京都大学 / 新制・課程博士 / 博士(工学) / 甲第24580号 / 工博第5086号 / 新制||工||1974(附属図書館) / 京都大学大学院工学研究科航空宇宙工学専攻 / (主査)教授 江利口 浩二, 教授 土屋 智由, 教授 平方 寛之 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

Software testing is essential for quality assurance, with automated techniques such as random testing and adaptive random testing being cost-effective solutions compared to others. Adaptive random testing seeks to enhance random testing, and there is a conception that adaptive random testing always should replace random testing. Our research question investigates this conception by addressing a gap in the literature, where a comparison between the two techniques in terms of certain key metrics is missing, namely defect detection efficiency and test case generation time. Defect detection efficiency is the amount of defects detected divided by the number defects in the system multiplied by one hundred. Test case generation time is the time it takes to generate all of the test case inputs. These metrics where chosen as they can be seen as a measurement of the techniques effectiveness and efficiency respectively. In order to address this research question we employ a quantitative experiment where we compare the performance of random testing and adaptive random testing with a sole focus on these two metrics. The comparison is performed by implementing and testing both algorithms on eight error-seeded numerical programs and measuring the results. The results displayed that adaptive random testing had a defect detection efficiency total average of 21.59% and a test case generation time total average of 35.37 (ms), while random testing had a defect detection efficiency total average of 22.28% and a test case generation time total average of 0.26 (ms). These results might contribute to disproving the conception that adaptive random testing always should replace random testing, as random testing evidently performed better on both the measured metrics.

Indiana University-Purdue University Indianapolis (IUPUI) / Congenital heart defects occur in approximately one percent of births every year, which makes it the most frequently occurring congenital defect in patients. The aim of this project was to use two mutant neural crest (NC) mouse models to study the mechanisms underlying congenital heart failure in utero. The first mouse model was a Pax3 systemic knockout, which was lethal by mouse gestational day 14, and had appreciably reduced numbers of migratory NC cells. The second mouse model was a Wnt1Cre-mediated NC genetic cell ablation model, which was surprisingly viable and survived to birth, despite an apparent lack of migratory NC cells. The resultant data indicated that both mouse models had similar heart structural defects including persistent truncus arteriosus, which was due to fewer or no migratory cardiac NC cells. However, in utero heart function was appreciably perturbed in Pax3 mutants when compared to that of the ablated mutant model. The loss of embryonic cardiac function in Pax3 mutants was directly attributed to a substantial decrease in the activity of the beta-adrenergic pathway. This was due to a lack of proper specification of trunk NC cells, leading to diminished levels of circulating catecholamine levels in the embryo. To definitively confirm this conclusion, poor cardiac function was successfully restored by pharmacological stimulation of the beta-adrenergic pathway via administration of isoproterenol and forskolin to pregnant dams, which led to embryonic survival of Pax3 mutants to birth. By comparison of these two mutant mouse models, perturbation in the beta-adrenergic pathway was identified as the underlying mechanism responsible for in utero heart failure and lethality in Pax3 mutant embryos. The results of this study are expected to be significant in developing future therapeutic targets for congenital heart failure in prenatal and newborn patients.

No description available.

The subject of this thesis is a qualitative analysis of the usability of three contemporary programming languages: Perl, Python, and Tcl. The relative usability of these languages was assessed by implementing two representative applications — a paint program and a chat program — in all three languages; using PSP to collect data on program development; and then comparing development times, defect counts, and defect fix times. Overall, Python was the easiest of the three languages to use, especially for new users. Perl is the hardest, and Tcl is somewhere between. Specifically, Perl's basic language constructs are the hardest to learn and to use. Python's are the easiest. Tcl's are closer to Python's. Python's and Tcl's mechanisms for supporting OO are much easier to learn for new users than Perl's. Python's Tk libraries are easier to use than Perl's and Tcl's. Tcl's [Incr Widget] (megawidgets) library is by far the hardest to master. Perl's Tk library is somewhere in between. Perl's I/O and Socket libraries are much harder to use than Python's and Tcl's. The primary recommendations of this study are as follows: For Perl, a better OO support will reduce a significant number of OO-related defects, especially to new users. For Tcl, Tcl's [Incr Tcl] megawidget library should wrap the standard raw Tk widgets into [Incr Tk] classes, to prevent a significant number of defects from mixture uses of raw Tk widgets and [incr Tk] megawidgets classes. Tcl's eventloop mechanism is a good model for handling non-blocking sockets and I/Os. Both Perl and Python should also provide such an easy model.

X-ray diffraction (XRD) patterns, electron probe microanalysis(EPMA), electron paramagnetic resonance (EPR) powder spectra (9 and 34 GHz) and the magnetic susceptibility of BaTiO3 + 0.04 BaO + 0.01 V2O5 ceramics were studied to investigate the valence states of V ions and their solubility in the BaTiO3 lattice. In samples sintered at 1400 °C in air, only about 0.1 mol% V is incorporated in the BaTiO3 lattice being in V4+ and V5+ valence state, respectively. 95% of the nominal V dopant content occurs in the secondary phase Ba3(V/Ti)2O8. All BaTiO3 samples investigated are in tetragonal phase at room temperature. In the as-sintered samples V4+ is detected at temperatures T < 20 K by its hyperfine structure (HFS) octet due to the nuclear spin 7/2 of 51V. Samples post annealed in H2/Ar atmosphere at 1200 °C exhibit a further HFS octet occurring at T > 25 K and vanishing at T > 250 K, which is caused by V2+ ions. This spectrum is characterized by a simultaneous HFS and fine structure splitting constituted by allowed and forbidden transitions. Both V4+ and V2+ ions are incorporated at Ti4+ sites of the BaTiO3 lattice

In this thesis the application of software defect prediction to predict unit test failure is investigated. Data for this purpose was collected from a Continuous Integration development environment. Experiments were performed using semantic features from the source code. As the data was imbalanced with defective samples being in minority different degrees of oversampling were also evaluated. The data collection process revealed that even though several different code commits were available few ever failed a unit test. Difficulties with linking a failure to a specific file were also encountered. The machine learning model used in the project produced poor results when compared against related work, from which it was based on. In F-measure, it on average achieve 53% of the mean performance of state-of-the-art for software defect prediction on bugs in Java source files. Specifically, it would appear that very little information was available for the model to learn defects in files not present in training data. / I denna avhandling undersöks applikationen av prognos för mjukvarudefekter för att förutse enhetstestfel. Data för detta syfte samlades in från en utvecklingsmiljö med kontinuerlig integration. Experimenten utfördes med användning av semantiska särdrag samlade från källkod. Då data var obalanserat med defekta exempel i minoritet evaluerades olika grader av översampling. Datainsamlingsprocessen visade att även om det fanns många kodinlämningar så misslyckades få någonsin ett enhetstest. Svårigheter med att länka testmisslyckanden till en specifik fil påträffades också. Den använda maskininlärningsmodellen uppvisade också dåliga resultat i jämförelse med relaterade värk. Mätt i F-measure uppnåddes i genomsnitt 53% av genomsnittlig prestandan av bästa möjliga prognos av mjukvarudefekter av buggar i Java källkod. Specifikt så framträdde det att väldigt lite information verkar finnas för modellen att lära sig defekter i filer som ej fanns med i träningsdata.

Laser powder bed fusion (LPBF) remains a predominately open-loop additive manufacturing process with minimal in-situ quality and process control. Some machines feature optical monitoring systems but lack automated analytical capabilities for real-time defect detection. Recent advances in machine learning (ML) and convolutional neural networks (CNN) present compelling solutions to analyze images in real-time and to develop in-situ monitoring. Approximately 30,000 selective laser melting (SLM) build images from 31 previous builds are gathered and labeled as either “okay” or “defect”. Then, 14 open-sourced CNN were trained using transfer learning to classify the SLM build images. These models were evaluated by F1 score and down selected to the top 3 models. The top 3 models were then retrained and evaluated using Dietterich’s 5x2 cross-validation and compared with pairwise student t-tests. The pairwise t-test results show no statistically significant difference in performance between VGG- 19, Xception, and InceptionResNet. All models are strong candidates for future development and refinement. Additional work addresses the entire model development process and establishes a foundation for future work. Collaborations with computer science students has produced an image pre-processing program to enhance as-taken SLM images. Other outcomes include initial work to overlay CAD layer images and preliminary hardware integration plan for the SLM machine. The results from this work have demonstrated the potential of an optical layer-wise image defect detection system when paired with a CNN.

Granular superconductivity at high temperatures in graphite can emerge at certain two-dimensional (2D) stacking faults (SFs) between regions with twisted (around the c-axis) or untwisted crystalline regions with Bernal (ABA…) and/or rhombohedral (ABCABCA…) stacking order. One way to observe experimentally such 2D superconductivity is to measure the frozen magnetic flux produced by a permanent current loop that remains after removing an external magnetic field applied normal to the SFs. Magnetic force microscopy was used to localize and characterize such a permanent current path found in one natural graphite sample out of ∼50 measured graphite samples of different origins. The position of the current path drifts with time and roughly follows a logarithmic time dependence similar to the one for flux creep in type II superconductors. We demonstrate that a ≃10 nm deep scratch on the sample surface at the position of the current path causes a change in its location. A further scratch was enough to irreversibly destroy the remanent state of the sample at room temperature. Our studies clarify some of the reasons for the difficulties of finding a trapped flux in a remanent state at room temperature in graphite samples with SFs.

The low-temperature joining technique (LTJT) by silver sintering is being implemented by major manufacturers of power electronics devices and modules for bonding power semiconductor chips. A common die-attach material used with LTJT is a silver paste consisting of silver powder (micron- or nano-size particles) mixed in organic solvent and binder formulation. It is believed that the drying of the paste during the bonding process plays a critical role in determining the quality of the sintered bond-line. In this study, a model based on the diffusion of solvent molecules and viscous mechanics of the paste was introduced to determine the stress and strain states of the silver bond-line. A numerical simulation algorithm of the model was developed and coded in the C++ programming language. The numerical simulation allows determination of the time-dependent physical properties of the silver bond-line as the paste is being dried with a heating profile. The properties studied were solvent concentration, weight loss, shrinkage, stress, and strain. The stress is the cause of cracks in the bond-line and bond-line delamination. The simulated results were verified by complementary experiments in which the formation of cracks in bond-line and interface delamination was observed during the pressure-free drying of a die-attach nanosilver paste. Furthermore, the important drying parameters, such as drying pressure, low temperature drying time and temperature ramp rate of nanosilver LTJT process, are experimentally studied and analyzed with the numerical simulation. The simulated results were consistent with the experimental findings that the quality of sintered silver bond-line increases with increasing external drying pressure, with increasing low temperature drying time, and with decreasing temperature ramp rate. The insight offered by this modeling study can be used to optimize the process profile that enable pressure-free, low-temperature sintering of the die-attach material to significantly lower the cost of implementing the LTJT in manufacturing. / Ph. D.