presterade bäst av de testade algoritmerna
Experimental characterization of pearlitic rail steel after thermomechanical straining
No full textRails are subjected to very high contact loads during service. The high contact loads cause the surface layer of the rails to be heavily deformed and aligned. The
anisotropic nature of the deformed surface layer is prone to crack initiation. The deformed surface layer is also very thin and has a large gradient of accumulated
strain. This large gradient makes it difficult to examine the material behavior with conventional testing methods because they requires a fairly uniform microstructure.
A predeformation method developed by CHARMEC researchers have proven to be able to produce a material with a fairly uniform microstructure which is consistent
with rail field samples with high accumulated shear strain.
The aim with the Master Thesis was to expand the knowledge of the material behaviour of pearlitic rail steels (grade R260) under combined thermal and cyclic
mechanical loading. The goal was to produce a microstructure with higher accumulated
strains compared to previous work. It was achieved by adding a heat treatment to the predeformation method. An axial-torsion test rig with an induction coil has
been used to deform and heat treat solid cylindrical test bars. This was done to obtain a microstructure that was similar to the one found in the field. The material
was compared with field samples in terms of microstructure and hardness.
The results of this thesis describes the mechanical behavior of a pearlitic rail steel during simultaneous axial compression and torsion with different compression loads at elevated temperature. The microstructures have been characterized and accumulated
strain and hardness have been measured. The highest amount of accumulated strain was obtained with constant heating at 350 °C with an axial compression of
350 MPa and twist rate of 1.5 °/s. The amount of twisting was 3.5 times higher compared to previous work. Heating in between the twisting cycles resulted in the
least amount of accumulated strain
