No description available.

No description available.

Quantitative Structure Activity Relationship (QSAR) is a very useful computa-tional method which has facilitated great progress in drug development [1]. Thismethod can be used to predict a molecule’s activity against a certain target justby comparing its structural characteristics (i.e., molecular descriptors) with thosebelonging to molecules of known activity. QSAR modeling is fueled by online freedatabases consisting of millions of active and inactive molecules and by MachineLearning (ML) Methods that enable data analysis. To ensure successful implemen-tation of ML models, there is a range of evaluation methods to estimate their perfor-mance and applicability domain. So far, a great deal of research has focused on theuse of Support Vector Machines (SVMs) to classify molecules with the use of theirMolecular Signature Fingerprints as descriptors [2]. However, another MachineLearning algorithm, Deep Neural Networks (DNNs), an improvement of single-layer Neural Networks, is rising in popularity in various fields including moleculeclassification. The two models were compared using CPSign software which intro-duces Conformal Prediction, to evaluate the reliability of model predictions basedon performance for individual compounds rather than mean performance on agiven test set. Three types of descriptors were used: Molecular Signature Finger-prints, Extended Connectivity Fingerprints and physicochemical descriptors. Thecomparison showed that Multilayer Perceptron (MLP) which was used as a DNNrepresentative in current context, had performance similar to the shallower SVMmodels but additionally demanded longer training times [3]. It can be concludedthat in the field of QSAR with the aforementioned descriptors, when the numberof examples used for training is not immense, Support Vector Machines might per-form equally well and demand less resources and time than the more sophisticated MLPs.

Additive manufacturing (AM) has received much attention from researchers over the past decades because of its diverse applications in various industries. AM is an advanced manufacturing process that facilitates the fabrication of complex geometries represented by computer-aided design (CAD) models. Traditionally, designed parts are fabricated by extruding material layer-by-layer using a tool path planning obtained from slicing programs by using CAD models as an input. Recently, there has been a growing interest in non-planar AM technologies, which offer the ability to fabricate multilayer constructs conforming to freeform surfaces. Non-planar AM processes have been utilized in various applications and involved objects of varying material properties and geometric characteristics. Although the current state of the art suggests AM can provide novel opportunities in conformal manufacturing, several challenges remain to be addressed. The identified challenges in non-planar AM fall into three categories: 1) conformal 3D printing on substrates with complex topography of which CAD model representation is not readily available, 2) understanding the relationship between the tool path planning and the quality of the 3D-printed construct, and 3) conformal 3D printing in the presence of mechanical disturbances. An open-loop non-planar tool path planning algorithm based on point cloud representations of object geometry and a closed-loop non-planar tool path planning algorithm based on position sensing were proposed to address these limitations and enable conformal 3D printing and spatiotemporal 3D sensing on objects of near-arbitrary organic shape. Three complementary studies have been completed towards the goal of improving the conformal tool path planning capabilities in various applications including fabrication of conformal electronics, in situ bioprinting, and spatiotemporal biosensing: i. A non-planar tool path planning algorithm for conformal microextrusion 3D printing based on point cloud data representations of object geometry was presented. Also, new insights into the origin of common conformal 3D printing defects, including tool-surface contact, were provided. The impact and utility of the proposed conformal microextrusion 3D printing process was demonstrated by the fabrication of 3D spiral and Hilbert-curve loop antennas on various non-planar substrates, including wrinkled and folded Kapton films and origami. ii. A new method for closed-loop controlled 3D printing on moving substrates, objects, and unconstrained human anatomy via real-time object position sensing was proposed. Monitoring of the tool position via real-time sensing of nozzle-surface offset using 1D laser displacement sensors enabled conformal 3D printing on moving substrates and objects. The proposed control strategy was demonstrated by microextrusion 3D printing on oscillating substrates and in situ bioprinting on an unconstrained human hand. iii. A reverse engineering-driven collision-free path planning program for automated inspection of macroscale biological specimens, such as tissue-based products and organs, was proposed. The path planning program for impedance-based spatiotemporal biosensing was demonstrated by the characterization of meat and fruit tissues using two impedimetric sensors: a cantilever sensor and a multifunctional fiber sensor. / Doctor of Philosophy / Additive Manufacturing (AM), commonly referred to as 3D printing, is a computer-aided manufacturing process that facilitates the fabrication of personalized and customized models, tissues, devices, and wearables. AM has several advantages over traditional manufacturing processes. For example, directing computer-driven robotics enables control over spatial structure and composition of parts. While 3D printing is typically performed using layer-by-layer planar tool paths generated by slicing programs, non-planar 3D printing is an emerging area that has recently been examined for various post-processing applications. Processes that enable material deposition conforming to complex geometric and freeform objects (e.g., anatomical structures), are central to various industries, including additive manufacturing, electronics manufacturing, and biomanufacturing. In this dissertation, tool path planning methods and real-time control strategies for non-planar 3D printing onto stationary and moving arbitrary surfaces, and various conformal electronics and in situ bioprinting applications will be presented. In addition to the tool path planning methods for 3D printing, a collision-free path planning program will be proposed for the inspection of large tissues and organs. The utility of the proposed method will be demonstrated through electrical impedance-based biosensing of meat and fruit to characterize their compositional and physiochemical properties which are used for quality assessment.

Non-planar additive manufacturing (AM) technologies, such as microextrusion 3D printing processes, offer the ability to fabricate conformal electronics with impressive structure and function on curvilinear substrates. Although various available methods offer conformal 3D printing capability on objects with limited geometric complexity, a number of challenges remain to improve feature resolution, throughput, materials compatibility, resultant function and properties of printed components, and application to substrates of varying topography. Hence, the overall objective of this dissertation was to create new non-planar AM processes that are compatible with personalized and anatomical computer-aided design workflows for the fabrication of conformal electronics and form-fitting wearables. After reviewing the current state of knowledge and state of the art, significant challenges in non-planar AM have been identified as: 1) limited non-planar AM path planning capability that synergizes with personalized or anatomical object surface modification, 2) limited approaches for printed and non-printed component integration on non-planar substrates. To address these challenges, a template-based reverse engineering workflow is proposed for conformal 3D printing electronics and form-fitting wearable devices on anatomical structures. This work was organized into three complementary tasks that enhance non-planar AM capabilities: 1） To achieve anatomical tissue-sensor integration, 3D scanning-based point cloud data acquisition and customized 3D printable conductive ink are proposed for capturing the topographical information of patient-specific malformations and integrating conformal sensing electronics across anatomical tissue-device interface. 2） To fabricate conformal antennas on flexible thin-film polymer substrates, a versatile method for microextrusion 3D printing of conformal antennas on thin film-based structures of random topography is proposed to control the ink deposition process across the curvilinear surfaces of freeform Kapton-based origami. 3） To simplify the fabrication process of form-fitting wearable devices with fiber-based form factors and self-powered capability, an innovative 3D printing process is proposed to achieve coaxial multi-material extrusion of metal-elastomer triboelectric fibers. By developing new advanced non-planar printing processes and conformal toolpath programming strategies, the utility of non-planar AM could be further expanded for fabricating various personalized implantable and wearable multi-functional systems, including novel 3D electronics. In summary, this work advances capability in additive manufacturing processes by providing new advances in multi-material extrusion processes and personalized device design and manufacturing workflows. / Doctor of Philosophy / The ability to assemble electronic devices on three-dimensional objects with complex geometry is essential for developing next-generation wearable devices. Additive manufacturing processes, commonly referred to as 3D printing, now offer the ability to fabricate conformal electronics on surfaces and objects with non-planar geometry. This dissertation aims to expand non-planar 3D printing capabilities for applications to objects with anatomical or personalized structures, such as patient-specific malformation and origami. The proposed methods in this dissertation are focused on addressing challenges, such as the acquisition of object 3D topographical data, material selection, and tool path programming for objects that exhibit anatomical geometry. The utility of the proposed methods is demonstrated with practical applications to 3D-printed conformal electronics and wearable devices for monitoring human behavior and organ healthcare. This dissertation contributes to improving manufacturing capability and outcomes of 3D-printed form-fitting wearable and implantable devices. Future work may emphasize developing biocompatible functional ink and toolpath programming algorithms with real-time adaptation capability.

The overall goal of this research work is to enhance the capabilities of large-format, polymer material extrusion, additive manufacturing (AM) systems. Specifically, the aims of this research are to (1) Construct, and develop a robust workflow for, a large-format, robotic, AM system; (2) Develop an algorithm for determining and relaying proper rotation commands for 5 degree of freedom (DoF) multi-axis deposition; and (3) Create a method for printing a removable support material in large-format AM. The development and systems-integration of a large-format, pellet-fed, polymer, material extrusion (ME), AM system that leverages an industrial robotic arm is presented. The robotic arm is used instead of the conventional gantry motion stage due to its multi-axis printing ability, ease of tool changes for multi-material deposition and/or subtraction, and relatively small machine footprint. A novel workflow is presented as a method to control the robotic arm for layer-wise fabrication of parts, and several machine modifications and workflow enhancements are presented to extend the multi-axis manufacturing capabilities of the robot. This workflow utilizes existing AM slicers to simplify the motion path planning for the robotic arm, as well as allowing the workflow to not be restricted to a single robotic deposition system. To enable multi-axis deposition, a method for generating tool orientations and resulting deposition toolpaths from a geometry's STL file was developed for 5-DoF conformal printing and validated via simulation using several different multi-DOF robotic arm platforms. Furthermore, this research proposes a novel method of depositing a secondary sacrificial support material was created for large-format AM to enable the fabrication of complex geometries with overhanging features. This method employs a simple tool change to deposit a secondary, water-soluble polymer at the interfaces between the part and supporting structures. In addition, a means to separate support material into smaller sections to extend the range of geometries able to be manufactured via large-format AM is presented. The resultant method was used to manufacture a geometry that would traditionally be considered unprintable on conventional large-format AM systems. / Master of Science / Additive manufacturing (AM), also known as 3D printing, is a method of manufacturing objects in a layer-by-layer technique. Large-format AM is typically defined as an AM system that can create an object larger than 1 m3. There are only a few manufacturers in the world of these systems, and all currently are built on gantry-based motion stages that only allow movement of the printer in three principal axes (X, Y, Z). The primary goal of this thesis is to construct a large-format AM system that uses a robotic arm to enable printing in any direction or orientation. The use of an industrial robotic arm enables printing in multiple planes, which can be used to print structures without support structures, print onto curved surfaces, and to purt with curved layers which produces a smoother external part surface. The design of the large-format AM system was validated through successful printing of objects as large as 1.0x0.5x1.2 m, simultaneous printing of a sacrificial support material to enable overhanging features, and through completing multi-axis printing. To enable multi-axis printing, an algorithm was developed to determine the proper toolpath location and relative orientation to the part surface. Using a part's STL file as input, the algorithm identifies the normal vector at each movement command, which is then used to calculate the required tool orientation. The tool orientations are then assembled with the movement commands to complete the multi-axis toolpath for the robot to perform. Finally, this research presents a method of using a second printing tool to deposit a secondary, water-soluble material to act as supporting structures for overhanging and bridging part features. While typical 3D printers can generally print sacrificial material for supporting overhangs, large-format printers produce layers up to 25 mm wide, rendering any support material impossible to remove without post-process machining. This limits the range of geometries able to be printed to just those with no steep overhangs, or those where the support material is easily reachable by a tool for removal. The solution presented in this work enables the large scale AM processes to create complex geometries.

In this thesis we explore CFTs and related nonperturbative phenomena through several various techniques. First we construct a novel way to latticize AdS via the triangle group. We characterize aspects of this tessellation in 2d as well extend the construction to 3d, where we find a bulk critical point for scalar 𝜙⁴ theory. We then study 2d minimal model CFTs in a bulk AdS₂ through powerful boundary CFT and JT/Schwarzian techniques. Finally, we use an algebraic formulation of a conformal interface separating two minimal models -- a RG brane -- to compute correlators in the presence of the brane to better understand the relation between different minimal model CFT theories.

2023 / I derive a formulation of the 2-dimensional critical Ising model on non-uniform simplicial lattices. Surprisingly, the derivation leads to a set of geometric constraints that a lattice must satisfy in order for the model to have a well-defined continuum limit. I perform Monte Carlo simulations of the critical Ising model on discretizations of several non-trivial manifolds including a twisted torus and a 2-sphere and I show that the simulations are in agreement with the 2d Ising CFT in the continuum limit. I discuss the inherent benefits of using non-uniform simplicial lattices to study quantum field theory and how the methods developed here can potentially be generalized for use with other theories.

Mémoire numérisé par la Division de la gestion de documents et des archives de l'Université de Montréal.

L'objet d'étude de cette thèse est une mesure infinie sur les boucles (lacets) naturellement associée à une large classe de processus de Markov et les processus ponctuels de Poisson d'intensité proportionnelle à cette mesure (paramètre d'intensité alpha>0). Ces processus ponctuels de Poisson portent le nom d'ensembles poissoniens de boucles markoviennes ou de soupes de boucles. La mesure sur les boucles est covariante par un certain nombre de transformations sur les processus de Markov, par exemple le changement de temps.Dans le cadre de soupe de boucles brownienne à l'intérieur d'un sous-domaine ouvert propre simplement connexe de C, il a été montré que les contours extérieurs des amas extérieurs de boucles sont, pour alpha<=1/2, des Conformal Loop Ensembles CLE(kappa), kappa dans (8/3,4]. D'autre part il a été montré pour une large classe de processus de Markov symétriques que lorsque alpha=1/2, le champ d'occupation d'une soupe de boucle (somme des temps passés par les boucles aux dessus des points) est le carré du champ libre gaussien. J'ai étudié d'abord les soupes de boucles associés aux processus de diffusion unidimensionnels, notamment leur champ d'occupation dont les zéros délimitent dans ce cas les amas de boucles. Puis j'ai étudié les soupes de boucles sur graphe discret ainsi que sur graphe métrique (arêtes remplacés par des fils continus). Sur graphe métrique on a d'une part une géométrie non triviale pour les boucles et d'autre part on a comme dans le cas unidimensionnel continu la propriété que les zéros du champ d'occupation délimitent les amas des boucles. En combinant les graphes métriques et l'isomorphisme avec le champ libre gaussien j'ai montré que alpha=1/2 est le paramètre d'intensité critique pour la percolation par soupe de boucles de marche aléatoire sur le demi plan discret Z*N (existence ou non d'un amas infini) et que pour alpha<=1/2 la limite d'échelle des contours extérieurs des amas extérieurs sur Z*N est un CLE(kappa) dans le demi-plan continu. / In this thesis I study an infinite measure on loops naturally associated to a wide range of Markovian processes and the Poisson point processes of intensity proportional to this measure (intensity parameter alpha>0). This Poissson point processes are called Poisson ensembles of Markov loops or loop soups. The measure on loops is covariant with some transformation on Markovian processes, for instance the change of time. In the setting of Brownian loop soups inside a proper open simply connected domain of C it was shown that the outer boundaries of outermost clusters of loops are, for alpha1/2, Conformal Loop Ensembles CLE(kappa), kappa in (8/3,4]. Besides, it was shown for a wide range of symmetric Markovian processes that for alpha=1/2 the occupation field of a loop soup (the sum of times spent by loops over points) is the square of the Gaussian free field. First I studied the loop soups associated to one-dimensional diffusions, and particularly the occupation field and its zeroes that delimit in this case the clusters of loops. Then I studied the loop soups on discrete graphs and metric graphs (edges replaced by continuous lines). On a metric graph on one hand the loops have a non-trivial geometry and on the other hand one has the same property as in the setting of one-dimensional diffusions that the zeroes of the occupation field delimit the clusters of loops. By combing metric graphs and the isomorphism with the Gaussian free field I have shown that alpha=1/2 is the critical parameter for random walk loop soup percolation on the discrete half-plane Z*N (existence or not of an infinite cluster of loops) and that for alpha<= 1/2 the scaling limit of outer boundaries of outermost clusters on Z*N is a CLE(kappa) on the continuum half plane.